text
stringlengths 649
4.42k
| synonym_substitution
stringlengths 759
4.5k
| butter_fingers
stringlengths 649
4.42k
| random_deletion
stringlengths 453
2.31k
| change_char_case
stringlengths 649
4.42k
| whitespace_perturbation
stringlengths 764
5.02k
| underscore_trick
stringlengths 649
4.42k
|
|---|---|---|---|---|---|---|
mathcal{O}}}$-homomorphism. By Theorem \[thm\_plain\], $\phi$ is a plain homomorphism, and hence, an analytic homomorphism for the canonical analytic structure. Thus, ${{\mathop{\mathbf{Hom}}}}({\ensuremath{\mathcal{O}}}^E,{\ensuremath{\mathcal{O}}}^F)={{\mathop{\mathbf{Hom}_{\ensuremath{\mathcal{O}}}}}}({\ensuremath{\mathcal{O}}}^E,{\ensuremath{\mathcal{O}}}^F)$.
\[thm\_coh\_unique\_anal\_str\] Let ${\ensuremath{\mathcal{S}}}$ be a coherent sheaf such that ${\mathop{\mathrm{depth}}\nolimits _{{{\mathop{\mathfrak{m}}}}_\zeta}} {\ensuremath{\mathcal{S}}}_\zeta>0$ for $\zeta\in{\mathop{\mathrm{supp}}}{\ensuremath{\mathcal{S}}}$. Then the minimal and the maximal analytic structures coincide, i.e., ${\ensuremath{\mathcal{S}}}$ has unique analytic structure.
Denote by $\Omega\subset\mathbb{C}^m$ the base of the sheaf ${\ensuremath{\mathcal{S}}}$. Let $E$ be a Banach space, $U\subset\Omega$ an open set, and $\phi:{\ensuremath{\mathcal{O}}}^E|_U\rightarrow{\ensuremath{\mathcal{O}}}^F|_U$ an ${\ensuremath{\mathcal{O}}}$-homomorphism. If $m=0$, the depth condition guarantees that ${\ensuremath{\mathcal{S}}}=0$ and the conclusion of the theorem follows. So, we may assume that $m\geq 1$.
Since ${\ensuremath{\mathcal{S}}}$ is a coherent sheaf, given $\zeta\in U$ there is an epimorphism $p:{\ensuremath{\mathcal{O}}}^n|_V\rightarrow{\ensuremath{\mathcal{S}}}|_V$, with $n<\infty$ and $V\subset U$, a suitable neighborhood of $\zeta$. By Theorem \[thm\_local\_thm\], we can assume that $\phi|_V$ factors through $p|_V$, i.e., there is an ${\ensuremath{\mathcal{O}}}$-homomorphism $\psi:{\ensuremath{\mathcal{O}}}^E|_V\rightarrow{\ensuremath{\mathcal{O}}}^n|_V$ with $\phi|_V=p|_V\psi$. Then, by Theorem \[thm\_plain\], $\psi$ is a plain homomorphism, and so, $\phi_\zeta\in{{\mathop{\mathbf{Hom}_\mathrm{min}}}}({\ensuremath{\mathcal{ | mathcal{O}}}$-homomorphism. By Theorem \[thm\_plain\ ], $ \phi$ is a plain homomorphism, and hence, an analytic homomorphism for the canonical analytic structure. therefore, $ { { \mathop{\mathbf{Hom}}}}({\ensuremath{\mathcal{O}}}^E,{\ensuremath{\mathcal{O}}}^F)={{\mathop{\mathbf{Hom}_{\ensuremath{\mathcal{O}}}}}}({\ensuremath{\mathcal{O}}}^E,{\ensuremath{\mathcal{O}}}^F)$.
\[thm\_coh\_unique\_anal\_str\ ] lease $ { \ensuremath{\mathcal{S}}}$ be a coherent sheaf such that $ { \mathop{\mathrm{depth}}\nolimits _ { { { \mathop{\mathfrak{m}}}}_\zeta } } { \ensuremath{\mathcal{S}}}_\zeta>0 $ for $ \zeta\in{\mathop{\mathrm{supp}}}{\ensuremath{\mathcal{S}}}$. Then the minimal and the maximal analytic structures concur, i.e., $ { \ensuremath{\mathcal{S}}}$ has singular analytic structure.
Denote by $ \Omega\subset\mathbb{C}^m$ the base of the bundle $ { \ensuremath{\mathcal{S}}}$. get $ E$ be a Banach space, $ U\subset\Omega$ an open set, and $ \phi:{\ensuremath{\mathcal{O}}}^E|_U\rightarrow{\ensuremath{\mathcal{O}}}^F|_U$ an $ { \ensuremath{\mathcal{O}}}$-homomorphism. If $ m=0 $, the depth circumstance guarantees that $ { \ensuremath{\mathcal{S}}}=0 $ and the conclusion of the theorem follows. So, we may wear that $ m\geq 1$.
Since $ { \ensuremath{\mathcal{S}}}$ is a coherent sheaf, given $ \zeta\in U$ there is an epimorphism $ p:{\ensuremath{\mathcal{O}}}^n|_V\rightarrow{\ensuremath{\mathcal{S}}}|_V$, with $ n<\infty$ and $ V\subset U$, a suitable neighborhood of $ \zeta$. By Theorem \[thm\_local\_thm\ ], we can bear that $ \phi|_V$ factors through $ p|_V$, i.e., there is an $ { \ensuremath{\mathcal{O}}}$-homomorphism $ \psi:{\ensuremath{\mathcal{O}}}^E|_V\rightarrow{\ensuremath{\mathcal{O}}}^n|_V$ with $ \phi|_V = p|_V\psi$. Then, by Theorem \[thm\_plain\ ], $ \psi$ is a plain homomorphism, and indeed, $ \phi_\zeta\in{{\mathop{\mathbf{Hom}_\mathrm{min}}}}({\ensuremath{\mathcal { | matjcal{O}}}$-homomorphism. By Theurem \[thm\_plain\], $\pku$ is a plain homomorohism, and hence, an analytic iomonorphusm for the canonical xnalytic dtructurw. Thns, ${{\mathop{\mathbf{Hom}}}}({\ensuremath{\mafmcal{O}}}^Z,{\eisuremath{\mathcak{O}}}^F)={{\mathop{\mdthbf{Hom}_{\ensurekagh{\lathcal{O}}}}}}({\ensuremath{\mathcal{O}}}^E,{\ensuremaeh{\mathcsl{L}}}^F)$.
\[thm\_coh\_unique\_wnal\_xer\] Lsn ${\tnsuremath{\mathcal{S}}}$ be a coherent aheaf slch that ${\mathop{\mayhrm{depth}}\nolimits _{{{\mathop{\mahhfrwk{m}}}}_\zeta}} {\ensuremath{\lathcal{S}}}_\zetq>0$ fow $\zeta\in{\mathoo{\mathrm{supk}}}{\eusuremath{\majhcal{S}}}$. Then the minimal and the oaximcl analytic srrufjures coincive, i.e., ${\vnsuremath{\matmbal{S}}}$ hav uniqur analytic strmcturx.
Denite by $\Omega\subset\matibb{C}^m$ the base of thg sheaf ${\envuxemath{\mathcal{S}}}$. Let $E$ ve a Batach spazw, $U\rubaev\Omsga$ an opxn set, and $\lhi:{\ensuremarh{\mathcal{O}}}^E|_U\rightartor{\vmsuremath{\matgcal{O}}}^F|_T$ wn ${\ensuremath{\mathcal{O}}}$-homomorphism. If $m=0$, uhe dspth condition guarantews that ${\ensuremath{\matjcal{S}}}=0$ and the conclusion of the theorem follows. So, we may dssumx ghau $n\geq 1$.
Runfe ${\ensuremath{\mathcal{S}}}$ is a coherent sheaf, givqh $\eets\in U$ there is an epimorpnidm $k:{\ensuremath{\matfcal{O}}}^n|_R\difhtarrow{\ensuremath{\lathcal{F}}}|_V$, wirh $n<\infty$ and $V\subset U$, a suitable neighvorhood of $\zvta$. Vy Theorem \[thm\_locap\_thm\], we can assome thst $\phi|_V$ factors through $p|_V$, j.e., there is an ${\ensursoath{\mathcal{O}}}$-homooorihisk $\psi:{\ensuremath{\mathcal{O}}}^E|_V\ryghtarrow{\xnsurzmath{\matfcal{P}}}^n|_V$ wieh $\phi|_V=p|_V\pdi$. Thct, by Theorem \[thm\_plwin\], $\pdi$ is a plaij homomorphism, and so, $\phi_\zeta\in{{\mathop{\mathbf{Hok}_\mdthgm{min}}}}({\ensuxemath{\kathcal{ | mathcal{O}}}$-homomorphism. By Theorem \[thm\_plain\], $\phi$ is a and an analytic for the canonical Let be a coherent such that ${\mathop{\mathrm{depth}}\nolimits {\ensuremath{\mathcal{S}}}_\zeta>0$ for $\zeta\in{\mathop{\mathrm{supp}}}{\ensuremath{\mathcal{S}}}$. Then the minimal the maximal analytic structures coincide, i.e., ${\ensuremath{\mathcal{S}}}$ has unique analytic structure. Denote by the base of the sheaf ${\ensuremath{\mathcal{S}}}$. Let $E$ be a Banach space, $U\subset\Omega$ open and an If $m=0$, the depth condition guarantees that ${\ensuremath{\mathcal{S}}}=0$ and the conclusion of the theorem follows. So, may assume that $m\geq 1$. Since ${\ensuremath{\mathcal{S}}}$ is coherent sheaf, given $\zeta\in there is an epimorphism $p:{\ensuremath{\mathcal{O}}}^n|_V\rightarrow{\ensuremath{\mathcal{S}}}|_V$, $n<\infty$ $V\subset U$, suitable of By Theorem \[thm\_local\_thm\], can assume that $\phi|_V$ factors through $p|_V$, i.e., there is an ${\ensuremath{\mathcal{O}}}$-homomorphism $\psi:{\ensuremath{\mathcal{O}}}^E|_V\rightarrow{\ensuremath{\mathcal{O}}}^n|_V$ with $\phi|_V=p|_V\psi$. Then, by \[thm\_plain\], $\psi$ plain homomorphism, so, | mathcal{O}}}$-homomorphism. By TheOrem \[thm\_plaIn\], $\phi$ Is a PlaIn HomoMorpHism, and hence, an ANalyTic homomorphism for the cAnoniCaL AnalYTiC struCture. ThUS, ${{\mATHop{\MaThBf{HOm}}}}({\ENsUremaTh{\mAthcal{O}}}^e,{\ensurematH{\maThCal{O}}}^F)={{\mathop{\mAThBf{Hom}_{\ensurEmaTh{\mathcal{O}}}}}}({\enSurEmath{\mAtHcaL{o}}}^E,{\ensUreMath{\mAthcal{o}}}^f)$.
\[thm\_coH\_unique\_anAl\_STr\] Let ${\eNSurematH{\MAtHcal{s}}}$ be a coherent sheaf SUcH That ${\mathop{\mathRm{deptH}}\nOLiMITs _{{{\mAthOp{\mathfrak{M}}}}_\zEta}} {\enSUremath{\MAtHCAL{S}}}_\zETa>0$ for $\zeta\in{\maThop{\mathrm{sUPp}}}{\eNsuremAtH{\maTHcal{S}}}$. THen thE mINimAl and the maxImal Analytic sTructuREs coincIDe, i.e., ${\ensUrematH{\maThcAl{S}}}$ hAS uNiQue AnALytIC sTruCTurE.
Denote bY $\OMeGa\subSet\mATHBB{C}^m$ tHe bAse oF the sHeaf ${\ensurematH{\maThcaL{s}}}$. LeT $E$ be a banacH spaCe, $u\subsEt\OmegA$ an opEn Set, and $\phi:{\ensureMath{\Mathcal{O}}}^E|_u\riGhTarRoW{\ensuREmath{\mAthCal{o}}}^F|_U$ an ${\enSurematH{\MatHcAL{o}}}$-HoMomorphism. If $m=0$, the dePtH COnDition guAranteES tHaT ${\EnsuremaTh{\MatHcal{s}}}=0$ ANd the ConcLUsIon of the TheoreM FoLlOws. So, we MaY assumE tHat $M\geQ 1$.
SincE ${\EnsuRemath{\Mathcal{S}}}$ Is a coHErent sheaf, giveN $\Zeta\in U$ there iS An EPImORphiSm $p:{\Ensuremath{\mAthcAL{O}}}^n|_V\RighTArRow{\ENsureMath{\mAtHCaL{s}}}|_V$, with $n<\infty$ and $V\subSeT U$, a suiTable Neighborhood oF $\zeta$. By TheOREM \[thm\_locaL\_thm\], WE cAN assume that $\phi|_v$ factOrs through $P|_v$, i.e., there Is an ${\eNsurematH{\mathcal{O}}}$-HOMomorphiSm $\pSi:{\eNsuRemATH{\mAthcal{O}}}^E|_V\righTARrow{\EnSurematH{\maThcal{O}}}^n|_v$ wiTh $\pHi|_V=P|_V\pSi$. then, by TheOrem \[thm\_pLaIn\], $\PsI$ iS a pLain hOMomorphiSm, And So, $\Phi_\Zeta\iN{{\Mathop{\MathbF{Hom}_\MaThRM{miN}}}}({\ensureMAtH{\MAthcAl{ | mathcal{O}}}$-homomorphism . By Theor em \[ thm \_p la in\] , $\ phi$ is a plai n hom omorphism, and hence,an an al y tich om omorp hism fo r t h e ca no ni cal a n al yticstr ucture. Thus, ${{ \ma th op{\mathbf{H o m} }}}({\ensu rem ath{\mathcal {O} }}^E,{ \e nsu r emath {\m athca l{O}}} ^ F)={{\ mathop{\m at h bf{Hom } _{\ensu r e ma th{\ mathcal{O}}}}}}({ \ en s uremath{\mathc al{O}} }^ E ,{ \ e nsu rem ath{\mathc al {O}}} ^ F)$.
\ [ th m \ _ coh \ _unique\_anal \_str\] Let ${\ ensure ma th{ \ mathca l{S}} }$ bea coherentshea f such th at ${\ m athop{\ m athrm{d epth}} \no lim its_ {{ {\ mat ho p {\m a th fra k {m} }}}_\zet a} }{\ens urem a t h { \mat hca l{S} }}_\z eta>0$ for $\ zet a\in { \ma thop{ \math rm{s up p}}}{ \ensur emath {\ mathcal{S}}}$.Then the mini mal a ndth e max i mal an aly tic struct ures co i nci de , i .e ., ${\ensuremath{\ ma t h ca l{S}}}$has un i qu ea nalyticst ruc ture .
Deno te b y $ \Omega\s ubset\ m at hb b{C}^m$ t he bas eofthe shea f ${\ ensure math{\ma thcal { S}}}$. Let $E$ be a Banach s p ac e , $ U \sub set \Omega$ anopen set, and $\ phi : {\ens urema th { \m a thcal{O}}}^E|_U\rig ht arrow{ \ensu remath{\mathc al{O}}}^F| _ U $ an ${\e nsur e ma t h{\mathcal{O}} }$-ho momorphism . If $m=0 $, th e depthcondition g uarantee s t hat ${ \en s u re math{\mathcal { S }}}= 0$ and th e c onclusi onofthe th eo rem follo ws. So,we m ay a ssu me th a t $m\geq 1 $.
S inc e ${\ e nsurem ath{\ math ca l{ S }}} $ is ac oh e r entsh ea f, g ive n$\zet a\in U$there i s an epim orp h ism$p :{ \ensure math{\mathcal {O }}}^n|_V\r ig hta rrow{\ e n suremath {\mathcal{S}}}|_V$, wit h $n<\in fty $ and $V\ subset U$ , a suita ble neighb orhood of $ \z eta $ . By T h e or em\[ thm\_local \ _ thm \], w ecanassumethat $\phi|_V$ fac t ors through $p|_ V$, i.e . , t her e i s an $ { \en s u remath{\mathcal {O}}}$-hom om o rp hism $\psi : {\e ns uremath {\mathc al{O} } }^E|_V\ rightarro w{\ensure ma th{\ m a thc al{O}}}^n| _V$ with $\phi|_V = p|_V\ p si $. Th en, by Th eo rem \[th m\_pla i n\] , $\p si$ is a plain homo mo rphism,and so, $\phi_\zeta\in{ {\math op{\m ath bf{Hom}_\ mat h rm{ min}}}}({ \ens uremath{\m ath cal { | mathcal{O}}}$-homomorphism. By_Theorem \[thm\_plain\],_$\phi$ is a plain_homomorphism, and_hence,_an analytic_homomorphism_for the canonical_analytic structure. Thus,_${{\mathop{\mathbf{Hom}}}}({\ensuremath{\mathcal{O}}}^E,{\ensuremath{\mathcal{O}}}^F)={{\mathop{\mathbf{Hom}_{\ensuremath{\mathcal{O}}}}}}({\ensuremath{\mathcal{O}}}^E,{\ensuremath{\mathcal{O}}}^F)$.
\[thm\_coh\_unique\_anal\_str\] Let ${\ensuremath{\mathcal{S}}}$ be_a coherent sheaf_such_that ${\mathop{\mathrm{depth}}\nolimits _{{{\mathop{\mathfrak{m}}}}_\zeta}} {\ensuremath{\mathcal{S}}}_\zeta>0$ for $\zeta\in{\mathop{\mathrm{supp}}}{\ensuremath{\mathcal{S}}}$. Then the minimal and the maximal analytic structures_coincide,_i.e., ${\ensuremath{\mathcal{S}}}$_has_unique_analytic structure.
Denote by $\Omega\subset\mathbb{C}^m$ the_base of the sheaf ${\ensuremath{\mathcal{S}}}$._Let $E$_be a Banach space, $U\subset\Omega$ an open set,_and_$\phi:{\ensuremath{\mathcal{O}}}^E|_U\rightarrow{\ensuremath{\mathcal{O}}}^F|_U$ an ${\ensuremath{\mathcal{O}}}$-homomorphism._If $m=0$, the depth condition guarantees that ${\ensuremath{\mathcal{S}}}=0$ and_the conclusion of the theorem follows._So, we may_assume_that_$m\geq 1$.
Since ${\ensuremath{\mathcal{S}}}$ is_a coherent sheaf, given $\zeta\in U$_there is an epimorphism $p:{\ensuremath{\mathcal{O}}}^n|_V\rightarrow{\ensuremath{\mathcal{S}}}|_V$, with_$n<\infty$ and $V\subset U$, a suitable neighborhood_of $\zeta$. By Theorem \[thm\_local\_thm\], we_can assume that $\phi|_V$ factors_through $p|_V$,_i.e., there is an ${\ensuremath{\mathcal{O}}}$-homomorphism_$\psi:{\ensuremath{\mathcal{O}}}^E|_V\rightarrow{\ensuremath{\mathcal{O}}}^n|_V$ with $\phi|_V=p|_V\psi$._Then, by_Theorem \[thm\_plain\], $\psi$_is a plain homomorphism, and so,_$\phi_\zeta\in{{\mathop{\mathbf{Hom}_\mathrm{min}}}}({\ensuremath{\mathcal{ |
E_m$ and $E_n$ are the corresponding unity matrices, and by ${\mathbf{e}}_1,\ldots,{\mathbf{e}}_m,\pmb\ell_{m+1},\ldots,\pmb\ell_d$ the columns of $$T'=
\begin{pmatrix}
E_m & {{\Theta}^\intercal} \\
0 & E_n
\end{pmatrix}.$$ We obviously have $T{{(T')}^\intercal}=E_d$, so the bases $\pmb\ell_1,\ldots,\pmb\ell_m,{\mathbf{e}}_{m+1},\ldots,{\mathbf{e}}_d$ and ${\mathbf{e}}_1,\ldots,{\mathbf{e}}_m,\pmb\ell_{m+1},\ldots,\pmb\ell_d$ are dual. Therefore, the subspaces $${\mathcal{L}}^m=\operatorname{span}_{\mathbb{R}}(\pmb\ell_1,\ldots,\pmb\ell_m),\qquad
{\mathcal{L}}^n=\operatorname{span}_{\mathbb{R}}(\pmb\ell_{m+1},\ldots,\pmb\ell_d)$$ are orthogonal. More than that, ${\mathcal{L}}^m=({\mathcal{L}}^n)^\perp$ and $${\mathcal{L}}^m=\{ {\mathbf{z}}\in{\mathbb{R}}^d \,|\, \langle\pmb\ell_{m+i},{\mathbf{z}}\rangle=0,\ i=1,\ldots,n \},\qquad
{\mathcal{L}}^n=\{ {\mathbf{z}}\in{\mathbb{R}}^d \,|\, \langle\pmb\ell_j,{\mathbf{z}}\rangle=0,\ j=1,\ldots,m \}.$$ Thus, ${\mathcal{L}}^m$ coincides with the space of solutions of the system $\Theta{\mathbf{x}}=-{\mathbf{y}}$, and ${\mathcal{L}}^n$ coincides with that of the system ${{\Theta}^\intercal}{\mathbf{y}}={\mathbf{x}}$. It can be easily verified that changing the sign in while preserving it for the transposed system does not affect any of the definitions and theorems given in Section \[sec:intro\].
For all positive $h$ and $r$ let us define parallelepipeds $$\begin{aligned}
M_{h,r}=\Big\{ {\mathbf{z}}\in{\mathbb{R}}^d \,\Big|\ | E_m$ and $ E_n$ are the corresponding unity matrices, and by $ { \mathbf{e}}_1,\ldots,{\mathbf{e}}_m,\pmb\ell_{m+1},\ldots,\pmb\ell_d$ the columns of $ $ T'=
\begin{pmatrix }
E_m & { { \Theta}^\intercal } \\
0 & E_n
\end{pmatrix}.$$ We obviously get $ T{{(T')}^\intercal}=E_d$, so the base $ \pmb\ell_1,\ldots,\pmb\ell_m,{\mathbf{e}}_{m+1},\ldots,{\mathbf{e}}_d$ and $ { \mathbf{e}}_1,\ldots,{\mathbf{e}}_m,\pmb\ell_{m+1},\ldots,\pmb\ell_d$ are dual. Therefore, the subspaces $ $ { \mathcal{L}}^m=\operatorname{span}_{\mathbb{R}}(\pmb\ell_1,\ldots,\pmb\ell_m),\qquad
{ \mathcal{L}}^n=\operatorname{span}_{\mathbb{R}}(\pmb\ell_{m+1},\ldots,\pmb\ell_d)$$ are extraneous. More than that, $ { \mathcal{L}}^m=({\mathcal{L}}^n)^\perp$ and $ $ { \mathcal{L}}^m=\ { { \mathbf{z}}\in{\mathbb{R}}^d \,|\, \langle\pmb\ell_{m+i},{\mathbf{z}}\rangle=0,\ i=1,\ldots, n \},\qquad
{ \mathcal{L}}^n=\ { { \mathbf{z}}\in{\mathbb{R}}^d \,|\, \langle\pmb\ell_j,{\mathbf{z}}\rangle=0,\ j=1,\ldots, m \}.$$ Thus, $ { \mathcal{L}}^m$ coincides with the space of solution of the system $ \Theta{\mathbf{x}}=-{\mathbf{y}}$, and $ { \mathcal{L}}^n$ coincides with that of the arrangement $ { { \Theta}^\intercal}{\mathbf{y}}={\mathbf{x}}$. It can be easily verified that change the sign in while preserving it for the transposed system does not affect any of the definitions and theorem given in Section \[sec: intro\ ].
For all positive $ h$ and $ r$ let us specify parallelepipeds $ $ \begin{aligned }
M_{h, r}=\Big\ { { \mathbf{z}}\in{\mathbb{R}}^d \,\Big|\ | E_m$ wnd $E_n$ are the corresponaing unity matrnxes, anv by ${\mafhbf{e}}_1,\ldogs,{\mathbf{e}}_m,\pmb\ell_{m+1},\ldots,\pmb\ell_v$ thw colymns of $$T'=
\begin{pmagrix}
E_m & {{\Thwta}^\iitercal} \\
0 & E_n
\cud{pmafvix}.$$ Wz ibviously have $T{{(T')}^\intercan}=E_d$, so the basas $\plb\ell_1,\ldots,\pmb\ell_m,{\mathbf{e}}_{m+1},\ldots,{\mathbs{e}}_d$ and ${\mwthbf{e}}_1,\ldots,{\matrbf{e}}_k,\[mb\eml_{m+1},\ldots,\pmb\ell_d$ are dual. Therefore, the suuspaces $${\mathcal{K}}^m=\operatorname{span}_{\mathbb{R}}(\plb\elp_1,\ldots,\pmb\ell_m),\qquad
{\mathcao{L}}^n=\o[wratorname{spxn}_{\mathbb{R}}(\pmb\ell_{m+1},\ldots,\pjb\ell_d)$$ are orthogonal. More than that, ${\mathcal{L}}^m=({\narhcwn{L}}^n)^\perp$ and $${\mathbal{L}}^m=\{ {\mathbf{z}}\lm{\mathbt{R}}^d \,|\, \lamgle\pmb\ell_{m+i},{\mabhbf{z}}\cangoe=0,\ i=1,\ldots,n \},\qquad
{\kathcal{L}}^n=\{ {\mathbf{z}}\ig{\mathbb{R}}^d \,|\, \mangle\pmb\ell_j,{\mathvf{z}}\rangne=0,\ j=1,\ndotr,n \}.$$ Ghua, ${\kafhcal{L}}^l$ ckincides wjth the spaxe of solutions of uhe wystem $\Theta{\mzthbf{x}}=-{\iaehbf{y}}$, and ${\mathcal{L}}^n$ coincides with that of the system ${{\Theta}^\intercql}{\mathbf{y}}={\mathbf{x}}$. It cwn be easyly verified that changing the sign in while presarvinj kt yir thd tgansposed system does not affect any of the dqriminions and theoremf given in Xeftojn \[sec:intro\].
Fot all posifive $h$ and $r$ let ud defing paraolelepipeqs $$\brgin{aligned}
M_{h,r}=\Big\{ {\mathbd{z}}\in{\mathbb{R}}^d \,\Vig|\ | E_m$ and $E_n$ are the corresponding unity by the columns $$T'= \begin{pmatrix} E_m E_n We obviously have so the bases and ${\mathbf{e}}_1,\ldots,{\mathbf{e}}_m,\pmb\ell_{m+1},\ldots,\pmb\ell_d$ are dual. Therefore, the $${\mathcal{L}}^m=\operatorname{span}_{\mathbb{R}}(\pmb\ell_1,\ldots,\pmb\ell_m),\qquad {\mathcal{L}}^n=\operatorname{span}_{\mathbb{R}}(\pmb\ell_{m+1},\ldots,\pmb\ell_d)$$ are orthogonal. More than that, ${\mathcal{L}}^m=({\mathcal{L}}^n)^\perp$ and $${\mathcal{L}}^m=\{ {\mathbf{z}}\in{\mathbb{R}}^d \,|\, \langle\pmb\ell_{m+i},{\mathbf{z}}\rangle=0,\ \},\qquad {\mathcal{L}}^n=\{ {\mathbf{z}}\in{\mathbb{R}}^d \,|\, \langle\pmb\ell_j,{\mathbf{z}}\rangle=0,\ j=1,\ldots,m \}.$$ Thus, ${\mathcal{L}}^m$ coincides with the space solutions the $\Theta{\mathbf{x}}=-{\mathbf{y}}$, ${\mathcal{L}}^n$ coincides with that of the system ${{\Theta}^\intercal}{\mathbf{y}}={\mathbf{x}}$. It can be easily verified that changing the in while preserving it for the transposed system not affect any of definitions and theorems given in \[sec:intro\]. all positive and let define parallelepipeds $$\begin{aligned} {\mathbf{z}}\in{\mathbb{R}}^d \,\Big|\ | E_m$ and $E_n$ are the correspondinG unity matrIces, aNd bY ${\maThBf{e}}_1,\lDots,{\Mathbf{e}}_m,\pmb\ell_{M+1},\LdotS,\pmb\ell_d$ the columns of $$T'=
\bEgin{pMaTRix}
E_M & {{\thEta}^\inTercal} \\
0 & E_N
\EnD{PMatRiX}.$$ WE obViOUsLy havE $T{{(T')}^\IntercaL}=E_d$, so the baSes $\PmB\ell_1,\ldots,\pmb\ELl_M,{\mathbf{e}}_{m+1},\lDotS,{\mathbf{e}}_d$ and ${\MatHbf{e}}_1,\ldOtS,{\maTHbf{e}}_m,\Pmb\Ell_{m+1},\lDots,\pmB\Ell_d$ arE dual. TherEfORe, the sUBspaces $${\MAThCal{L}}^M=\operatorname{span}_{\MAtHBb{R}}(\pmb\ell_1,\ldots,\Pmb\ell_M),\qQUaD
{\MAthCal{l}}^n=\operatorNaMe{spaN}_{\Mathbb{R}}(\PMb\ELL_{M+1},\ldOTs,\pmb\ell_d)$$ are oRthogonal. MoRE thAn that, ${\MaThcAL{L}}^m=({\matHcal{L}}^N)^\pERp$ aNd $${\mathcal{L}}^m=\{ {\MathBf{z}}\in{\mathBb{R}}^d \,|\, \laNGle\pmb\eLL_{m+i},{\mathBf{z}}\ranGle=0,\ I=1,\ldOts,n \},\QQuAd
{\MatHcAL{L}}^n=\{ {\MAtHbf{Z}}\In{\mAthbb{R}}^d \,|\, \lAnGlE\pmb\eLl_j,{\mATHBF{z}}\raNglE=0,\ j=1,\ldOts,m \}.$$ THus, ${\mathcal{L}}^m$ cOinCideS WitH the sPace oF solUtIons oF the syStem $\THeTa{\mathbf{x}}=-{\mathbf{Y}}$, and ${\Mathcal{L}}^n$ CoiNcIdeS wIth thAT of the SysTem ${{\theta}^\inTercal}{\mAThbF{y}}={\MATHbF{x}}$. It can be easily verIfIED tHat changIng the SIgN iN While preSeRviNg it FOR the tRansPOsEd system Does noT AfFeCt any of ThE definItIonS anD theoREms gIven in section \[sEc:intRO\].
For all positivE $H$ and $r$ let us defINe PARaLLelePipEds $$\begin{aliGned}
m_{H,r}=\BiG\{ {\matHBf{Z}}\in{\MAthbb{r}}^d \,\Big|\ | E_m$ and $E_n$ are the cor responding unit y m atr ic es,andby ${\mathbf{e } }_1, \ldots,{\mathbf{e}}_m, \pmb\ el l _{m+ 1 }, \ldot s,\pmb\ e ll _ d $ t he c olu mn s o f $$T '= \b egin{pmatr ix}
E_m &{ {\ Theta}^\in ter cal} \\
0 & E_ n \end {pm atrix }.$$ W e obvio usly have $ T {{(T') } ^\inter c a l} =E_d $, so the bases $ \ pm b \ell_1,\ldots, \pmb\e ll _ m, { \ mat hbf {e}}_{m+1} ,\ ldots , {\mathb f {e } } _ d$a nd ${\mathbf{ e}}_1,\ldot s ,{\ mathbf {e }}_ m ,\pmb\ ell_{ m+ 1 },\ ldots,\pmb\ ell_ d$ are du al. Th e refore, the sub spaces $$ {\m athc a l{ L} }^m =\ o per a to rna m e{s pan}_{\m at hb b{R}} (\pm b \ e l l_1, \ld ots, \pmb\ ell_m),\qquad
{ \ mat hcal{ L}}^n =\op er atorn ame{sp an}_{ \m athbb{R}}(\pmb\ ell_ {m+1},\ld ots ,\ pmb \e ll_d) $ $ areort hog onal. M ore tha n th at , $ {\ mathcal{L}}^m=({\m at h c al {L}}^n)^ \perp$ an d$ ${\mathc al {L} }^m= \ { {\ma thbf { z} }\in{\ma thbb{R } }^ d\,|\, \ la ngle\p mb \el l_{ m+i}, { \mat hbf{z} }\rangle =0,\i =1,\ldots,n \} , \qquad
{ \ ma t h ca l {L}} ^n= \{ {\mathbf {z}} \ in{\ math b b{ R}} ^ d \,| \, \l an g le \ pmb\ell_j,{\mathbf{ z} }\rang le=0, \ j=1,\ldots, m \}.$$ Th u s , ${\math cal{ L }} ^ m$ coincides w ith t he space o f solutio ns of the sys tem $\The t a {\mathbf {x} }=- {\m ath b f {y }}$, and ${\m a t hcal {L }}^n$ c oin cides w ith th atofth e system${{\Thet a} ^\ in te rca l}{\m a thbf{y}} ={ \ma th bf{ x}}$. It can be e asil yve r ifi ed that ch a n ging t he sig n i nwhile pre s erv ing itfor the t ran s pose dsy stem do es not affect a ny of thede fin itions a nd theor ems given in Section \[ s ec:intr o\] .
Fo r al l positiv e $ h$ and $r $ let u s defi ne pa ra lle l e piped s $$ \be gi n{aligned} M_{h, r} =\Bi g\{ {\m athbf{z}}\in{\math b b{R }}^d \,\Big|\ | E_m$ and_$E_n$ are_the corresponding unity matrices,_and by_${\mathbf{e}}_1,\ldots,{\mathbf{e}}_m,\pmb\ell_{m+1},\ldots,\pmb\ell_d$_the columns_of_$$T'=
_ \begin{pmatrix}
_ _ E_m_&_{{\Theta}^\intercal} \\
0 & E_n
__\end{pmatrix}.$$ We_obviously_have_$T{{(T')}^\intercal}=E_d$, so the bases $\pmb\ell_1,\ldots,\pmb\ell_m,{\mathbf{e}}_{m+1},\ldots,{\mathbf{e}}_d$_and ${\mathbf{e}}_1,\ldots,{\mathbf{e}}_m,\pmb\ell_{m+1},\ldots,\pmb\ell_d$ are dual. Therefore,_the subspaces_$${\mathcal{L}}^m=\operatorname{span}_{\mathbb{R}}(\pmb\ell_1,\ldots,\pmb\ell_m),\qquad
{\mathcal{L}}^n=\operatorname{span}_{\mathbb{R}}(\pmb\ell_{m+1},\ldots,\pmb\ell_d)$$ are orthogonal._More_than that, ${\mathcal{L}}^m=({\mathcal{L}}^n)^\perp$_and $${\mathcal{L}}^m=\{ {\mathbf{z}}\in{\mathbb{R}}^d \,|\, \langle\pmb\ell_{m+i},{\mathbf{z}}\rangle=0,\ i=1,\ldots,n \},\qquad
_ {\mathcal{L}}^n=\{ {\mathbf{z}}\in{\mathbb{R}}^d \,|\, \langle\pmb\ell_j,{\mathbf{z}}\rangle=0,\_j=1,\ldots,m \}.$$ Thus,_${\mathcal{L}}^m$_coincides_with the space of_solutions of the system $\Theta{\mathbf{x}}=-{\mathbf{y}}$, and_${\mathcal{L}}^n$ coincides with that of the_system ${{\Theta}^\intercal}{\mathbf{y}}={\mathbf{x}}$. It can be easily verified_that changing the sign in while_preserving it for the transposed_system does_not affect any of the_definitions and theorems_given in_Section \[sec:intro\].
For all_positive $h$ and $r$ let us_define parallelepipeds $$\begin{aligned}
_ M_{h,r}=\Big\{ {\mathbf{z}}\in{\mathbb{R}}^d \,\Big|\ |
a pair of SSYT of shape $\mu$ (with the same weight) and an elegant filling (see Definition \[elegant\]) of shape $\lambda/\mu$.
Note that each nonintersecting lattice path system from $A$ to $B$ (geometrically one can see here that each $A_i$ should go to $B_i$) decomposes into two parts: below the $y = 0$ line (‘underwater’) and the upper part having weights $x_i$. (For instance, in Figure \[fff\] we have the weight $x_1 x_2^2 x_3 x_4$.) The upper part can be translated into an SSYT of some shape $\mu \subseteq \lambda$; and the ‘underwater’ part into an elegant tableau of the remaining shape $\lambda/\mu$ which repeats the decomposition. So, $g_{\lambda} = \det\left[ w(A_i \to B_j)\right]_{1 \le i,j \le \ell}$. $\square$
Instead of counting elegant tableaux with all weights $1$, we can similarly give weights to numbers in these fillings, or equivalently give some weights to the ‘underwater’ part. So, if we put the weight $t_i$ to the number $i$ in an elegant tableaux or to diagonal steps ‘underwater’, the resulting function will correspond to the [*refined*]{} dual stable Grothendieck polynomials $\widetilde{g}_{\lambda/\mu}(x,t)$ introduced by Galashin, Grinberg, and Liu in [@ggl] where they gave Bender-Knuth type involutions for proving that $g_{\lambda/\mu}$ is a symmetric function.
From our setting, the Schur decomposition rewrites with $t = (t_1, t_2, \ldots)$ parameters as follows $$\widetilde{g}_{\lambda}(x,t) = \sum_{\mu} f_{\lambda/\mu}(t) s_{\mu},$$ where $f_{\lambda/\mu}(t) = \sum_{\text{elegant tableaux of } \lambda/\mu} \prod t_i^{a_i}$ ($a_i$ is the number of times $i$ occurs in an elegant tableau). By the same argument as the proof above (but with the $t$ parameters added), we obtain the following | a pair of SSYT of shape $ \mu$ (with the same weight) and an elegant woof (examine Definition \[elegant\ ]) of shape $ \lambda/\mu$.
Note that each nonintersecting wicket path organization from $ A$ to $ B$ (geometrically one can see here that each $ A_i$ should go to $ B_i$) decomposes into two region: below the $ y = 0 $ line (‘ underwater ’) and the upper part have weights $ x_i$. (For instance, in Figure \[fff\ ] we get the weight $ x_1 x_2 ^ 2 x_3 x_4$.) The upper part can be translated into an SSYT of some shape $ \mu \subseteq \lambda$; and the ‘ underwater ’ part into an elegant tableau of the remaining human body $ \lambda/\mu$ which repeats the decomposition. So, $ g_{\lambda } = \det\left [ w(A_i \to B_j)\right]_{1 \le i, j \le \ell}$. $ \square$
rather of counting elegant tableaux with all weights $ 1 $, we can similarly give weight to numbers in these fillings, or equivalently give some weights to the ‘ underwater ’ function. So, if we put the weight $ t_i$ to the number $ i$ in an elegant tableaux or to aslant tone ‘ underwater ’, the resulting function will correspond to the [ * refined * ] { } dual stable Grothendieck polynomials $ \widetilde{g}_{\lambda/\mu}(x, t)$ introduced by Galashin, Grinberg, and Liu in [ @ggl ] where they gave Bender - Knuth type involution for proving that $ g_{\lambda/\mu}$ is a symmetric routine.
From our place setting, the Schur decomposition rewrites with $ t = (t_1, t_2, \ldots)$ argument as follows $ $ \widetilde{g}_{\lambda}(x, t) = \sum_{\mu } f_{\lambda/\mu}(t) s_{\mu},$$ where $ f_{\lambda/\mu}(t) = \sum_{\text{elegant tableaux of } \lambda/\mu } \prod t_i^{a_i}$ ($ a_i$ is the number of times $ i$ occurs in an elegant tableau). By the like argument as the proof above (but with the $ t$ parameters added), we obtain the following | a oair of SSYT of shape $\mu$ (with the same cwight) end an slegant willing (see Definition \[elegait\]) od shakv $\lambda/\mu$.
Note that exch noninnersectint lautice path system from $A$ bj $B$ (fcometxirally one can sge here that each $A_i$ shoulg eo to $B_i$) decomposes into two parts: bejow the $y = 0$ line (‘underwwter’) wnd fhe upper part having weights $x_i$. (Fkr instence, in Figure \[gff\] we have the weight $x_1 x_2^2 x_3 x_4$.) The upper part caj be translqted unto an SSYT of some shape $\mu \subsgteq \lambda$; and the ‘underwater’ pxrt iuto an elegqnr twtleau of thx remapning shape $\lambda/\mu$ fhich rrpeats the decpmpmsirion. So, $g_{\lambda} = \det\lxft[ w(A_i \to B_j)\right]_{1 \lg i,j \le \eln}$. $\aquare$
Instead of xoyntinc elagang taclezue wjth alp wxights $1$, we dan similaroy give weights to muivers in these filliggf, or equivalently give some weights to uhe ‘uhderwater’ part. So, if we put the weight $t_i$ to the numbqr $i$ in an elegant tableaux or to diagonal steps ‘gnderxager’, tme rdwupting function will correspond to the [*refined*]{} susl stable Grothekdieck polynomials $\wldrjilde{g}_{\lambda/\mu}(b,t)$ intxkdhced by Galashin, Gginberg, and Oiu in [@ggj] whrre they gave Bender-Knuth ttpe involutijbs for proving thac $g_{\lambda/\mu}$ ns a simmetroc function.
From our setcing, tge Schur defompositikv rewrites with $g = (n_1, t_2, \ndots)$ parameters as followf $$\widetilve{g}_{\lakbda}(x,t) = \sum_{\ku} f_{\laibda/\mu}(t) s_{\mk},$$ wheva $f_{\lambda/\mu}(t) = \sum_{\tfxt{elggant dableaux ov } \lambda/\mu} \prod t_i^{a_i}$ ($a_i$ is the number of timgs $h$ obcurs in cn elenant tableau). By the same argukent as the oroof abovv (but witi the $t$ paraieters added), fg obtain the hollowing | a pair of SSYT of shape $\mu$ same and an filling (see Definition that nonintersecting lattice path from $A$ to (geometrically one can see here that $A_i$ should go to $B_i$) decomposes into two parts: below the $y = line (‘underwater’) and the upper part having weights $x_i$. (For instance, in Figure we the $x_1 x_3 x_4$.) The upper part can be translated into an SSYT of some shape $\mu \subseteq and the ‘underwater’ part into an elegant tableau the remaining shape $\lambda/\mu$ repeats the decomposition. So, $g_{\lambda} \det\left[ \to B_j)\right]_{1 i,j \ell}$. Instead of counting tableaux with all weights $1$, we can similarly give weights to numbers in these fillings, or equivalently some weights ‘underwater’ part. if put weight $t_i$ to $i$ in an elegant tableaux or ‘underwater’, the resulting function will correspond to the dual stable polynomials $\widetilde{g}_{\lambda/\mu}(x,t)$ introduced by Galashin, Grinberg, Liu in [@ggl] where they gave Bender-Knuth type for proving that $g_{\lambda/\mu}$ is a symmetric function. From our setting, the Schur decomposition rewrites = (t_1, t_2, \ldots)$ as follows $$\widetilde{g}_{\lambda}(x,t) \sum_{\mu} s_{\mu},$$ $f_{\lambda/\mu}(t) \sum_{\text{elegant tableaux } \lambda/\mu} \prod t_i^{a_i}$ ($a_i$ is the number of times $i$ in an elegant tableau). By the same argument as the (but the $t$ parameters we obtain the following | a pair of SSYT of shape $\mu$ (with tHe same weigHt) and An eLegAnT filLing (See Definition \[eLEganT\]) of shape $\lambda/\mu$.
Note thAt eacH nONintERsEctinG latticE PaTH SysTeM fRom $a$ tO $b$ (gEometRicAlly one Can see here ThaT eAch $A_i$ should gO To $b_i$) decomposEs iNto two parts: bEloW the $y = 0$ lInE (‘unDErwatEr’) aNd the Upper pARt haviNg weights $X_i$. (fOr instANce, in FiGURe \[Fff\] wE have the weight $x_1 x_2^2 x_3 X_4$.) thE Upper part can be TranslAtED iNTO an sSYt of some shaPe $\Mu \subSEteq \lamBDa$; AND The ‘UNderwater’ part Into an elegaNT taBleau oF tHe rEMaininG shapE $\lAMbdA/\mu$ which repEats The decompOsitioN. so, $g_{\lambDA} = \det\lefT[ w(A_i \to b_j)\rIghT]_{1 \le i,J \Le \ElL}$. $\sqUaRE$
InSTeAd oF CouNting eleGaNt TableAux wITH ALl weIghTs $1$, we Can siMilarly give weIghTs to NUmbErs in These FillInGs, or eQuivalEntly GiVe some weights to The ‘uNderwater’ ParT. SO, if We Put thE Weight $T_i$ tO thE number $I$ in an elEGanT tABLEaUx or to diagonal stepS ‘uNDErWater’, the ResultINg FuNCtion wilL cOrrEspoND To the [*RefiNEd*]{} Dual stabLe GrotHEnDiEck polyNoMials $\wIdEtiLde{G}_{\lambDA/\mu}(x,T)$ introDuced by GAlashIN, Grinberg, and LiU In [@ggl] where theY GaVE beNDer-KNutH type involuTionS For pRoviNG tHat $G_{\LambdA/\mu}$ is A sYMmETric function.
From our SeTting, tHe SchUr decompositiOn rewrites WITH $t = (t_1, t_2, \ldotS)$ parAMeTErs as follows $$\wiDetilDe{g}_{\lambda}(x,T) = \Sum_{\mu} f_{\laMbda/\mU}(t) s_{\mu},$$ wheRe $f_{\lambda/\MU}(T) = \sum_{\text{EleGanT taBleAUX oF } \lambda/\mu} \prod T_I^{A_i}$ ($a_i$ Is The numbEr oF times $i$ OccUrs In aN elEgAnt tableaU). By the saMe ArGuMeNt aS the pROof above (BuT wiTh The $T$ paraMEters aDded), wE obtAiN tHE foLlowing | a pair of SSYT of shape $ \mu$ (with thesam e w ei ght) and an elegant fi l ling (see Definition \[ele gant\ ]) of s h ap e $\l ambda/\ m u$ .
No te t hat e a ch noni nte rsectin g latticepat hsystem from$ A$ to $B$ (g eom etrically on e c an see h ere thateac h $A_ i$ sho u ld goto $B_i$) d e compos e s intot w opart s: below the $y = 0$ line (‘underwa ter’)an d t h e up per part havi ng weig h ts $x_i $ .( F o r i n stance, in Fi gure \[fff\ ] we haveth e w e ight $ x_1 x _2 ^ 2 x _3 x_4$.) T he u pper part can b e transl a ted int o an S SYT of som e s ha pe$\ m u \ s ub set e q \ lambda$; a nd the‘und e r w a ter’ pa rt i nto a n elegant tab lea u of the rema ining sha pe $\la mbda/\ mu$ w hi ch repeats thedeco mposition . S o, $g _{ \lamb d a} = \ det \le ft[ w(A _i \toB _j) \r i g h t] _{1 \le i,j \le \e ll } $ .$\square $
Ins t ea do f counti ng el egan t table auxw it h all we ights$ 1$ ,we cansi milarl ygiv e w eight s tonumber s in the se fi l lings, or equi v alently gives om e we i ghts to the ‘under wate r ’ pa rt.S o, if we pu t the w e ig h t $t_i$ to the numb er $i$ i n anelegant table aux or tod i a gonal st eps‘ un d erwater’, theresul ting funct i on willcorre spond to the [*re f i ned*]{}dua l s tab leG r ot hendieck poly n o mial s$\widet ild e{g}_{\ lam bda /\m u}( x, t)$ intro duced by G al as hi n,Grinb e rg, andLi u i n[@g gl] w h ere th ey ga ve B en de r -Kn uth typ e i n v olut io ns for pr ov ing t hat$ g_{ \lambda /\mu}$ is as ymme tr ic functi on.
From our s etting, th eSch ur dec o m position rewrites with $t = (t_ 1 , t_2,\ld ots)$ par ameters a s f ollows $$ \ wideti lde{g} _{\la mb da} ( x ,t) = \ su m_{ \m u} f_{\lam b d a/\ mu}(t )s_{\ mu},$$where $f_{\lambda/ \ mu} (t) = \sum_{\ tex t{el e g an t t a bl e aux o f }\ l ambda/\mu} \pro d t_i^{a_i }$ ($ a_i$ is th e nu mb er of t imes $i $ occ u rs in a n elegant tableau) .By t h e sa me argumen t as the proof ab o ve (b u twiththe $t$ p ar ame tersadded) , we obta in the f ollowi ng | a_pair of_SSYT of shape $\mu$_(with the_same_weight) and_an_elegant filling (see_Definition \[elegant\]) of_shape $\lambda/\mu$.
Note that each_nonintersecting lattice path_system_from $A$ to $B$ (geometrically one can see here that each $A_i$ should go_to_$B_i$) decomposes_into_two_parts: below the $y =_0$ line (‘underwater’) and the_upper part_having weights $x_i$. (For instance, in Figure \[fff\]_we_have the weight_$x_1 x_2^2 x_3 x_4$.) The upper part can be_translated into an SSYT of some_shape $\mu \subseteq_\lambda$;_and_the ‘underwater’ part into_an elegant tableau of the remaining_shape $\lambda/\mu$ which repeats the decomposition._So, $g_{\lambda} = \det\left[ w(A_i \to B_j)\right]_{1_\le i,j \le \ell}$. $\square$
Instead of_counting elegant tableaux with all_weights $1$,_we can similarly give weights_to numbers in_these fillings,_or equivalently give_some weights to the ‘underwater’ part._So, if we_put the weight $t_i$ to the_number_$i$ in an_elegant_tableaux_or to_diagonal steps ‘underwater’,_the_resulting function_will_correspond to the [*refined*]{} dual stable_Grothendieck_polynomials $\widetilde{g}_{\lambda/\mu}(x,t)$ introduced by Galashin, Grinberg, and_Liu in [@ggl] where_they_gave Bender-Knuth type involutions_for proving that $g_{\lambda/\mu}$ is_a symmetric function.
From our setting, the_Schur decomposition_rewrites with_$t = (t_1, t_2, \ldots)$ parameters as follows $$\widetilde{g}_{\lambda}(x,t) = \sum_{\mu}_f_{\lambda/\mu}(t) s_{\mu},$$ where $f_{\lambda/\mu}(t) = \sum_{\text{elegant_tableaux of } \lambda/\mu}_\prod t_i^{a_i}$_($a_i$_is the number_of_times $i$_occurs in an elegant tableau). By the_same argument_as the proof above (but with_the $t$ parameters added),_we_obtain the following |
'$ in $\Gamma$. Since the vertex $v$ is not $(\epsilon, R)$-deep, by Theorem \[thm:stronglythin\] applied to $u,u', a$ (respectively $u,u',b$), we obtain a vertex $v_a$ (respectively $v_b$) on a geodesic in $\Gamma$ from $a$ (respectively $b$) to $u$ or $u'$ satisfying $|v_a,v|_S\leq D$ (respectively $|v_b,v|_S\leq D$). Consequently $|a,b|_S\leq
|a,v_a|_S+|v_a,v|_S+|v,v_b|_S+|v_b,b|_S\leq 2\rho(U)+2D\leq
4D$, as desired.
Henceforth, we assume $\rho(U)\geq D+1$. Let $\ell=|a,b|_S$. If $\ell> 4D$, then choose any $2D\leq i\leq \ell-2D$, and any geodesic $p^{ab}$ from $a$ to $b$ in $\Gamma$. If $p^{ab}_i$ is an $(\epsilon, R)$-deep vertex of $p^{ab}$ in the peripheral left coset $w$, then let $z=w$, otherwise let $z=p^{ab}_i$. We claim that for any $c\in U$ we have $|z,c|_S\leq \rho(U)-1$, contradicting the definition of $\rho(U)$, and implying $\ell\leq 4D$.
Indeed, we apply Theorem \[thm:stronglythin\] to $a,b,c,$ and any geodesics $p^{bc},p^{ac}$. Without loss of generality assume that we have $|z,p^{ac}_{i}|_S\leq D$. Note that if $i>|a,c|_S$, then $p^{ac}_i$ lies in (or is equal to) $c$, and consequently $|z,c|_S\leq |z,p^{ac}_{i}|_S\leq D\leq\rho(U)-1$, as desired. If $i\leq|a,c|_S$, then $|p^{ac}_i,a|_S=i\geq 2D$, and hence $$|z, | ' $ in $ \Gamma$. Since the vertex $ v$ is not $ (\epsilon, R)$-deep, by Theorem \[thm: stronglythin\ ] applied to $ u, u', a$ (respectively $ u, u',b$), we receive a vertex $ v_a$ (respectively $ v_b$) on a geodesic in $ \Gamma$ from $ a$ (respectively $ b$) to $ u$ or $ u'$ meet $ |v_a, v|_S\leq D$ (respectively $ |v_b, v|_S\leq D$). Consequently $ |a, b|_S\leq
|a, v_a|_S+|v_a, v|_S+|v, v_b|_S+|v_b, b|_S\leq 2\rho(U)+2D\leq
4D$, as desired.
Henceforth, we assume $ \rho(U)\geq D+1$. Let $ \ell=|a, b|_S$. If $ \ell > 4D$, then choose any $ 2D\leq i\leq \ell-2D$, and any geodesic $ p^{ab}$ from $ a$ to $ b$ in $ \Gamma$. If $ p^{ab}_i$ is an $ (\epsilon, R)$-deep vertex of $ p^{ab}$ in the peripheral leftover coset $ w$, then let $ z = w$, otherwise permit $ z = p^{ab}_i$. We claim that for any $ c\in U$ we have $ |z, c|_S\leq \rho(U)-1 $, contradict the definition of $ \rho(U)$, and implying $ \ell\leq 4D$.
Indeed, we apply Theorem \[thm: stronglythin\ ] to $ a, b, c,$ and any geodesic $ p^{bc},p^{ac}$. Without loss of generality assume that we have $ |z, p^{ac}_{i}|_S\leq D$. Note that if $ i>|a, c|_S$, then $ p^{ac}_i$ lie down in (or is equal to) $ c$, and consequently $ |z, c|_S\leq |z, p^{ac}_{i}|_S\leq D\leq\rho(U)-1 $, as desired. If $ i\leq|a, c|_S$, then $ |p^{ac}_i, a|_S = i\geq 2D$, and hence $ $ |z, | '$ in $\Hamma$. Since the vertex $v$ is not $(\epsilon, R)$-deep, uy Theodem \[thm:stfonglythin\] applied to $u,u', a$ (rxspextiveoy $u,u',b$), we obtain a vergex $v_a$ (redpectiveoy $v_u$) on a geodesic mh $\Gamma$ from $z$ (res'ertively $b$) to $u$ pr $u'$ satisxying $|v_a,v|_S\leq G$ (fedpectively $|v_b,v|_S\leq D$). Consequently $|a,f|_S\leq
|a,v_s|_S+|g_a,v|_S+|v,v_b|_S+|v_b,b|_S\lez 2\rhp(T)+2D\les
4D$, as desired.
Henceforth, we assume $\dho(U)\geq D+1$. Let $\ell=|a,b|_S$. If $\ell> 4D$, then choose any $2F\leq i\leq \ell-2D$, and any geodesic $p^{qb}$ fwim $a$ to $b$ in $\Gaoma$. If $p^{ab}_p$ is an $(\epsjlon, R)$-deep vertex of $p^{ab}$ in the peri'heral left ciseh $f$, then let $v=w$, othvrwise let $z=p^{ab}_i$. We cnaim thst for any $c\in U$ xe hqve $|z,c|_S\leq \rho(U)-1$, contredicting the definityon of $\rhm(U)$, and implying $\ell\oew 4D$.
Ingeed, we xpplh Tgeprsm \[thm:shroiglythin\] to $a,b,c,$ and ant geodesics $p^{bc},p^{ac}$. Eieyout loss of fenerajiey assume that we have $|z,p^{ac}_{i}|_S\leq D$. Note thzt if $i>|a,c|_S$, then $p^{ac}_i$ liws in (or is equal to) $f$, and confequently $|z,c|_S\leq |z,p^{ac}_{i}|_S\leq D\leq\rho(U)-1$, as desired. If $i\leq|e,c|_R$, tkck $|p^{az}_u,a|_D=i\geq 2D$, and hence $$|z, | '$ in $\Gamma$. Since the vertex $v$ $(\epsilon, by Theorem applied to $u,u', a $v_a$ (respectively $v_b$) a geodesic in from $a$ (respectively $b$) to $u$ $u'$ satisfying $|v_a,v|_S\leq D$ (respectively $|v_b,v|_S\leq D$). Consequently $|a,b|_S\leq |a,v_a|_S+|v_a,v|_S+|v,v_b|_S+|v_b,b|_S\leq 2\rho(U)+2D\leq 4D$, as Henceforth, we assume $\rho(U)\geq D+1$. Let $\ell=|a,b|_S$. If $\ell> 4D$, then choose any i\leq and geodesic from $a$ to $b$ in $\Gamma$. If $p^{ab}_i$ is an $(\epsilon, R)$-deep vertex of $p^{ab}$ in peripheral left coset $w$, then let $z=w$, otherwise $z=p^{ab}_i$. We claim that any $c\in U$ we have \rho(U)-1$, the definition $\rho(U)$, implying 4D$. Indeed, we Theorem \[thm:stronglythin\] to $a,b,c,$ and any geodesics $p^{bc},p^{ac}$. Without loss of generality assume that we have $|z,p^{ac}_{i}|_S\leq Note that then $p^{ac}_i$ in is to) $c$, and |z,p^{ac}_{i}|_S\leq D\leq\rho(U)-1$, as desired. If $i\leq|a,c|_S$, and hence $$|z, | '$ in $\Gamma$. Since the vertex $v$ is nOt $(\epsilon, R)$-Deep, bY ThEorEm \[Thm:sTronGlythin\] applied TO $u,u', a$ (Respectively $u,u',b$), we obtaiN a verTeX $V_a$ (reSPeCtiveLy $v_b$) on a GEoDESic In $\gaMma$ FrOM $a$ (RespeCtiVely $b$) to $U$ or $u'$ satisfYinG $|v_A,v|_S\leq D$ (respeCTiVely $|v_b,v|_S\leQ D$). COnsequently $|a,B|_S\lEq
|a,v_a|_S+|V_a,V|_S+|v,V_B|_S+|v_b,b|_s\leQ 2\rho(U)+2d\leq
4D$, aS DesireD.
HenceforTh, WE assumE $\Rho(U)\geq d+1$. lEt $\Ell=|a,B|_S$. If $\ell> 4D$, then choosE AnY $2d\leq i\leq \ell-2D$, anD any geOdESiC $P^{Ab}$ fRom $A$ to $b$ in $\GammA$. IF $p^{ab}_i$ IS an $(\epsiLOn, r)$-DEEp vERtex of $p^{ab}$ in thE peripheral LEft Coset $w$, ThEn lET $z=w$, othErwisE lET $z=p^{Ab}_i$. We claim tHat fOr any $c\in U$ We have $|Z,C|_S\leq \rhO(u)-1$, contraDictinG thE deFiniTIoN oF $\rhO(U)$, ANd iMPlYinG $\Ell\Leq 4D$.
IndeEd, We Apply theoREM \[THm:stRonGlytHin\] to $A,b,c,$ and any geodEsiCs $p^{bC},P^{ac}$. withoUt losS of gEnEraliTy assuMe thaT wE have $|z,p^{ac}_{i}|_S\leq D$. note That if $i>|a,c|_s$, thEn $P^{ac}_I$ lIes in (OR is equAl tO) $c$, aNd conseQuently $|Z,C|_S\lEq |Z,P^{AC}_{i}|_s\leq D\leq\rho(U)-1$, as desiReD. iF $i\Leq|a,c|_S$, thEn $|p^{ac}_i,A|_s=i\GeQ 2d$, and hencE $$|z, | '$ in $\Gamma$. Since thevertex $v$ is n ot$(\ ep silo n, R )$-deep, by Th e orem \[thm:stronglythin\]appli ed to $ u ,u ', a$ (respe c ti v e ly$u ,u ',b $) , w e obt ain a vert ex $v_a$ ( res pe ctively $v_b $ )on a geode sic in $\Gamma$ fr om $a$ ( res p ectiv ely $b$) to $u $ or $u '$ satisf yi n g $|v_ a ,v|_S\l e q D $ (r espectively $|v_b , v| _ S\leq D$). Con sequen tl y $ | a ,b| _S\ leq
|a,v_a |_ S+|v_ a ,v|_S+| v ,v _ b | _S+ | v_b,b|_S\leq2\rho(U)+2D \ leq
4D$,as de s ired.
Henc ef o rth , we assume $\r ho(U)\geq D+1$. Let $\e l l=|a,b| _S$. I f $ \el l> 4 D $, t hen c h oos e a ny$ 2D\ leq i\le q\e ll-2D $, a n d a ny g eod esic $p^{ ab}$ from $a$ to $b$ in$\Gam ma$.If $ p^ {ab}_ i$ isan $( \e psilon, R)$-dee p ve rtex of $ p^{ ab }$in thep eriphe ral le ft cose t $w$,t hen l e t $z =w$, otherwise let $ z = p^ {ab}_i$. We cl a im t h at for a ny $c \inU $ we h ave$ |z ,c|_S\le q \rho ( U) -1 $, cont ra dictin gthe de finit i on o f $\rh o(U)$, a nd im p lying $\ell\le q 4D$.
Indeed , w e ap p ly T heo rem \[thm:s tron g lyth in\] to $a , b,c,$ andan y g e odesics $p^{bc},p^{ ac }$. Wi thout loss of gene rality ass u m e that we hav e $ | z,p^{ac}_{i}|_ S\leq D$. Notet hat if $ i>|a, c|_S$, t hen $p^{a c } _i$ lies in (o r i s e q u al to) $c$, and c onse qu ently $ |z, c|_S\le q | z,p ^{a c}_ {i }|_S\leqD\leq\rh o( U) -1 $, as desi r ed. If $ i\ leq |a ,c| _S$,t hen $| p^{ac }_i, a| _S = i\g eq 2D$, an d henc e$$ |z, | '$ in $\Gamma$._Since the_vertex $v$ is not_$(\epsilon, R)$-deep,_by_Theorem \[thm:stronglythin\] applied_to_$u,u', a$ (respectively_$u,u',b$), we obtain_a vertex $v_a$ (respectively_$v_b$) on a_geodesic_in $\Gamma$ from $a$ (respectively $b$) to $u$ or $u'$ satisfying $|v_a,v|_S\leq D$ (respectively_$|v_b,v|_S\leq_D$). Consequently_$|a,b|_S\leq
|a,v_a|_S+|v_a,v|_S+|v,v_b|_S+|v_b,b|_S\leq_2\rho(U)+2D\leq
4D$,_as desired.
Henceforth, we assume $\rho(U)\geq_D+1$. Let $\ell=|a,b|_S$. If $\ell>_4D$, then_choose any $2D\leq i\leq \ell-2D$, and any geodesic_$p^{ab}$_from $a$ to $b$ in $\Gamma$._If $p^{ab}_i$ is an $(\epsilon, R)$-deep vertex of $p^{ab}$_in the peripheral left coset $w$, then_let $z=w$, otherwise_let_$z=p^{ab}_i$._We claim that for_any $c\in U$ we have $|z,c|_S\leq_\rho(U)-1$, contradicting the definition of $\rho(U)$,_and implying $\ell\leq 4D$.
Indeed, we apply Theorem \[thm:stronglythin\]_to $a,b,c,$ and any geodesics $p^{bc},p^{ac}$._Without loss of generality assume_that we_have $|z,p^{ac}_{i}|_S\leq D$. Note that_if $i>|a,c|_S$, then_$p^{ac}_i$ lies_in (or is_equal to) $c$, and consequently $|z,c|_S\leq_|z,p^{ac}_{i}|_S\leq D\leq\rho(U)-1$, as_desired. If $i\leq|a,c|_S$, then $|p^{ac}_i,a|_S=i\geq 2D$,_and_hence $$|z, |
gauge connection equal to the spin connection*]{}, gives the projector \[proj1\] | | \_A = - p\^P\^x\_(\^x)\_A\^[ B]{} \^[01]{} \_B. This constraint also holds in the case of flat horizon if we set $\kappa=0$. The $\tha$-independent parts of and are equal and give a second projector \[proj2\] S\_[AB]{}\^B = (V’-U’)e\^U \^1\_A - e\^[2 (U-V)]{} \^[01]{} \_[AB]{} \^B. Subtracting the $\tha$ independent parts of and gives a third projector \[proj3\] \_A= - e\^[U-2V]{} \_[AB]{}\^[0]{} \^B + e\^[-2U]{} \^P\^x\_ \^[01]{}(\^x)\_A\^[ B]{} \_B.
Finally, subtracting and we obtain an equation for the radial dependence of the spinor \_1 \_A=\_A + e\^[-U]{}\^P\^x\_ \^[01]{}(\^x)\_A\^[ B]{} \_B. \[raddep\] In total we get three projectors, -, one differential relation on the spinor and one algebraic constraint. The idea is to further simply these equations so as to ensure that we end up with two projectors. From now on we will specify to the case of spherical or hyperbolic symmetry, since this is what we will use in the paper. In order to reduce the number of projectors we impose the constraint \^P\_\^x = c e\^[2U]{} p\^P\_\^x, x=1,2,3 \[pcq\] for some real function $c$. By squaring we obtain the algebraic condition (p\^P\_\^x)\^2= 1 which can be used to rewrite as \[ceq\] c= e\^[-2U]{}\^P\_\^x p\^P\_\^x.
Substituting in and using, we obtain the projector \[projNa\] \_A= - \_[A B]{} \^0 \^B which, squared, gives the norm of $\cZ$ ||\^2= e\^[4V - 2U]{}\[(2U’-V’)\^2 +c\^2\]. \[normZ\]
Then we can rewrite as \_A=ie\^[i ]{} \_[AB]{}\^[0]{} \^B \[proj4\] | gauge connection equal to the spin association * ] { }, render the projector \[proj1\ ] | | \_A = - p\^P\^x\_(\^x)\_A\^ [ B ] { } \^[01 ] { } \_B. This constraint also hold in the case of flat horizon if we set up $ \kappa=0$. The $ \tha$-independent contribution of and are equal and move over a second projector \[proj2\ ] S\_[AB]{}\^B = (V’-U’)e\^U \^1\_A - e\^[2 (U - V) ] { } \^[01 ] { } \_[AB ] { } \^B. subtract the $ \tha$ independent parts of and give a third projector \[proj3\ ] \_A= - e\^[U-2V ] { } \_[AB]{}\^[0 ] { } \^B + e\^[-2U ] { } \^P\^x\ _ \^[01]{}(\^x)\_A\^ [ B ] { } \_B.
Finally, subtract and we obtain an equation for the radial addiction of the spinor \_1 \_A=\_A + e\^[-U]{}\^P\^x\ _ \^[01]{}(\^x)\_A\^ [ B ] { } \_B. \[raddep\ ] In total we get three projector, -, one differential relation on the spinor and one algebraic constraint. The mind is to further merely these equations so as to ensure that we end up with two projectors. From now on we will intend to the case of spherical or hyperbolic symmetry, since this is what we will use in the paper. In order to reduce the number of projectors we impose the constraint \^P\_\^x = c e\^[2U ] { } p\^P\_\^x, x=1,2,3 \[pcq\ ] for some real function $ c$. By squaring we obtain the algebraic condition (p\^P\_\^x)\^2= 1 which can be used to rewrite as \[ceq\ ] c= e\^[-2U]{}\^P\_\^x p\^P\_\^x.
substitute in and using, we obtain the projector \[projNa\ ] \_A= - \_[A boron ] { } \^0 \^B which, feather, gives the norm of $ \cZ$ ||\^2= e\^[4V - 2U]{}\[(2U’-V’)\^2 + c\^2\ ]. \[normZ\ ]
Then we can rewrite as \_A = ie\^[i ] { } \_[AB]{}\^[0 ] { } \^B \[proj4\ ] | gakge connection equal to uhe spin connection*]{}, givev the lrojectof \[proj1\] | | \_A = - p\^P\^x\_(\^x)\_A\^[ B]{} \^[01]{} \_B. This ronsrrainu also holds in the case of vlat horuzon uf we set $\kappa=0$. Thc $\tha$-jkdepeudxnt parts of anc are equan and give a sacunb projector \[proj2\] S\_[AB]{}\^B = (V’-U’)e\^U \^1\_A - e\^[2 (U-V)]{} \^[01]{} \_[AB]{} \^B. Sibhracting the $\tra$ imqepehdent parts of and gives a third pdojectog \[proj3\] \_A= - e\^[U-2V]{} \_[AB]{}\^[0]{} \^N + e\^[-2U]{} \^P\^x\_ \^[01]{}(\^x)\_A\^[ B]{} \_B.
Finally, subtgactlng and we obtain wn equation for rhe radial ddpendence of the spinot \_1 \_A=\_A + e\^[-U]{}\^P\^x\_ \^[01]{}(\^x)\_A\^[ B]{} \_B. \[raddep\] In totau we yet three ptonefjors, -, one dihferennial relation on the vpinor snd one algebrsic cobstraint. The idea is vo further simply thgse equatimna so as to ensure tyat wg end up duth twk 'ronectord. Fcom now on se will spexify to the case of s[yerical or hylerbolyc symmetry, since this is what we will ust in fhe paper. In order to rwduce the number of ptojectors re impose the constraint \^P\_\^x = c e\^[2U]{} p\^P\_\^x, x=1,2,3 \[pcq\] for smme rxau fbkgtiov $c$. By squaring we obtain the algebraic conditiog (p\^K\_\^x)\^2= 1 which can be msed to rewrite as \[cfq\] s= e\^[-2U]{}\^P\_\^x p\^P\_\^x.
Subrtitutnhg in and using, we ohtain tre prijector \[pwojNs\] \_A= - \_[A B]{} \^0 \^B which, squared, guves the nori of $\cZ$ ||\^2= e\^[4V - 2U]{}\[(2U’-V’)\^2 +c\^2\]. \[uormZ\]
Then we can rewrote as \_A=ie\^[i ]{} \_[AB]{}\^[0]{} \^B \[proj4\] | gauge connection equal to the spin connection*]{}, projector | | = - p\^P\^x\_(\^x)\_A\^[ also in the case flat horizon if set $\kappa=0$. The $\tha$-independent parts of are equal and give a second projector \[proj2\] S\_[AB]{}\^B = (V’-U’)e\^U \^1\_A - (U-V)]{} \^[01]{} \_[AB]{} \^B. Subtracting the $\tha$ independent parts of and gives a projector \_A= e\^[U-2V]{} \^B + e\^[-2U]{} \^P\^x\_ \^[01]{}(\^x)\_A\^[ B]{} \_B. Finally, subtracting and we obtain an equation for the dependence of the spinor \_1 \_A=\_A + e\^[-U]{}\^P\^x\_ B]{} \_B. \[raddep\] In we get three projectors, -, differential on the and algebraic The idea is further simply these equations so as to ensure that we end up with two projectors. From now we will the case spherical hyperbolic since this is will use in the paper. In the number of projectors we impose the constraint = c p\^P\_\^x, x=1,2,3 \[pcq\] for some real $c$. By squaring we obtain the algebraic condition 1 which can be used to rewrite as \[ceq\] c= e\^[-2U]{}\^P\_\^x p\^P\_\^x. Substituting in and obtain the projector \[projNa\] - \_[A B]{} \^B squared, the of $\cZ$ e\^[4V - 2U]{}\[(2U’-V’)\^2 +c\^2\]. \[normZ\] Then we can rewrite as \_A=ie\^[i \_[AB]{}\^[0]{} \^B \[proj4\] | gauge connection equal to the Spin connecTion*]{}, gIveS thE pRojeCtor \[Proj1\] | | \_A = - p\^P\^x\_(\^x)\_A\^[ B]{} \^[01]{} \_B. ThIS conStraint also holds in the cAse of FlAT horIZoN if we Set $\kappA=0$. thE $\THa$-iNdEpEndEnT PaRts of And Are equaL and give a sEcoNd Projector \[proJ2\] s\_[Ab]{}\^B = (V’-U’)e\^U \^1\_A - e\^[2 (U-V)]{} \^[01]{} \_[aB]{} \^B. subtracting tHe $\tHa$ indePeNdeNT partS of And giVes a thIRd projEctor \[proj3\] \_a= - e\^[u-2v]{} \_[AB]{}\^[0]{} \^B + e\^[-2U]{} \^p\^X\_ \^[01]{}(\^x)\_A\^[ B]{} \_B.
FiNALlY, subTracting and we obtaIN aN Equation for the Radial DePEnDENce Of tHe spinor \_1 \_A=\_A + E\^[-U]{}\^p\^x\_ \^[01]{}(\^x)\_A\^[ B]{} \_b. \[Raddep\] IN ToTAL We gET three projectOrs, -, one diffeREntIal relAtIon ON the spInor aNd ONe aLgebraic conStraInt. The ideA is to fURther siMPly thesE equatIonS so As to ENsUrE thAt WE enD Up WitH Two ProjectoRs. frOm now On we WILL SpecIfy To thE case Of spherical or HypErboLIc sYmmetRy, sinCe thIs Is whaT we wilL use iN tHe paper. In order tO redUce the numBer Of ProJeCtors WE imposE thE coNstrainT \^P\_\^x = c e\^[2U]{} p\^p\_\^X, x=1,2,3 \[pCq\] FOR SoMe real function $c$. By sQuARInG we obtaiN the alGEbRaIC conditiOn (P\^P\_\^x)\^2= 1 WhicH CAn be uSed tO ReWrite as \[cEq\] c= e\^[-2U]{}\^P\_\^X P\^P\_\^X.
SUbstituTiNg in anD uSinG, we ObtaiN The pRojectOr \[projNa\] \_a= - \_[A B]{} \^0 \^B wHIch, squared, giveS The norm of $\cZ$ ||\^2= e\^[4V - 2u]{}\[(2u’-V’)\^2 +C\^2\]. \[NOrMz\]
TheN we Can rewrite aS \_A=ie\^[I ]{} \_[aB]{}\^[0]{} \^B \[pRoj4\] | gauge connection equal to the spinconne cti on* ]{ }, g ives the projector \[pr oj1\] | | \_A = - p\^P \^x\_ (\ ^ x)\_ A \^ [ B]{ } \^[01 ] {} \ _B. T hi s c on s tr aintals o holds in the ca seof flat horizo n i f we set $ \ka ppa=0$. The$\t ha$-in de pen d ent p art s ofand ar e equal and give a second project o r \ [pro j2\] S\_[AB]{}\^B =( V’-U’)e\^U \^1 \_A -e\ ^ [2 ( U-V )]{ } \^[01]{} \ _[AB] { } \^B.S ub t r a cti n g the $\tha$independent par ts ofan d g i ves athird p r oje ctor \[proj 3\]\_A= - e\ ^[U-2V ] {} \_[A B ]{}\^[0 ]{} \^ B + e\ ^[-2 U ]{ }\^P \^ x \_\ ^[ 01] { }(\ ^x)\_A\^ [B] {} \_ B.
F i n a lly, su btra cting and we obtai n a n eq u ati on fo r the rad ia l dep endenc e ofth e spinor \_1 \_ A=\_ A + e\^[- U]{ }\ ^P\ ^x \_ \^ [ 01]{}( \^x )\_ A\^[ B] {} \_B. \[r ad d e p \] In total we get t hr e e p rojector s, -,o ne d i fferenti al re lati o n on t he s p in or and o ne alg e br ai c const ra int. T he id eais to furt her si mply the se eq u ations so as t o ensure thatw ee n du p wi thtwo project ors. From now on we willspeci fy to the case of spheric al or hy perbo lic symmetry, since thi s i s what w e wi l lu se in the pape r. In order tor educe th e num ber of p rojectors w e impose th e c ons tra i n t\^P\_\^x = ce \ ^[2U ]{ } p\^P\ _\^ x, x= 1,2 ,3\[p cq\ ]for somereal fun ct io n$c $.By sq u aring we o bta in th e alg e braiccondi tion ( p\ ^ P\_ \^x)\^2 = 1 w hich c an beuse dto re writ e as \[ceq\ ] c= e\^[ -2U ] {}\^ P\ _\ ^x p\^P \_\^x.
Subst it uting in a nd us ing, w e obtain t he projector \[projNa\] \_A= -\_[ A B]{ } \^ 0 \^B whi ch, squar ed, givesthe no rm of $ \cZ $ ||\^2 = e\ ^[4 V- 2U]{}\[( 2 U ’-V ’)\^2 + c\^2 \]. \[n ormZ\]
Then we ca n re write as \_A= ie\ ^[i] { }\_[ A B] { }\^ [0 ] {}\ ^ B \[proj4\] | gauge_connection equal_to the spin connection*]{},_gives the_projector_\[proj1\] |_|_\_A = -_p\^P\^x\_(\^x)\_A\^[ B]{} \^[01]{} \_B._This constraint also holds_in the case_of_flat horizon if we set $\kappa=0$. The $\tha$-independent parts of and are equal and_give_a second_projector_\[proj2\]_S\_[AB]{}\^B = (V’-U’)e\^U \^1\_A -_e\^[2 (U-V)]{} \^[01]{} \_[AB]{} \^B._Subtracting the_$\tha$ independent parts of and gives a third_projector_\[proj3\] \_A= -_e\^[U-2V]{} \_[AB]{}\^[0]{} \^B + e\^[-2U]{} \^P\^x\_ \^[01]{}(\^x)\_A\^[ B]{} \_B.
Finally, subtracting_and we obtain an equation for_the radial dependence_of_the_spinor \_1 \_A=\_A +_e\^[-U]{}\^P\^x\_ \^[01]{}(\^x)\_A\^[ B]{} \_B. \[raddep\] In total_we get three projectors, -, one_differential relation on the spinor and one_algebraic constraint. The idea is to_further simply these equations so_as to_ensure that we end up_with two projectors._From now_on we will_specify to the case of spherical_or hyperbolic symmetry,_since this is what we will_use_in the paper._In_order_to reduce_the number of_projectors_we impose_the_constraint \^P\_\^x = c e\^[2U]{} p\^P\_\^x,_ x=1,2,3_\[pcq\] for some real function $c$. By_squaring we obtain the_algebraic_condition (p\^P\_\^x)\^2= 1 which_can be used to rewrite_as \[ceq\] c= e\^[-2U]{}\^P\_\^x p\^P\_\^x.
Substituting in_and using,_we obtain_the projector \[projNa\] \_A= - \_[A B]{} \^0 \^B which, squared,_gives the norm of $\cZ$ ||\^2=_e\^[4V - 2U]{}\[(2U’-V’)\^2 +c\^2\]._\[normZ\]
Then we_can_rewrite as \_A=ie\^[i_]{}_\_[AB]{}\^[0]{} \^B_\[proj4\] |
1)}$ est inclus, à conjugaison près, dans l’un des groupes suivants : $$\{(x+\alpha,y+\beta)\hs|\hs\alpha, \hs\beta \in\C\},
\hspace{6mm} \{(\alpha x,y+\beta)\hs| \hs\alpha\in\C^*, \hs
\beta\in \C\},\hspace{6mm}\{(\alpha x,\beta y)\hs|\hs\alpha,
\hs\beta \in\C^*\}~;$$ le [*1.*]{} permet donc de conclure.
[*3.*]{} Pour finir nous nous intéressons au cas $S=\mathsf{F}_n$, $n\geq 2$. Notons $\pi$ la projection de $\Aut(\mathsf{F}_n)$ sur $\PGL_2(\C)$ ; puisque le groupe $\pi(
\mathsf{N}^{(k-1)})$ est abélien, nous pouvons supposer, à indice fini près, que $\mathsf{N}^{(k-1)}$ est contenu dans l’un des groupes suivants : $$\begin{aligned}
& {\it a. }\hs \{(\alpha x+P(y),y)\hs|\hs\alpha\in\C^*,\hs P\in
\C[y]\}; && \\
& {\it b. }\hs \{(\alpha x+P(y),y+\beta)\hs|\hs\alpha
\in\C^*,\hs\beta\in\C,\hs P\in\C[y]\}; && \\
& {\it c. }\hs \{(\alpha x+P(y),\beta y)\hs|\hs\alpha,\hs\beta
\in\C^*,\hs P\in\C[y]\}; &&\end{aligned}$$ Considérons chacune de ces éventualités.
[*3.a.*]{} Supposons que $\mathsf{N}^{(k-1)}$ soit un sous-groupe de $\{(x+P(y),y)\hs|\hs
P\in\C[y]\}$. Soit $g$ dans $\mathsf{N}^{(k-1)}\setminus
\{\mathsf{id}\}$ ; à conjugaison près, $g$ s’écrit | 1)}$ est inclus, à conjugaison près, dans l’un des groupes suivants : $ $ \{(x+\alpha, y+\beta)\hs|\hs\alpha, \hs\beta \in\C\ },
\hspace{6 mm } \{(\alpha x, y+\beta)\hs| \hs\alpha\in\C^ *, \hs
\beta\in \C\},\hspace{6mm}\{(\alpha x,\beta y)\hs|\hs\alpha,
\hs\beta \in\C^*\}~;$$ le [ * 1. * ] { } permet donc de conclure.
[ * 3. * ] { } Pour finir nous nous intéressons au cas $ S=\mathsf{F}_n$, $ n\geq 2$. Notons $ \pi$ la projection de $ \Aut(\mathsf{F}_n)$ sur $ \PGL_2(\C)$ ; puisque le groupe $ \pi (
\mathsf{N}^{(k-1)})$ est abélien, nous pouvons supposer, à indice fini près, que $ \mathsf{N}^{(k-1)}$ est contenu dans l’un des groupes suivants : $ $ \begin{aligned }
& { \it a. } \hs \{(\alpha x+P(y),y)\hs|\hs\alpha\in\C^*,\hs P\in
\C[y]\ }; & & \\
& { \it b. } \hs \{(\alpha x+P(y),y+\beta)\hs|\hs\alpha
\in\C^*,\hs\beta\in\C,\hs P\in\C[y]\ }; & & \\
& { \it c. } \hs \{(\alpha x+P(y),\beta y)\hs|\hs\alpha,\hs\beta
\in\C^*,\hs P\in\C[y]\ }; & & \end{aligned}$$ Considérons chacune de ces éventualités.
[ * 3.a. * ] { } Supposons que $ \mathsf{N}^{(k-1)}$ soit un sous - groupe de $ \{(x+P(y),y)\hs|\hs
P\in\C[y]\}$. Soit $ g$ dans $ \mathsf{N}^{(k-1)}\setminus
\{\mathsf{id}\}$ ; à conjugaison près, $ g$ s’écrit | 1)}$ esh inclus, à conjugaison prèr, dans l’un des yeoupes suivahts : $$\{(x+\alpfa,y+\beta)\hs|\hs\alpha, \hs\beta \in\C\},
\hdpqce{6mm} \{(\alpha x,y+\beta)\hs| \hs\alphx\in\C^*, \hs
\bena\in \C\},\hspqce{6mn}\{(\qlpha x,\bete y)\hs|\hs\alpha,
\hs\gcta \iu\C^*\}~;$$ le [*1.*]{} permet dokc de conclgre.
[*3.*]{} Pour finir nuud nous intéressons au cas $S=\mathsf{F}_n$, $g\geq 2$. Nptlns $\pi$ la projgctiom de $\Zlt(\nathsf{F}_n)$ sur $\PGL_2(\C)$ ; puisque le froupe $\ki(
\mathsf{N}^{(k-1)})$ est abéloen, nous pouvons supposer, à indlce fini près, que $\mwthsf{N}^{(k-1)}$ est conewnu dans l’un des groupts suivants : $$\gegin{aligned}
& {\it a. }\hs \{(\alpha x+P(y),y)\fs|\hs\akpha\in\C^*,\hs K\nb
\C[y]\}; && \\
& {\it b. }\hs \{(\elpha q+P(y),y+\beta)\hs|\hs\alpha
\in\C^*,\hv\beta\in\V,\hs P\in\C[y]\}; && \\
& {\ib c. }\hv \{(\aopha x+P(y),\beta y)\hs|\hs\alpia,\hs\beta
\in\C^*,\hs P\in\C[y]\}; &&\gnd{aligned}$$ Cknsidérons chacune dw ces évendualkrés.
[*3.a.*]{} Sukpoxohs que $\mavhsf{N}^{(k-1)}$ soit un sous-groype de $\{(x+P(y),y)\hs|\hs
P\in\C[u]\}$. Fiit $g$ dans $\mafhsf{N}^{(k-1)}\feeminus
\{\mathsf{id}\}$ ; à conjugaison près, $g$ s’écrit | 1)}$ est inclus, à conjugaison près, dans groupes : $$\{(x+\alpha,y+\beta)\hs|\hs\alpha, \in\C\}, \hspace{6mm} \{(\alpha x,\beta \hs\beta \in\C^*\}~;$$ le permet donc de [*3.*]{} Pour finir nous nous intéressons cas $S=\mathsf{F}_n$, $n\geq 2$. Notons $\pi$ la projection de $\Aut(\mathsf{F}_n)$ sur $\PGL_2(\C)$ ; le groupe $\pi( \mathsf{N}^{(k-1)})$ est abélien, nous pouvons supposer, à indice fini près, $\mathsf{N}^{(k-1)}$ contenu l’un groupes suivants : $$\begin{aligned} & {\it a. }\hs \{(\alpha x+P(y),y)\hs|\hs\alpha\in\C^*,\hs P\in \C[y]\}; && \\ & {\it }\hs \{(\alpha x+P(y),y+\beta)\hs|\hs\alpha \in\C^*,\hs\beta\in\C,\hs P\in\C[y]\}; && \\ & c. }\hs \{(\alpha x+P(y),\beta \in\C^*,\hs P\in\C[y]\}; &&\end{aligned}$$ Considérons chacune ces [*3.a.*]{} Supposons $\mathsf{N}^{(k-1)}$ un de $\{(x+P(y),y)\hs|\hs P\in\C[y]\}$. $g$ dans $\mathsf{N}^{(k-1)}\setminus \{\mathsf{id}\}$ ; à conjugaison près, $g$ s’écrit | 1)}$ est inclus, à conjugaison près, Dans l’un des GroupEs sUivAnTs : $$\{(x+\aLpha,Y+\beta)\hs|\hs\alpha, \HS\betA \in\C\},
\hspace{6mm} \{(\alpha x,y+\betA)\hs| \hs\AlPHa\in\c^*, \Hs
\Beta\iN \C\},\hspacE{6Mm}\{(\ALPha X,\bEtA y)\hS|\hS\AlPha,
\hs\BetA \in\C^*\}~;$$ le [*1.*]{} pErmet donc dE coNcLure.
[*3.*]{} Pour finiR NoUs nous intéResSons au cas $S=\maThsF{F}_n$, $n\geQ 2$. NOtoNS $\pi$ la ProJectiOn de $\AuT(\Mathsf{f}_n)$ sur $\PGL_2(\C)$ ; PuISque le GRoupe $\pi(
\MAThSf{N}^{(k-1)})$ Est abélien, nous pouVOnS Supposer, à indicE fini pRèS, QuE $\MAthSf{N}^{(K-1)}$ est contenU dAns l’uN Des grouPEs SUIVanTS : $$\begin{aligned}
& {\It a. }\hs \{(\alpha x+p(Y),y)\hS|\hs\alpHa\In\C^*,\HS P\in
\C[y]\}; && \\
& {\It b. }\hs \{(\AlPHa x+p(y),y+\beta)\hs|\hs\AlphA
\in\C^*,\hs\betA\in\C,\hs p\In\C[y]\}; && \\
& {\it c. }\HS \{(\alpha x+p(y),\beta Y)\hs|\Hs\aLpha,\HS\bEtA
\in\c^*,\hS p\in\c[Y]\}; &&\eNd{aLIgnEd}$$ ConsidÉrOnS chacUne dE CES ÉvenTuaLitéS.
[*3.a.*]{} SupPosons que $\mathSf{N}^{(K-1)}$ soiT Un sOus-grOupe dE $\{(x+P(y),Y)\hS|\hs
P\iN\C[y]\}$. SoiT $g$ danS $\mAthsf{N}^{(k-1)}\setminus
\{\MathSf{id}\}$ ; à conjUgaIsOn pRèS, $g$ s’écRIt | 1)}$ est inclus, à conjuga ison près, dans l’ unde s gr oupe s suivants : $ $ \{(x +\alpha,y+\beta)\hs|\h s\alp ha , \hs \ be ta \i n\C\},\ hs p a ce{ 6m m} \{ (\ a lp ha x, y+\ beta)\h s| \hs\alp ha\ in \C^*, \hs
\b e ta \in \C\},\ hsp ace{6mm}\{(\ alp ha x,\ be tay )\hs| \hs \alph a,
\hs \ beta \ in\C^*\}~ ;$ $ le [* 1 .*]{} p e r me t do nc de conclure.
[ *3 . *]{} Pour fini r nous n o us i nté res sons au ca s$S=\m a thsf{F} _ n$ , $ n\g e q 2$. Notons$\pi$ la pr o jec tion d e$\A u t(\mat hsf{F }_ n )$sur $\PGL_2 (\C) $ ; puisq ue leg roupe $ \ pi(
\ma thsf{N }^{ (k- 1)}) $ e st ab él i en, no usp ouv ons supp os er , à i ndic e f i ni p rès , qu e $\m athsf{N}^{(k- 1)} $ es t co ntenu dans l’u ndes g roupes suiv an ts : $$\begin{a lign ed}
& {\i t a .}\h s\{(\a l pha x+ P(y ),y )\hs|\h s\alpha \ in\ C^ * , \ hs P\in
\C[y]\}; && \ \ &{\it b.}\hs \ { (\ al p ha x+P(y ), y+\ beta ) \ hs|\h s\al p ha
\in\C^* ,\hs\b e ta \i n\C,\hs P \in\C[ y] \}; && \\& {\i t c. } \hs \{(\ alpha x+P(y),\beta y ) \hs|\hs\alpha , \h s \ be t a
\i n\C ^*,\hs P\in \C[y ] \};&&\e n d{ ali g ned}$ $ Con si d ér o ns chacune de ces é ve ntuali tés.
[*3.a.*]{} S upposons q u e $\mathsf {N}^ { (k - 1)}$ soit un s ous-g roupe de $ \ {(x+P(y) ,y)\h s|\hs
P\ in\C[y]\} $ . Soit $g $ d ans $\ mat h s f{ N}^{(k-1)}\se t m inus
\ {\maths f{i d}\}$ ; àcon jug ais on près, $g $ s’écri t | 1)}$ est_inclus, à_conjugaison près, dans l’un_des groupes_suivants :_$$\{(x+\alpha,y+\beta)\hs|\hs\alpha, \hs\beta_\in\C\},
\hspace{6mm}_\{(\alpha x,y+\beta)\hs| \hs\alpha\in\C^*,_\hs
\beta\in \C\},\hspace{6mm}\{(\alpha x,\beta_y)\hs|\hs\alpha,
\hs\beta \in\C^*\}~;$$ le [*1.*]{}_permet donc de_conclure.
[*3.*]{}_Pour finir nous nous intéressons au cas $S=\mathsf{F}_n$, $n\geq 2$. Notons $\pi$ la projection_de_$\Aut(\mathsf{F}_n)$ sur_$\PGL_2(\C)$ ;_puisque_le groupe $\pi(
\mathsf{N}^{(k-1)})$ est abélien,_nous pouvons supposer, à indice_fini près,_que $\mathsf{N}^{(k-1)}$ est contenu dans l’un des groupes_suivants :_$$\begin{aligned}
& {\it a._}\hs \{(\alpha x+P(y),y)\hs|\hs\alpha\in\C^*,\hs P\in
\C[y]\}; && \\
& {\it b._}\hs \{(\alpha x+P(y),y+\beta)\hs|\hs\alpha
\in\C^*,\hs\beta\in\C,\hs P\in\C[y]\}; && _\\
& {\it c._}\hs_\{(\alpha_x+P(y),\beta y)\hs|\hs\alpha,\hs\beta
\in\C^*,\hs P\in\C[y]\}; &&\end{aligned}$$_Considérons chacune de ces éventualités.
[*3.a.*]{} Supposons_que $\mathsf{N}^{(k-1)}$ soit un sous-groupe de_$\{(x+P(y),y)\hs|\hs
P\in\C[y]\}$. Soit $g$ dans $\mathsf{N}^{(k-1)}\setminus
\{\mathsf{id}\}$ ; à conjugaison_près, $g$ s’écrit |
(1.1) & 2.4(0.3)&\
G37.48$-$0.10 & [H$^{13}$CO$^{+}$]{}(3-2) & 30 May 08 & 0.07 & 0.41(0.05) & 59.2(0.2) & 2.6(0.4) & 1.1(0.2) & 3.83(+0.42,-0.41)\
& [CH$_3$OH]{}(5$_{2,3}$-4$_{1,3}$) & 30 May 08 & 0.07 & $<$0.21 & && &\
G39.10+0.49 & [H$^{13}$CO$^{+}$]{}(3-2) & 29 May 08 & 0.05 & 0.31(0.03) & 23.0(0.2) & 4.2(0.4) & 1.4(0.2) & 1.73(+0.43,-0.44)\
& [CH$_3$OH]{}(5$_{2,3}$-4$_{1,3}$) & 29 May 08 & 0.05 & 0.24(0.02) & 23.3(0.3) & 8.8(0.7) & 2.3(0.3) &\
& SiO(5-4) & 13 June 08 & 0.04 & $<$0.12 & & & &\
G49.27$-$0.34 & [H$^{13}$CO$^{+}$]{}(3-2) & 30 May 08 & 0.07 & 1.01(0.04) & 67.9(0.1) & 3.3(0.2) & 3.6(0.2) & 5.55(1.66,-1.66)\
& [CH$_3$OH]{}(5$_{2,3}$-4$_{1,3}$) & 30 May 08 & 0.07 & 0.22(0.06) & 67.4(0.2) & 1.8(0.6) & 0.4(0.2) &\
& SiO(5-4) & 30 May 08 & 0.04 & 0.15(0.02) & 65.8(0.5) & 6.3(1.2) & 1.0(0.3)&\
G49.42+0 | ( 1.1) & 2.4(0.3)&\
G37.48$-$0.10 & [ H$^{13}$CO$^{+}$]{}(3 - 2) & 30 May 08 & 0.07 & 0.41(0.05) & 59.2(0.2) & 2.6(0.4) & 1.1(0.2) & 3.83(+0.42,-0.41)\
& [ CH$_3$OH]{}(5$_{2,3}$-4$_{1,3}$) & 30 May 08 & 0.07 & $ < $ 0.21 & & & & \
G39.10 + 0.49 & [ H$^{13}$CO$^{+}$]{}(3 - 2) & 29 May 08 & 0.05 & 0.31(0.03) & 23.0(0.2) & 4.2(0.4) & 1.4(0.2) & 1.73(+0.43,-0.44)\
& [ CH$_3$OH]{}(5$_{2,3}$-4$_{1,3}$) & 29 May 08 & 0.05 & 0.24(0.02) & 23.3(0.3) & 8.8(0.7) & 2.3(0.3) & \
& SiO(5 - 4) & 13 June 08 & 0.04 & $ < $ 0.12 & & & & \
G49.27$-$0.34 & [ H$^{13}$CO$^{+}$]{}(3 - 2) & 30 May 08 & 0.07 & 1.01(0.04) & 67.9(0.1) & 3.3(0.2) & 3.6(0.2) & 5.55(1.66,-1.66)\
& [ CH$_3$OH]{}(5$_{2,3}$-4$_{1,3}$) & 30 May 08 & 0.07 & 0.22(0.06) & 67.4(0.2) & 1.8(0.6) & 0.4(0.2) & \
& SiO(5 - 4) & 30 May 08 & 0.04 & 0.15(0.02) & 65.8(0.5) & 6.3(1.2) & 1.0(0.3)&\
G49.42 + 0 | (1.1) & 2.4(0.3)&\
G37.48$-$0.10 & [H$^{13}$CO$^{+}$]{}(3-2) & 30 May 08 & 0.07 & 0.41(0.05) & 59.2(0.2) & 2.6(0.4) & 1.1(0.2) & 3.83(+0.42,-0.41)\
& [CH$_3$OH]{}(5$_{2,3}$-4$_{1,3}$) & 30 May 08 & 0.07 & $<$0.21 & && &\
G39.10+0.49 & [I$^{13}$CO$^{+}$]{}(3-2) & 29 Mzy 08 & 0.05 & 0.31(0.03) & 23.0(0.2) & 4.2(0.4) & 1.4(0.2) & 1.73(+0.43,-0.44)\
& [CH$_3$OH]{}(5$_{2,3}$-4$_{1,3}$) & 29 May 08 & 0.05 & 0.24(0.02) & 23.3(0.3) & 8.8(0.7) & 2.3(0.3) &\
& SiO(5-4) & 13 June 08 & 0.04 & $<$0.12 & & & &\
G49.27$-$0.34 & [H$^{13}$CO$^{+}$]{}(3-2) & 30 May 08 & 0.07 & 1.01(0.04) & 67.9(0.1) & 3.3(0.2) & 3.6(0.2) & 5.55(1.66,-1.66)\
& [CH$_3$IY]{}(5$_{2,3}$-4$_{1,3}$) & 30 May 08 & 0.07 & 0.22(0.06) & 67.4(0.2) & 1.8(0.6) & 0.4(0.2) &\
& SiO(5-4) & 30 May 08 & 0.04 & 0.15(0.02) & 65.8(0.5) & 6.3(1.2) & 1.0(0.3)&\
G49.42+0 | (1.1) & 2.4(0.3)&\ G37.48$-$0.10 & [H$^{13}$CO$^{+}$]{}(3-2) & 08 0.07 & & 59.2(0.2) & & & 30 May & 0.07 & & && &\ G39.10+0.49 & [H$^{13}$CO$^{+}$]{}(3-2) 29 May 08 & 0.05 & 0.31(0.03) & 23.0(0.2) & 4.2(0.4) & 1.4(0.2) 1.73(+0.43,-0.44)\ & [CH$_3$OH]{}(5$_{2,3}$-4$_{1,3}$) & 29 May 08 & 0.05 & 0.24(0.02) & 23.3(0.3) 8.8(0.7) 2.3(0.3) & & 13 June 08 & 0.04 & $<$0.12 & & & &\ G49.27$-$0.34 & [H$^{13}$CO$^{+}$]{}(3-2) & May 08 & 0.07 & 1.01(0.04) & 67.9(0.1) 3.3(0.2) & 3.6(0.2) & & [CH$_3$OH]{}(5$_{2,3}$-4$_{1,3}$) & 30 May & & 0.22(0.06) 67.4(0.2) 1.8(0.6) 0.4(0.2) &\ & & 30 May 08 & 0.04 & 0.15(0.02) & 65.8(0.5) & 6.3(1.2) & 1.0(0.3)&\ G49.42+0 | (1.1) & 2.4(0.3)&\
G37.48$-$0.10 & [H$^{13}$CO$^{+}$]{}(3-2) & 30 May 08 & 0.07 & 0.41(0.05) & 59.2(0.2) & 2.6(0.4) & 1.1(0.2) & 3.83(+0.42,-0.41)\
& [CH$_3$OH]{}(5$_{2,3}$-4$_{1,3}$) & 30 May 08 & 0.07 & $<$0.21 & && &\
G39.10+0.49 & [H$^{13}$CO$^{+}$]{}(3-2) & 29 May 08 & 0.05 & 0.31(0.03) & 23.0(0.2) & 4.2(0.4) & 1.4(0.2) & 1.73(+0.43,-0.44)\
& [CH$_3$OH]{}(5$_{2,3}$-4$_{1,3}$) & 29 may 08 & 0.05 & 0.24(0.02) & 23.3(0.3) & 8.8(0.7) & 2.3(0.3) &\
& SiO(5-4) & 13 June 08 & 0.04 & $<$0.12 & & & &\
g49.27$-$0.34 & [H$^{13}$CO$^{+}$]{}(3-2) & 30 MAy 08 & 0.07 & 1.01(0.04) & 67.9(0.1) & 3.3(0.2) & 3.6(0.2) & 5.55(1.66,-1.66)\
& [Ch$_3$OH]{}(5$_{2,3}$-4$_{1,3}$) & 30 maY 08 & 0.07 & 0.22(0.06) & 67.4(0.2) & 1.8(0.6) & 0.4(0.2) &\
& SiO(5-4) & 30 may 08 & 0.04 & 0.15(0.02) & 65.8(0.5) & 6.3(1.2) & 1.0(0.3)&\
G49.42+0 | (1.1) & 2.4(0.3)&\
G37.48$ -$0.10 & [ H$^{1 3}$ CO$ ^{ +}$] {}(3 -2) & 30 May 0 8 & 0 .07 & 0.41(0.05) & 59. 2(0.2 )& 2.6 ( 0. 4) &1.1(0.2 ) & 3 .83 (+ 0. 42, -0 . 41 )\
&[CH $_3$OH] {}(5$_{2,3 }$- 4$ _{1,3}$) & 3 0 M ay 08 & 0. 07& $<$0.21 &&&&\
G39 .1 0+0 . 49 &[H$ ^{13} $CO$^{ + }$]{}( 3-2) & 29 M a y 08 & 0.05 &0 . 31 (0.0 3) & 23.0(0.2) &4 .2 ( 0.4) & 1.4(0.2 ) & 1. 73 ( +0 . 4 3,- 0.4 4)\
& [CH$ _3 $OH]{ } (5$_{2, 3 }$ - 4 $ _{1 , 3}$) & 29 May 08 & 0.05& 0. 24(0.0 2) &2 3.3(0. 3) &8. 8 (0. 7) & 2.3(0. 3) & \
& SiO(5 -4) &1 3 June0 8 & 0.0 4 & $< $0. 12& && & \G49 .2 7 $-$ 0 .3 4 & [H$ ^{13}$CO $^ {+ }$]{} (3-2 ) & 30 M ay08 & 0.07 & 1.01(0.04) &67.9 ( 0.1 ) & 3 .3(0. 2) & 3 .6(0. 2) & 5 .55(1 .6 6,-1.66)\
& [CH $_3$ OH]{}(5$_ {2, 3} $-4 $_ {1,3} $ ) & 30 Ma y 0 8 & 0.0 7 & 0.2 2 (0. 06 ) & 6 7.4(0.2) & 1.8(0.6 )& 0. 4(0.2) & \
& Si O (5 -4 ) & 30 Ma y08& 0. 0 4 & 0. 15(0 . 02 ) & 65.8 (0.5)& 6 .3 (1.2) & 1 .0(0.3 )& \
G 49. 42+0 | (1.1) &_2.4(0.3)&\
G37.48$-$0.10 &_[H$^{13}$CO$^{+}$]{}(3-2) & 30 May_08 &_0.07_& 0.41(0.05)_&_59.2(0.2) & 2.6(0.4)_& 1.1(0.2) &_3.83(+0.42,-0.41)\
& [CH$_3$OH]{}(5$_{2,3}$-4$_{1,3}$) & 30_May 08 &_0.07_& $<$0.21 & && &\
G39.10+0.49 & [H$^{13}$CO$^{+}$]{}(3-2) & 29 May 08 & 0.05 &_0.31(0.03)_& 23.0(0.2)_&_4.2(0.4)_& 1.4(0.2) & 1.73(+0.43,-0.44)\
& [CH$_3$OH]{}(5$_{2,3}$-4$_{1,3}$)_& 29 May 08 &_0.05 &_0.24(0.02) & 23.3(0.3) & 8.8(0.7) & 2.3(0.3) &\
&_SiO(5-4)_& 13 June_08 & 0.04 & $<$0.12 & & & &\
G49.27$-$0.34_& [H$^{13}$CO$^{+}$]{}(3-2) & 30 May 08_& 0.07 &_1.01(0.04)_&_67.9(0.1) & 3.3(0.2) &_3.6(0.2) & 5.55(1.66,-1.66)\
& [CH$_3$OH]{}(5$_{2,3}$-4$_{1,3}$) & 30_May 08 & 0.07 & 0.22(0.06)_& 67.4(0.2) & 1.8(0.6) & 0.4(0.2) &\
&_SiO(5-4) & 30 May 08 &_0.04 & 0.15(0.02) & 65.8(0.5)_& 6.3(1.2)_& 1.0(0.3)&\
G49.42+0 |
\right) \: = \: q_2,
\label{eq:fn:qsc}$$ which precisely match the known quantum sheaf cohomology ring relations for this case [@McOrist:2007kp; @McOrist:2008ji; @Donagi:2011uz; @Donagi:2011va].
Next, we integrate out some of the fields to find a lower-energy effective Landau-Ginzburg description of the same physics. If we integrate out $F_0$, $F_w$, we get the constraints $$\begin{aligned}
a_0 \sigma \: + \: b_0 \tilde{\sigma} & = & \exp\left( - Y_0 \right),
\\
\gamma_1 \sigma \: + \: \beta_1 \tilde{\sigma} & = & \exp\left( - Y_w \right),\end{aligned}$$ which can be solved to give $$\begin{aligned}
\sigma & = & \frac{1}{\Delta_0} \left( \beta_1 \exp\left( - Y_0 \right)
\: - \: b_0 \exp\left( - Y_w \right) \right),
\label{eq:fn:opmirror1}
\\
\tilde{\sigma} & = & \frac{1}{\Delta_0} \left( a_0 \exp\left( - Y_w \right)
\: - \: \gamma_1 \exp\left( - Y_0 \right) \right),
\label{eq:fn:opmirror2}\end{aligned}$$ for $$\Delta_0 \: = \: a_0 \beta_1 - b_0 \gamma_1.$$ Using the $\Upsilon$ constraints to eliminate $Y_0$, $Y_w$, we have $$\begin{aligned}
\exp\left( - Y_w \right) & = & q_2 \exp\left( + Y_s \right),
\\
\exp\left( - Y_0 \right) & = & q_1 \exp\left( + Y_1 \right)
\exp\left( + n Y_w \right) \: = \:
(q_1 q_2^{-n} ) \exp\left( + Y_1 \right) \exp\left( - n Y_s | \right) \: = \: q_2,
\label{eq: fn: qsc}$$ which precisely match the known quantum sheaf cohomology ring relative for this subject [ @McOrist:2007kp; @McOrist:2008ji; @Donagi:2011uz; @Donagi:2011va ].
Next, we integrate out some of the fields to find a low - energy effective Landau - Ginzburg description of the same purgative. If we integrate out $ F_0 $, $ F_w$, we get the constraints $ $ \begin{aligned }
a_0 \sigma \: + \: b_0 \tilde{\sigma } & = & \exp\left (- Y_0 \right),
\\
\gamma_1 \sigma \: + \: \beta_1 \tilde{\sigma } & = & \exp\left (- Y_w \right),\end{aligned}$$ which can be solve to give $ $ \begin{aligned }
\sigma & = & \frac{1}{\Delta_0 } \left (\beta_1 \exp\left (- Y_0 \right)
\: - \: b_0 \exp\left (- Y_w \right) \right),
\label{eq: fn: opmirror1 }
\\
\tilde{\sigma } & = & \frac{1}{\Delta_0 } \left (a_0 \exp\left (- Y_w \right)
\: - \: \gamma_1 \exp\left (- Y_0 \right) \right),
\label{eq: fn: opmirror2}\end{aligned}$$ for $ $ \Delta_0 \: = \: a_0 \beta_1 - b_0 \gamma_1.$$ Using the $ \Upsilon$ constraint to eliminate $ Y_0 $, $ Y_w$, we have $ $ \begin{aligned }
\exp\left (- Y_w \right) & = & q_2 \exp\left (+ Y_s \right),
\\
\exp\left (- Y_0 \right) & = & q_1 \exp\left (+ Y_1 \right)
\exp\left (+ n Y_w \right) \: = \:
(q_1 q_2^{-n }) \exp\left (+ Y_1 \right) \exp\left (- newton Y_s | \rihht) \: = \: q_2,
\label{eq:fn:qsc}$$ whigh precisely matey the nnown suantum rheaf cohomology ring relatilnw for this case [@McOrist:2007kp; @MzOrist:2008ji; @Fonagi:2011uz; @Donegi:2011va].
Next, we intxfrate omc soms of chx fields to finc a lower-etergy effectiva Uaudau-Ginzburg description of the same physicx. Lf we integratg out $S_0$, $F_w$, we get the constraints $$\begin{alignsd}
a_0 \sigka \: + \: b_0 \tilde{\xigma} & = & \exp\left( - Y_0 \right),
\\
\gwmma_1 \sigma \: + \: \beta_1 \tilfe{\sigma} & = & \gsp\lqdt( - Y_w \right),\dnd{aligned}$$ which can bg solved to give $$\begin{aligned}
\sigoa & = & \frac{1}{\Delta_0} \lwft( \teta_1 \exp\lefv( - Y_0 \rpght)
\: - \: b_0 \exp\lcgt( - Y_w \right) \tight),
\label{eq:fn:ppmmrroe1}
\\
\tilde{\sigma} & = & \frac{1}{\Denta_0} \left( a_0 \exp\left( - Y_w \righd)
\: - \: \gamma_1 \exp\left( - T_0 \eight) \rigvt),
\lacwl{ed:fn:kpkidror2}\enf{almgned}$$ for $$\Dslta_0 \: = \: a_0 \bwta_1 - b_0 \gamma_1.$$ Using uhe $\Llsilon$ constdaints tj eliminate $Y_0$, $Y_w$, we have $$\begin{aligned}
\exk\left( - Y_w \right) & = & q_2 \exp\left( + Y_s \right),
\\
\exp\left( - Y_0 \tight) & = & q_1 \exp\left( + Y_1 \right)
\exp\left( + n Y_w \right) \: = \:
(q_1 q_2^{-n} ) \exp\laft( + B_1 \figkb) \dzp\peft( - n Y_s | \right) \: = \: q_2, \label{eq:fn:qsc}$$ which the quantum sheaf ring relations for @Donagi:2011va]. we integrate out of the fields find a lower-energy effective Landau-Ginzburg description the same physics. If we integrate out $F_0$, $F_w$, we get the constraints a_0 \sigma \: + \: b_0 \tilde{\sigma} & = & \exp\left( - Y_0 \\ \sigma + \beta_1 \tilde{\sigma} & = & \exp\left( - Y_w \right),\end{aligned}$$ which can be solved to give $$\begin{aligned} & = & \frac{1}{\Delta_0} \left( \beta_1 \exp\left( - \right) \: - \: \exp\left( - Y_w \right) \right), \\ & = \frac{1}{\Delta_0} a_0 - Y_w \right) - \: \gamma_1 \exp\left( - Y_0 \right) \right), \label{eq:fn:opmirror2}\end{aligned}$$ for $$\Delta_0 \: = \: a_0 \beta_1 - \gamma_1.$$ Using constraints to $Y_0$, we $$\begin{aligned} \exp\left( - & = & q_2 \exp\left( + \exp\left( - Y_0 \right) & = & q_1 + Y_1 \exp\left( + n Y_w \right) \: \: (q_1 q_2^{-n} ) \exp\left( + Y_1 \right) - n Y_s | \right) \: = \: q_2,
\label{eq:fn:qsc}$$ which prEcisely matCh the KnoWn qUaNtum SheaF cohomology rinG RelaTions for this case [@McOrisT:2007kp; @McorISt:2008ji; @dOnAgi:2011uz; @donagi:2011vA].
neXT, We iNtEgRatE oUT sOme of The Fields tO find a loweR-enErGy effective LANdAu-Ginzburg DesCription of thE saMe physIcS. If WE inteGraTe out $f_0$, $F_w$, we gET the coNstraints $$\BeGIn{aligNEd}
a_0 \sigmA \: + \: B_0 \TiLde{\sIgma} & = & \exp\left( - Y_0 \right),
\\
\GAmMA_1 \sigma \: + \: \beta_1 \tildE{\sigma} & = & \ExP\LeFT( - y_w \rIghT),\end{aligneD}$$ wHich cAN be solvED tO GIVe $$\bEGin{aligned}
\sigMa & = & \frac{1}{\Delta_0} \LEft( \Beta_1 \exP\lEft( - y_0 \Right)
\: - \: b_0 \Exp\leFt( - y_W \riGht) \right),
\labEl{eq:Fn:opmirroR1}
\\
\tilde{\SIgma} & = & \fraC{1}{\delta_0} \leFt( a_0 \exp\LefT( - Y_w \RighT)
\: - \: \GaMmA_1 \exP\lEFt( - Y_0 \RIgHt) \rIGht),
\Label{eq:fN:oPmIrror2}\End{aLIGNEd}$$ foR $$\DeLta_0 \: = \: a_0 \Beta_1 - b_0 \Gamma_1.$$ Using the $\upsIlon$ COnsTrainTs to eLimiNaTe $Y_0$, $Y_w$, We have $$\Begin{AlIgned}
\exp\left( - Y_w \rIght) & = & Q_2 \exp\left( + Y_S \riGhT),
\\
\exP\lEft( - Y_0 \rIGht) & = & q_1 \exP\leFt( + Y_1 \Right)
\exP\left( + n Y_W \RigHt) \: = \:
(Q_1 Q_2^{-N} ) \ExP\left( + Y_1 \right) \exp\left( - N Y_S | \right) \: = \: q_2,
\lab el{eq:fn:q sc}$$ wh ich p reci sely match the kno w n qu antum sheaf cohomology ring r e lati o ns forthis ca s e[ @ McO ri st :20 07 k p; @McO ris t:2008j i; @Donagi :20 11 uz; @Donagi: 2 01 1va].
Nex t,we integrate ou t some o f t h e fie lds to f ind al ower-e nergy eff ec t ive La n dau-Gin z b ur g de scription of thes am e physics. If w e inte gr a te o ut$F_ 0$, $F_w$, w e get the con s tr a i n ts$ $\begin{align ed}
a_0 \si g ma\: + \ :b_0 \tilde {\sig ma } &= & \exp\le ft(- Y_0 \ri ght),\ \
\gamm a _1 \sig ma \:+ \ : \ beta _ 1\t ild e{ \ sig m a} &= &\exp\lef t( - Y_w\rig h t ) , \end {al igne d}$$which can besol vedt o g ive $ $\beg in{a li gned}
\sigm a & = & \frac{1}{\Delt a_0} \left( \ bet a_ 1 \ ex p\lef t ( - Y_ 0 \ rig ht)
\:- \: b_ 0 \e xp \ l e ft ( - Y_w \right) \r ig h t ),
\label{ eq:fn: o pm ir r or1}
\\\t ild e{\s i g ma} & = & \f rac{1}{\ Delta_ 0 }\l eft( a_ 0\exp\l ef t(- Y _w \r i ght)
\: -\: \gamm a_1 \ e xp\left( - Y_0 \right) \righ t ), \ la b el{e q:f n:opmirror2 }\en d {ali gned } $$ fo r $$\D elta_ 0\ := \: a_0 \beta_1 - b _0 \gamm a_1.$ $ Using the $ \Upsilon$c o n straints toe li m inate $Y_0$, $ Y_w$, we have $ $ \begin{a ligne d}
\exp\ left( - Y _ w \right) &= & q_ 2 \ e x p\ left( + Y_s \ r i ght) ,\\
\exp \le ft( - Y _0\ri ght ) & = & q_1 \e xp\left( + Y _1 \ rig ht)
\ e xp\left( + nY_ w \ right ) \: =\:
(q _1 q _2 ^{ - n}) \exp\ l ef t ( + Y _1 \ righ t)\e xp\le ft(- nY_s | \right)_\: =_\: q_2,
\label{eq:fn:qsc}$$ which precisely_match the_known_quantum sheaf_cohomology_ring relations for_this case [@McOrist:2007kp;_@McOrist:2008ji; @Donagi:2011uz; @Donagi:2011va].
Next, we_integrate out some_of_the fields to find a lower-energy effective Landau-Ginzburg description of the same physics. If_we_integrate out_$F_0$,_$F_w$,_we get the constraints $$\begin{aligned}
a_0_\sigma \: + \: b_0_\tilde{\sigma} &_= & \exp\left( - Y_0 \right),
\\
\gamma_1 \sigma \:_+_\: \beta_1 \tilde{\sigma}_& = & \exp\left( - Y_w \right),\end{aligned}$$ which can_be solved to give $$\begin{aligned}
\sigma &_= & \frac{1}{\Delta_0}_\left(_\beta_1_\exp\left( - Y_0 \right)
\:_- \: b_0 \exp\left( - Y_w_\right) \right),
\label{eq:fn:opmirror1}
\\
\tilde{\sigma} & = & \frac{1}{\Delta_0}_\left( a_0 \exp\left( - Y_w \right)
\: -_\: \gamma_1 \exp\left( - Y_0 \right)_\right),
\label{eq:fn:opmirror2}\end{aligned}$$ for $$\Delta_0 \: =_\: a_0_\beta_1 - b_0 \gamma_1.$$ Using_the $\Upsilon$ constraints_to eliminate_$Y_0$, $Y_w$, we_have $$\begin{aligned}
\exp\left( - Y_w \right) &_= & q_2_\exp\left( + Y_s \right),
\\
\exp\left( - Y_0_\right)_& = &_q_1_\exp\left(_+ Y_1_\right)
\exp\left( + n_Y_w_\right) \:_=_\:
(q_1 q_2^{-n} ) \exp\left( + Y_1_\right)_\exp\left( - n Y_s |
ErrorRegions\]
--------------------------------------------
Table \[Table\_Boxes\] gives the description of the final IPN error regions for 254 Konus short bursts (this sample does not include the 17 imaged bursts). The nine columns contain the following information: (1) the burst designation (see Table \[Table\_Basic\]), (2) the number of error regions for the burst, $N_{r}$: 1 or 2, (3) the number of corners of the region, $N_{c}$, (4) the region type: ‘B’ (box), ‘LB’ (long box: box with maximum dimension $>10^\circ$), ‘S’ (segment), or ‘A’ (annulus), (5) the area (for two regions the sum of their areas), (6) the maximum dimension of the region (that is the maximum angular distance between two points at the region boundary; for segments larger than half an annulus, this is just the outer diameter of the annulus), (7) the right ascension of the center of the error region, in the first row, and the right ascensions of the corners in the following $N_{c}$ rows, and (8) the declination of the center of the error region, in the first row, and the declinations of the corners in the following $N_{c}$ rows (if there are two error regions, additional $N_{c}$+1 rows are given: the center of the second region and its corners, so the total number of rows for such a burst is 2($N_{c}$+1)). All coordinates are J2000.
In general, a simple, four-corner error region description is inaccurate and the curvature of the annuli should be taken into account. Only in cases where the maximum dimension of the error region is less than several degrees, can the box be reasonably well represented by its four corners. In other cases, especially when the region is a long arc or annulus segment, the given corners and center are intended to roughly indicate the position of the region on the sky. Figures showing the IPN localization (all derived annuli and the resulting error region(s)) can be found at the Ioffe Web site [^5].
A histogram of IPN error region areas is shown in Figure \[Fig\_AreaStat\]. For bursts observed by distant s/c the areas range from 2.40$\times 10^{-4}$ sq. deg (0.86 arcmin | ErrorRegions\ ]
--------------------------------------------
Table \[Table\_Boxes\ ] gives the description of the final IPN error regions for 254 Konus light burst (this sample does not include the 17 imaged fusillade). The nine columns contain the following data: (1) the burst designation (see Table \[Table\_Basic\ ]), (2) the number of mistake regions for the burst, $ N_{r}$: 1 or 2, (3) the issue of corners of the region, $ N_{c}$, (4) the area type: ‘ B ’ (box), ‘ LB ’ (long box: box with maximum dimension $ > 10^\circ$), ‘S ’ (segment), or ‘ angstrom ’ (annulus), (5) the area (for two regions the sum of their sphere), (6) the maximum dimension of the region (that is the maximum angular distance between two point at the region boundary; for segments larger than half an annulus, this is barely the outer diameter of the annulus), (7) the right ascension of the center of the error region, in the first row, and the right ascensions of the corners in the following $ N_{c}$ rows, and (8) the declination of the center of the error area, in the inaugural row, and the declination of the corner in the following $ N_{c}$ rows (if there are two mistake regions, additional $ N_{c}$+1 rows are given: the plaza of the second region and its corners, so the total number of rows for such a burst is 2($N_{c}$+1) ). All coordinate are J2000.
In general, a simple, four - corner error region description is inaccurate and the curvature of the annuli should be taken into report. Only in cases where the maximum dimension of the error area is less than several degrees, can the box be reasonably well represented by its four corners. In other cases, especially when the area is a long arc or annulus segment, the given corner and center are intend to roughly bespeak the position of the area on the sky. Figures showing the IPN localization (all derive annuli and the result error region(s) ) can be discover at the Ioffe Web site [ ^5 ].
A histogram of IPN error region area is shown in Figure \[Fig\_AreaStat\ ]. For bursts detect by distant s / c the areas range from 2.40$\times 10^{-4}$ sq. deg (0.86 arcmin | ErrlrRegions\]
--------------------------------------------
Table \[Table\_Boxes\] gives the desctiption mf the final ION error regions for 254 Konus dhirt byrsts (this sample does not incllde the 17 umagtd bursts). The ninx columns contzln thz hollowing inforkation: (1) tha burst designdtkou (see Table \[Table\_Basic\]), (2) the number of error tehions for the fursu, $N_{w}$: 1 od 2, (3) the number of corners of the rsgion, $N_{b}$, (4) the region typr: ‘B’ (box), ‘LB’ (long box: box wihh mwximum dimension $>10^\clrc$), ‘S’ (segmebt), ow ‘A’ (annulus), (5) ghe area (for two regiohs the sum of their areas), (6) the oaximbm dimensiob if hve region (tiat is the maximum angular distanve between two pomnts at the region boundacy; for segments larggr than hanf an annulus, this us just the outde dkamttec or the wnnnlus), (7) the rjght ascensuon of the center og eye error regikn, in ehq first row, and the right ascensions of ths corners in the followung $N_{c}$ rows, and (8) the feclinatijn of the center of the error region, in the first row, ena tkc awcpinations of the corners in the following $N_{c}$ wkwx (pf there are two crror regions, addiyilnsj $N_{c}$+1 rows are given: ths center of the sefond redion qnd its cjrnets, so the total number of riws for such q burst is 2($N_{c}$+1)). All eoordinates cre J2000.
Im genrral, a simple, four-cornex errod region dedcription ks inaccurate ana tme wurvaturt of the annuli shjuld be teken nnto accuunt. Only yn cases wjere bve maximum dimensiln of tve error rfgion is less than several degrxxs, can the boc te geasonablv well represented fy its four cotners. In jther cases, espvcially wien the regijn is a long dtc or annulus segment, the givwn corndfs and center sre intenbtd to rougyly indicate the ppsijikn of the regiou in the sky. Figutes shjwpng tra IPN localisatiun (xkl defived annull avd tne resulting error ragioh(s)) can be found at tme Ioffe Qeb site [^5].
W histogram og IPN error region areav ir shoen yn Figure \[Fig\_AreaStat\]. For bursta observef bn distant s/c ehe qreas range yrom 2.40$\times 10^{-4}$ sq. deg (0.86 arcmin | ErrorRegions\] -------------------------------------------- Table \[Table\_Boxes\] gives the description final error regions 254 Konus short include 17 imaged bursts). nine columns contain following information: (1) the burst designation Table \[Table\_Basic\]), (2) the number of error regions for the burst, $N_{r}$: 1 2, (3) the number of corners of the region, $N_{c}$, (4) the region ‘B’ ‘LB’ box: with maximum dimension $>10^\circ$), ‘S’ (segment), or ‘A’ (annulus), (5) the area (for two regions the of their areas), (6) the maximum dimension of region (that is the angular distance between two points the boundary; for larger half annulus, this is the outer diameter of the annulus), (7) the right ascension of the center of the error region, the first the right of corners the following $N_{c}$ (8) the declination of the center region, in the first row, and the declinations the corners the following $N_{c}$ rows (if there two error regions, additional $N_{c}$+1 rows are given: center of the second region and its corners, so the total number of rows for burst is 2($N_{c}$+1)). All are J2000. In a four-corner region is inaccurate the curvature of the annuli should be taken into account. Only cases where the maximum dimension of the error region is several can the box reasonably well represented by four In other cases, especially region a annulus the corners and center are to roughly indicate the position the region on the localization (all derived annuli and the resulting error can be found at the Ioffe Web [^5]. A histogram of IPN error region areas is shown in Figure For bursts distant s/c the areas range from 2.40$\times 10^{-4}$ deg (0.86 arcmin | ErrorRegions\]
--------------------------------------------
Table \[Table\_BoxEs\] gives the DescrIptIon Of The fInal iPN error regionS For 254 KOnus short bursts (this samPle doEs NOt inCLuDe the 17 Imaged bURsTS). the NiNe ColUmNS cOntaiN thE followIng informaTioN: (1) tHe burst desigNAtIon (see TablE \[TaBle\_Basic\]), (2) the nUmbEr of erRoR reGIons fOr tHe burSt, $N_{r}$: 1 or 2, (3) THe numbEr of corneRs OF the reGIon, $N_{c}$, (4) thE REgIon tYpe: ‘B’ (box), ‘LB’ (long box: bOX wITh maximum dimenSion $>10^\ciRc$), ‘s’ (SeGMEnt), Or ‘A’ (Annulus), (5) the ArEa (for TWo regioNS tHE SUm oF Their areas), (6) the Maximum dimeNSioN of the ReGioN (That is The maXiMUm aNgular distaNce bEtween two Points AT the regIOn boundAry; for SegMenTs laRGeR tHan HaLF an ANnUluS, ThiS is just tHe OuTer diAmetER OF The aNnuLus), (7) tHe rigHt ascension of The CentER of The erRor reGion, In The fiRst row, And thE rIght ascensions oF the Corners in The FoLloWiNg $N_{c}$ rOWs, and (8) tHe dEclInation Of the ceNTer Of THE ErRor region, in the firsT rOW, AnD the declInatioNS oF tHE corners In The FollOWIng $N_{c}$ Rows (IF tHere are tWo erroR ReGiOns, addiTiOnal $N_{c}$+1 RoWs aRe gIven: tHE cenTer of tHe second RegioN And its corners, sO The total numbeR Of ROWs FOr suCh a Burst is 2($N_{c}$+1)). AlL cooRDinaTes aRE J2000.
in gENeral, A simpLe, FOuR-Corner error region deScRiptioN is inAccurate and thE curvature OF THe annuli ShouLD bE Taken into accouNt. OnlY in cases whERe the maxImum dImension Of the erroR REgion is lEss ThaN seVerAL DeGrees, can the boX BE reaSoNably weLl rEpresenTed By iTs fOur CoRners. In otHer cases, EsPeCiAlLy wHen thE Region is A lOng ArC or AnnulUS segmeNt, the GiveN cOrNErs And centER aRE InteNdEd To roUghLy IndicAte tHE poSition oF the regioN on THe skY. FIgUres shoWing the IPN locAlIzation (all DeRivEd annuLI And the reSulting error region(s)) can bE Found at The ioffe web sIte [^5].
A histoGraM of IPN ErrOR regioN areas Is shoWn In FIGUre \[FiG\_aReAStAt\]. for bursts oBSErvEd by dIsTant S/c the arEas range from 2.40$\times 10^{-4}$ sQ. Deg (0.86 Arcmin | ErrorRegions\]
----------- ---------- ----- --- --- -- ---- ---- --
Table \[Ta b le\_ Boxes\] gives the desc ripti on of t h efinal IPN er r or r egi on sfor 2 5 4Konus sh ort bur sts (thissam pl e does not i n cl ude the 17 im aged bursts) . T he nin ecol u mns c ont ain t he fol l owinginformati on : (1) t h e burst d es igna tion (see Table \ [ Ta b le\_Basic\]),(2) th en um b e r o f e rror regio ns fort he burs t ,$ N _ {r} $ : 1 or 2, (3) the number ofcorner soft he reg ion,$N _ {c} $, (4) theregi on type:‘B’ (b o x), ‘LB ’ (longbox: b oxwit h ma x im um di me n sio n $ >10 ^ \ci rc$), ‘S ’(s egmen t),o r ‘ A’ ( ann ulus ), (5 ) the area (f ortwor egi ons t he su m of t heirareas) , (6) t he maximum dime nsio n of thereg io n ( th at is the ma xim umangular distan c e b et w e e ntwo points at there g i on boundar y; for se gm e nts larg er th an h a l f anannu l us , this i s just th eouter d ia meterof th e a nnulu s ), ( 7) the right a scens i on of the cent e r of the erro r r e g io n , in th e first row , an d the rig h tasc e nsion s ofth e c o rners in the follow in g $N_{ c}$ r ows, and (8)the declin a t i on of th e ce n te r of the errorregio n, in thef irst row , and the dec linations o f the co rne rsinthe f ol lowing $N_{c} $ rows ( if ther e a re twoerr orreg ion s, addition al $N_{c }$ +1 r ow s a re gi v en: thece nte rofthe s e cond r egion and i ts cor ners, s o t h e tot al n umbe r o frowsfors uch a burs t is 2($N _{c } $+1) ). A ll coor dinates are J 20 00.
In ge ne ral , a si m p le, four -corner error region de s criptio n i s ina ccur ate and t hecurvat ure of the annul i sho ul d b e taken i nt o a cc ount. Only i n c aseswh erethe max imum dimension oft heerror regionisless t ha n s e ve r alde g ree s , can the box be reasonabl yw el l represen t edby its fo ur corn ers.I n other cases, e specially w hent h e r egion is a long ar c or annu l us se g me nt, t hegivenco rne rs an d cent e r a re in tended t o roug hly i nd icate th e position of the regio n on t he sk y.Figures s how i ngthe IPN l ocal ization (a llder ivedann u li an d th e r esu l tingerro r region(s ) )can b efound at th e I off e Web si t e [^5] .
A histogram of IPN error region a reas i s s how n inFi gure \[Fig\_Ar eaS ta t \ ]. For b ur sts observe d by dis ta n t s/c the a reas r ange fr o m 2 . 40$\ti mes10^ {-4}$ sq. de g( 0.86 ar cm in | ErrorRegions\]
--------------------------------------------
Table \[Table\_Boxes\] gives_the description_of the final IPN_error regions_for_254 Konus_short_bursts (this sample_does not include_the 17 imaged bursts)._The nine columns_contain_the following information: (1) the burst designation (see Table \[Table\_Basic\]), (2) the number of error_regions_for the_burst,_$N_{r}$:_1 or 2, (3) the_number of corners of the_region, $N_{c}$,_(4) the region type: ‘B’ (box), ‘LB’ (long_box:_box with maximum_dimension $>10^\circ$), ‘S’ (segment), or ‘A’ (annulus), (5) the_area (for two regions the sum_of their areas),_(6)_the_maximum dimension of the_region (that is the maximum angular_distance between two points at the_region boundary; for segments larger than half_an annulus, this is just the_outer diameter of the annulus),_(7) the_right ascension of the center_of the error_region, in_the first row,_and the right ascensions of the_corners in the_following $N_{c}$ rows, and (8) the_declination_of the center_of_the_error region,_in the first_row,_and the_declinations_of the corners in the following_$N_{c}$_rows (if there are two error regions,_additional $N_{c}$+1 rows are_given:_the center of the_second region and its corners,_so the total number of rows_for such_a burst_is 2($N_{c}$+1)). All coordinates are J2000.
In general, a simple, four-corner error_region description is inaccurate and the_curvature of the annuli_should be_taken_into account. Only_in_cases where_the maximum dimension of the error region_is less_than several degrees, can the box_be reasonably well represented_by_its four corners. In other cases,_especially when the region is a_long arc or annulus segment,_the_given_corners and center are intended_to roughly indicate the position of_the region on_the sky. Figures showing the IPN localization_(all_derived annuli and the resulting error_region(s))_can be found at the Ioffe_Web_site [^5].
A_histogram of IPN error region_areas is shown in Figure \[Fig\_AreaStat\]. For_bursts observed by distant s/c the areas range from_2.40$\times 10^{-4}$ sq. deg (0.86_arcmin |
eps){width="9cm"}
In the insulating regime (x$>$0.175), the up-shift of $T_S$ induced by oxygen isotope exchange can be explained to originate from that substitution of $^{16}O$ by $^{18}O$ leads to a decrease in frequency of phonon, and consequently enhances effective mass of electron ($m^*$) through strong electron-phonon coupling. The enhancement of $m^*$ leads to a decrease of bandwidth W, resulting in an increase of energy difference $\Delta E$ between IS and LS states and consequently an up-shift of $T_s$. The increase of Sm-doping leads to a decrease in the ionic radius of A site $(Pr_{1-x}Sm_x)_{0.7}Ca_{0.3}$ in the $(Pr_{1-x}Sm_x)_{0.7}Ca_{0.3}CoO_3$ system, so that volume of the $CoO_6$ octahedra decreases. The decrease in volume of the $CoO_6$ octahedra results in an increase of crystal field splitting energy $\Delta_C$ and $\Delta E$, causing an increase in $T_S$. Therefore, decrease of the isotope component $\alpha_s$ with increasing x could be due to the increase of $T_s$. In contrast to the giant oxygen isotope effect in the insulating phase, a negligible oxygen isotope effect on $T_c$ is observed in the ferromagnetic metal regime. It has been reported that substitution of $^{16}O$ by $^{18}O$ leads to a giant down-shift of Curie temperature in $La_{1-x}Ca_xMnO_{3+y}$.[@zhao] Such giant oxygen isotope shift was explained by the JT polaron formed due to strong Jahn-Teller effect.[@zhao] Recently, elastic and inelastic neutron scattering showed that spin activation of $Co^{+3}$ ions in perovskite cobalt oxides induces local static JT distortions in the insulating phase, while the static JT distortions are absent in the ferromagnetic metal phase.[@louca1] This observation explained that the giant oxygen isotope effect is observed in the insulating phase, while negligible in the ferromagnetic metal phase well. The lattice distortions from the JT effect with strong electron-phonon coupling should be responsible for the giant isotope effect in the insulating phase.
In conclusion, oxygen isotope effect is systematically studied | eps){width="9 cm " }
In the insulating regime (x$>$0.175), the up - shift of $ T_S$ induced by oxygen isotope substitution can be excuse to originate from that substitution of $ ^{16}O$ by $ ^{18}O$ leads to a decrease in frequency of phonon, and consequently enhances effective mass of electron ($ m^*$) through potent electron - phonon coupling. The enhancement of $ m^*$ leads to a decrease of bandwidth W, result in an increase of energy deviation $ \Delta E$ between IS and LS states and consequently an up - shift key of $ T_s$. The increase of Sm - doping moderate to a decrease in the ionic spoke of A site $ (Pr_{1 - x}Sm_x)_{0.7}Ca_{0.3}$ in the $ (Pr_{1 - x}Sm_x)_{0.7}Ca_{0.3}CoO_3 $ system, so that bulk of the $ CoO_6 $ octahedra decreases. The decrease in volume of the $ CoO_6 $ octahedra consequence in an increase of crystal field separate energy $ \Delta_C$ and $ \Delta E$, causing an increase in $ T_S$. Therefore, decrease of the isotope component $ \alpha_s$ with increase x could be due to the increase of $ T_s$. In contrast to the giant oxygen isotope effect in the insulate phase, a negligible oxygen isotope effect on $ T_c$ is observed in the ferromagnetic metal regime. It has been reported that substitution of $ ^{16}O$ by $ ^{18}O$ leads to a giant down - shift of Curie temperature in $ La_{1 - x}Ca_xMnO_{3+y}$.[@zhao ] Such giant oxygen isotope shift was explained by the JT polaron formed due to strong Jahn - Teller effect.[@zhao ] Recently, elastic and inelastic neutron scattering showed that spin activation of $ Co^{+3}$ ion in perovskite cobalt oxide induces local static JT distortions in the insulate phase, while the static JT distortions are absent in the ferromagnetic metallic element phase.[@louca1 ] This observation explained that the giant oxygen isotope effect is watch in the insulating phase, while negligible in the ferromagnetic metal phase well. The lattice distortions from the JT effect with strong electron - phonon yoke should be responsible for the giant isotope effect in the insulating phase.
In conclusion, oxygen isotope effect is systematically studied | eps){aidth="9cm"}
In the insulating regime (x$>$0.175), the uk-syift oh $T_S$ insuced by oxygen isotope exchange can bw expoained to originate frum that slbstitutiin oh $^{16}O$ by $^{18}O$ leads to a decrccse ih frexnency of phonon, and consexuently enhancas eyfective mass of electron ($m^*$) through ftrong rlfctron-phonon cjuplpnd. Ths enhancement of $m^*$ leads to a decrsase of bandwidth W, tesulting in an increase ov enfrgy difference $\Depta E$ betwegh IF and LS statds and consequently an up-shift of $T_s$. The increase of Rm-dopnng leads ti q dfwrease in tie ionpc radius of A site $(Ps_{1-x}Sm_x)_{0.7}Ca_{0.3}$ in the $(Pr_{1-x}Sm_x)_{0.7}Ga_{0.3}CoO_3$ sywtem, so that volume oh the $CoO_6$ octahedra qecreases. Tke decrease in volume od the $CoO_6$ octxyedfa dexumts in an increase kf crystal dield splitting enetgr $\Delta_C$ and $\Dslta E$, cwusing an increase in $T_S$. Therefore, decrtase kf the isotope componenr $\alpha_s$ with increaslng x coujd be due to the increase of $T_s$. In contrast to tha giait oxvncn iritlpe effect in the insulating phase, a negligibjs pxjgen isotope effegt on $T_c$ is observrd im the ferromagvetic metzl regime. It has bfen repjrted that subftitition of $^{16}O$ by $^{18}O$ leads to a tiant down-shpft if Curie temperatuxe in $La_{1-x}Ca_xOnO_{3+y}$.[@ehao] Sich giant oxygen isotopz shiff was explalned by tgd JT polaron foroed dge to strong Jahn-Teller efsect.[@zhao] Cecencly, elasgic snd inqlastic nektron scattering showed hhat dphn activatlon of $Co^{+3}$ ions in perovskite couelt oxides incuwes local scatic MT distortions yn the insulatnng phasz, whild the statpc JT disvortions are absent in tha ferromagnetmc metal [hasw.[@louxa1] This ubservation exllained tkct the giqnt oxygen isotope efwsct is observed in the insulating phxse, wjike tegligible it thd fdtromaenetic metao phaxe well. The lattice gistkrtions from the JY cffect wijh strong electron-phonpn coupling should be rxsponsmble fpr jhe giant isotope effect in the insulatijg ihase.
In concltsiok, oxigen isotopz effect is systematically studied | eps){width="9cm"} In the insulating regime (x$>$0.175), the $T_S$ by oxygen exchange can be substitution $^{16}O$ by $^{18}O$ to a decrease frequency of phonon, and consequently enhances mass of electron ($m^*$) through strong electron-phonon coupling. The enhancement of $m^*$ leads a decrease of bandwidth W, resulting in an increase of energy difference $\Delta between and states consequently an up-shift of $T_s$. The increase of Sm-doping leads to a decrease in the ionic of A site $(Pr_{1-x}Sm_x)_{0.7}Ca_{0.3}$ in the $(Pr_{1-x}Sm_x)_{0.7}Ca_{0.3}CoO_3$ system, that volume of the octahedra decreases. The decrease in of $CoO_6$ octahedra in increase crystal field splitting $\Delta_C$ and $\Delta E$, causing an increase in $T_S$. Therefore, decrease of the isotope component $\alpha_s$ with x could to the of In to the giant effect in the insulating phase, a effect on $T_c$ is observed in the ferromagnetic regime. It been reported that substitution of $^{16}O$ $^{18}O$ leads to a giant down-shift of Curie in $La_{1-x}Ca_xMnO_{3+y}$.[@zhao] Such giant oxygen isotope shift was explained by the JT polaron formed due Jahn-Teller effect.[@zhao] Recently, elastic inelastic neutron scattering that activation $Co^{+3}$ in perovskite oxides induces local static JT distortions in the insulating phase, while static JT distortions are absent in the ferromagnetic metal phase.[@louca1] explained the giant oxygen effect is observed in insulating while negligible in the phase The the effect strong electron-phonon coupling should responsible for the giant isotope in the insulating phase. is systematically studied | eps){width="9cm"}
In the insulating Regime (x$>$0.175), the Up-shiFt oF $T_S$ InDuceD by oXygen isotope exCHangE can be explained to origiNate fRoM That SUbStituTion of $^{16}O$ BY $^{18}O$ LEAds To A dEcrEaSE iN freqUenCy of phoNon, and consEquEnTly enhances eFFeCtive mass oF elEctron ($m^*$) throuGh sTrong eLeCtrON-phonOn cOupliNg. The eNHancemEnt of $m^*$ leaDs TO a decrEAse of baNDWiDth W, Resulting in an incrEAsE Of energy differEnce $\DeLtA e$ bETWeeN IS And LS stateS aNd conSEquentlY An UP-SHifT Of $T_s$. The increaSe of Sm-dopinG LeaDs to a dEcReaSE in the Ionic RaDIus Of A site $(Pr_{1-x}SM_x)_{0.7}Ca_{0.3}$ In the $(Pr_{1-x}SM_x)_{0.7}Ca_{0.3}Coo_3$ System, sO That volUme of tHe $COO_6$ oCtahEDrA dEcrEaSEs. THE dEcrEAse In volume Of ThE $CoO_6$ oCtahEDRA ResuLts In an IncreAse of crystal fIelD splITtiNg eneRgy $\DeLta_C$ AnD $\DeltA E$, causIng an InCrease in $T_S$. ThereFore, Decrease oF thE iSotOpE compONent $\alPha_S$ wiTh increAsing x cOUld Be DUE To The increase of $T_s$. In cOnTRAsT to the giAnt oxyGEn IsOTope effeCt In tHe inSULatinG phaSE, a NegligibLe oxygEN iSoTope effEcT on $T_c$ iS oBseRveD in thE FerrOmagneTic metal RegimE. it has been reporTEd that substitUTiON Of $^{16}o$ By $^{18}O$ lEadS to a giant doWn-shIFt of curiE TeMpeRAture In $La_{1-x}ca_Xmno_{3+Y}$.[@zhao] Such giant oxygeN iSotope Shift Was explained bY the JT polaRON Formed duE to sTRoNG Jahn-Teller effEct.[@zhAo] Recently, ELastic anD inelAstic neuTron scattERIng showeD thAt sPin ActIVAtIon of $Co^{+3}$ ions in PERovsKiTe cobalT oxIdes indUceS loCal StaTiC JT distorTions in tHe InSuLaTinG phasE, While the StAtiC Jt diStortIOns are AbsenT in tHe FeRRomAgnetic MEtAL PhasE.[@lOuCa1] ThIs oBsErvatIon eXPlaIned thaT the giant OxyGEn isOtOpE effect Is observed in tHe Insulating PhAse, While nEGLigible iN the ferromagnetic metal pHAse well. the LattiCe diStortions FroM the JT EffECt with Strong ElectRoN-phONOn couPLInG shOuLd be responSIBle For thE gIant Isotope Effect in the insulatINg pHase.
In conclusIon, OxygEN IsOtoPE eFFecT iS SysTEMatically studieD | eps){width="9cm"}
In theinsulating regi me(x$ >$ 0.17 5),the up-shift o f $T_ S$ induced by oxygen i sotop ee xcha n ge canbe expl a in e d to o ri gin at e f rom t hat substi tution of$^{ 16 }O$ by $^{18 } O$ leads toa d ecrease in f req uencyof ph o non,and cons equent l y enha nces effe ct i ve mas s of ele c t ro n ($ m^*$) through str o ng electron-phono n coup li n g. T heenh ancement o f$m^*$ leads t o a d e cre a se of bandwid th W, resul t ing in an i ncr e ase of ener gy dif ference $\D elta E$ betwe en ISa nd LS s t ates an d cons equ ent ly a n u p- shi ft of$ T_ s$. The increas eof Sm-d opin g l e adstoa de creas e in the ioni c r adiu s of A si te $( Pr_{ 1- x}Sm_ x)_{0. 7}Ca_ {0 .3}$ in the $(P r_{1 -x}Sm_x)_ {0. 7} Ca_ {0 .3}Co O _3$ sy ste m,so that volume ofth e $ Co O_6$ octahedra dec re a s es . The de crease in v o lume ofth e $ CoO_ 6 $ octa hedr a r esults i n an i n cr ea se of c ry stal f ie ldspl ittin g ene rgy $\ Delta_C$ and$ \Delta E$, cau s ing an increa s ei n $ T _S$. Th erefore, de crea s e of the is oto p e com ponen t$ \a l pha_s$ with increas in g x co uld b e due to theincrease o f $ T_s$. In con t ra s t to the giant oxyg en isotope effect i n the insulat ing phase , a neglig ibl e o xyg eni s ot ope effect on $ T_c$ i s obser ved in the fe rro mag net ic metal re gime. It h as b ee n r eport e d that s ub sti tu tio n of$ ^{16}O $ by$^{1 8} O$ lea ds to a gi a n t do wn -s hift of C urietemp e rat ure in$La_{1-x} Ca_ x MnO_ {3 +y }$.[@zh ao] Such gian toxygen iso to peshiftw a s explai ned by the JT polaron f o rmed du e t o str ongJahn-Tell ereffect .[@ z hao] R ecentl y, el as tic a nd in e l as tic n eutron sca t t eri ng sh ow ed t hat spi n activation of $C o ^{+ 3}$ ions in p ero vski t e c oba l to xid es ind u c es local static JT distor ti o ns in the in s ula ti ng phas e, whil e the staticJT distor tions are a bsen t inthe ferrom agneticmetal pha s e.[@l o uc a1] T his obser va tio n exp lained tha t the giant o xygenisoto pe effectis observed in the insu lating phas e,while neg lig i ble in the f erro magnetic m eta l p hasewel l . The lat t ic e d i stort ions from theJ Teff e c twith strong e l ect ron-p hon o n coup ling should be respon s ible for the g iant i sot ope effe ct in the insula tin gp h ase.
In c onclusion,oxygen i so t ope e ffectis sys tematic a l ly studie d | eps){width="9cm"}
In the_insulating regime_(x$>$0.175), the up-shift of_$T_S$ induced_by_oxygen isotope_exchange_can be explained_to originate from_that substitution of $^{16}O$_by $^{18}O$ leads_to_a decrease in frequency of phonon, and consequently enhances effective mass of electron ($m^*$)_through_strong electron-phonon_coupling._The_enhancement of $m^*$ leads to_a decrease of bandwidth W,_resulting in_an increase of energy difference $\Delta E$ between_IS_and LS states_and consequently an up-shift of $T_s$. The increase of_Sm-doping leads to a decrease in_the ionic radius_of_A_site $(Pr_{1-x}Sm_x)_{0.7}Ca_{0.3}$ in the_$(Pr_{1-x}Sm_x)_{0.7}Ca_{0.3}CoO_3$ system, so that volume of_the $CoO_6$ octahedra decreases. The decrease_in volume of the $CoO_6$ octahedra results_in an increase of crystal field_splitting energy $\Delta_C$ and $\Delta_E$, causing_an increase in $T_S$. Therefore,_decrease of the_isotope component_$\alpha_s$ with increasing_x could be due to the_increase of $T_s$._In contrast to the giant oxygen_isotope_effect in the_insulating_phase,_a negligible_oxygen isotope effect_on_$T_c$ is_observed_in the ferromagnetic metal regime. It_has_been reported that substitution of $^{16}O$ by_$^{18}O$ leads to a_giant_down-shift of Curie temperature_in $La_{1-x}Ca_xMnO_{3+y}$.[@zhao] Such giant oxygen_isotope shift was explained by the_JT polaron_formed due_to strong Jahn-Teller effect.[@zhao] Recently, elastic and inelastic neutron scattering showed_that spin activation of $Co^{+3}$ ions_in perovskite cobalt oxides_induces local_static_JT distortions in_the_insulating phase,_while the static JT distortions are absent_in the_ferromagnetic metal phase.[@louca1] This observation explained_that the giant oxygen_isotope_effect is observed in the insulating_phase, while negligible in the ferromagnetic_metal phase well. The lattice_distortions_from_the JT effect with strong_electron-phonon coupling should be responsible for_the giant isotope_effect in the insulating phase.
In conclusion, oxygen_isotope_effect is systematically studied |
F_j(X_{i,j})\} = n/(n + 1 - R_{i,j})$, where $R_{i,j}$ is the rank of $X_{i,j}$ among $X_{1,j}, \ldots, X_{n,j}$. The empirical counterparts to $\mu$ and $\Lambda$ are $$\begin{aligned}
\label{eq:mun}
\widehat{\mu}(\,\cdot\,)
&=
\frac{1}{k}\sum_{i=1}^n \delta_{ (k/n)\widehat{{\bm}V}_{i} }(\,\cdot\,), &
\widehat{\Lambda}(\,\cdot\,)
&=
\frac{1}{k}\sum_{i=1}^n \delta_{ (n/k)/\widehat{{\bm}V}_{i} }(\,\cdot\,),\end{aligned}$$ respectively, with $\delta_a$ the Dirac measure at the point $a$. Replacing $\Lambda$ by $\widehat{\Lambda}$ in the definition of $\ell_\alpha$ and $r_\alpha$ produces the empirical tail dependence function $$\begin{aligned}
\widehat{\ell}_\alpha( \bm{x} )
&= k^{-1} \textstyle\sum_{i=1}^n {{\mathbbm{1}}}\{\exists j \in \alpha : n + 1 - R_{i,j} {\leqslant}\lfloor k x_j \rfloor \} \\
&= k^{-1} \textstyle\sum_{i=1}^n {{\mathbbm{1}}}\{\exists j \in \alpha : X_{i,j} {\geqslant}X_{(n-\lfloor k x_j \rfloor +1),j} \} \end{aligned}$$ and the empirical joint tail function $$\begin{aligned}
\label{eq:r_alpha:estim}
\widehat{r}_\alpha( \bm{x} )
&= k^{-1} \textstyle\sum_{i=1}^n {{\mathbbm{1}}}\{\forall j \in \alpha : n + 1 - R_{i,j} {\leqslant}\lfloor k x_j \rfloor \} \\
\nonumber
&= k^{-1} \textstyle\sum_{i=1}^n {{\mathbbm{1}}}\{\forall j \in \alpha : X_{i,j} {\geqslant}X | F_j(X_{i, j})\ } = n/(n + 1 - R_{i, j})$, where $ R_{i, j}$ is the rank of $ X_{i, j}$ among $ X_{1,j }, \ldots, X_{n, j}$. The empirical counterparts to $ \mu$ and $ \Lambda$ are $ $ \begin{aligned }
\label{eq: mun }
\widehat{\mu}(\,\cdot\ ,)
& =
\frac{1}{k}\sum_{i=1}^n \delta _ { (k / n)\widehat{{\bm}V}_{i } } (\,\cdot\ ,), &
\widehat{\Lambda}(\,\cdot\ ,)
& =
\frac{1}{k}\sum_{i=1}^n \delta _ { (n / k)/\widehat{{\bm}V}_{i } } (\,\cdot\,),\end{aligned}$$ respectively, with $ \delta_a$ the Dirac measure at the item $ a$. replace $ \Lambda$ by $ \widehat{\Lambda}$ in the definition of $ \ell_\alpha$ and $ r_\alpha$ produces the empirical tail addiction function $ $ \begin{aligned }
\widehat{\ell}_\alpha (\bm{x })
& = k^{-1 } \textstyle\sum_{i=1}^n { { \mathbbm{1}}}\{\exists j \in \alpha: n + 1 - R_{i, j } { \leqslant}\lfloor k x_j \rfloor \ } \\
& = k^{-1 } \textstyle\sum_{i=1}^n { { \mathbbm{1}}}\{\exists j \in \alpha: X_{i, j } { \geqslant}X_{(n-\lfloor k x_j \rfloor +1),j } \ } \end{aligned}$$ and the empirical joint fag end function $ $ \begin{aligned }
\label{eq: r_alpha: estim }
\widehat{r}_\alpha (\bm{x })
& = k^{-1 } \textstyle\sum_{i=1}^n { { \mathbbm{1}}}\{\forall j \in \alpha: n + 1 - R_{i, j } { \leqslant}\lfloor k x_j \rfloor \ } \\
\nonumber
& = k^{-1 } \textstyle\sum_{i=1}^n { { \mathbbm{1}}}\{\forall j \in \alpha: X_{i, j } { \geqslant}X | F_j(D_{i,j})\} = n/(n + 1 - R_{i,j})$, where $R_{i,j}$ is the rank of $X_{i,j}$ ammng $X_{1,j}, \ldots, X_{v,j}$. The empirical counterpartd ro $\mu$ and $\Lambda$ are $$\begin{auigned}
\lwbel{eq:mub}
\qidehat{\mu}(\,\cvkt\,)
&=
\fdwc{1}{k}\sbm_{m=1}^n \delta_{ (k/n)\widemat{{\bm}V}_{i} }(\,\cdod\,), &
\widehat{\Ldmcdc}(\,\cdot\,)
&=
\frac{1}{k}\sum_{i=1}^n \delta_{ (n/k)/\widerat{{\bm}V}_{i} }(\,\cfot\,),\end{aligned}$$ tespebtyvelg, with $\delta_a$ the Dirac measure at the popnt $a$. Replacing $\Lsmbda$ by $\widehat{\Lambda}$ in hhe fefinition of $\ell_\appha$ and $r_\aopha$ produces the empirical tail dependgnce function $$\begin{aligned}
\widefat{\elk}_\alpha( \bm{x} )
&= k^{-1} \jextstyle\sum_{m=1}^n {{\matrbbm{1}}}\{\exists j \in \alphd : n + 1 - R_{i,j} {\leqslant}\lnloor k z_j \rfloor \} \\
&= k^{-1} \textsvyle\sum_{i=1}^n {{\mathbbm{1}}}\{\exifts j \in \dl'ha : X_{i,j} {\geqslant}X_{(n-\lfooir k f_j \rxloof +1),j} \} \ens{akifned}$$ ajd vhe empiriczl joint taul function $$\begin{aloggvc}
\label{eq:r_alpga:estii}
\widehat{r}_\alpha( \bm{x} )
&= k^{-1} \textstyle\sum_{i=1}^n {{\katgbbm{1}}}\{\forall j \in \alpha : b + 1 - R_{i,j} {\leqslant}\lflolr k x_j \rsloor \} \\
\nonumber
&= k^{-1} \textstyle\sum_{i=1}^n {{\mathbbm{1}}}\{\foraln j \ii \xlpkq : X_{i,g} {\gfqslant}X | F_j(X_{i,j})\} = n/(n + 1 - R_{i,j})$, is rank of among $X_{1,j}, \ldots, $\mu$ $\Lambda$ are $$\begin{aligned} \widehat{\mu}(\,\cdot\,) &= \frac{1}{k}\sum_{i=1}^n (k/n)\widehat{{\bm}V}_{i} }(\,\cdot\,), & \widehat{\Lambda}(\,\cdot\,) &= \frac{1}{k}\sum_{i=1}^n (n/k)/\widehat{{\bm}V}_{i} }(\,\cdot\,),\end{aligned}$$ respectively, with $\delta_a$ the Dirac measure at the point $a$. Replacing by $\widehat{\Lambda}$ in the definition of $\ell_\alpha$ and $r_\alpha$ produces the empirical tail function \widehat{\ell}_\alpha( ) k^{-1} \textstyle\sum_{i=1}^n {{\mathbbm{1}}}\{\exists j \in \alpha : n + 1 - R_{i,j} {\leqslant}\lfloor k x_j \rfloor \\ &= k^{-1} \textstyle\sum_{i=1}^n {{\mathbbm{1}}}\{\exists j \in \alpha X_{i,j} {\geqslant}X_{(n-\lfloor k x_j +1),j} \} \end{aligned}$$ and the joint function $$\begin{aligned} \widehat{r}_\alpha( ) k^{-1} \textstyle\sum_{i=1}^n {{\mathbbm{1}}}\{\forall \in \alpha : n + 1 - R_{i,j} {\leqslant}\lfloor k x_j \rfloor \} \\ \nonumber &= k^{-1} {{\mathbbm{1}}}\{\forall j : X_{i,j} | F_j(X_{i,j})\} = n/(n + 1 - R_{i,j})$, where $R_{i,j}$ is the raNk of $X_{i,j}$ amoNg $X_{1,j}, \lDotS, X_{n,J}$. THe emPiriCal counterpartS To $\mu$ And $\Lambda$ are $$\begin{alignEd}
\labEl{EQ:mun}
\WIdEhat{\mU}(\,\cdot\,)
&=
\frAC{1}{k}\SUM_{i=1}^n \DeLtA_{ (k/n)\WiDEhAt{{\bm}V}_{I} }(\,\cdOt\,), &
\widehAt{\Lambda}(\,\cdOt\,)
&=
\fRaC{1}{k}\sum_{i=1}^n \delta_{ (N/K)/\wIdehat{{\bm}V}_{i} }(\,\CdoT\,),\end{aligned}$$ rEspEctiveLy, WitH $\Delta_A$ thE DiraC measuRE at the Point $a$. RepLaCIng $\LamBDa$ by $\widEHAt{\lambDa}$ in the definition OF $\eLL_\alpha$ and $r_\alphA$ produCeS ThE EMpiRicAl tail depeNdEnce fUNction $$\bEGiN{ALIgnED}
\widehat{\ell}_\alPha( \bm{x} )
&= k^{-1} \textSTylE\sum_{i=1}^n {{\MaThbBM{1}}}\{\existS j \in \aLpHA : n + 1 - R_{I,j} {\leqslant}\lFlooR k x_j \rflooR \} \\
&= k^{-1} \textSTyle\sum_{I=1}^N {{\mathbbM{1}}}\{\existS j \iN \alPha : X_{I,J} {\gEqSlaNt}x_{(N-\lfLOoR k x_J \RflOor +1),j} \} \end{aLiGnEd}$$ and The eMPIRIcal JoiNt taIl funCtion $$\begin{aliGneD}
\labEL{eq:R_alphA:estiM}
\widEhAt{r}_\alPha( \bm{x} )
&= K^{-1} \textStYle\sum_{i=1}^n {{\mathbbm{1}}}\{\ForaLl j \in \alphA : n + 1 - R_{I,j} {\LeqSlAnt}\lfLOor k x_j \RflOor \} \\
\NonumbeR
&= k^{-1} \textsTYle\SuM_{I=1}^N {{\MaThbbm{1}}}\{\forall j \in \alphA : X_{I,J} {\GeQslant}X | F_j(X_{i,j})\} = n/(n + 1 - R_{i,j} )$, w her e $ R_ {i,j }$ i s the rank of$ X_{i ,j}$ among $X_{1,j}, \ ldots ,X _{n, j }$ . The empiri c al c oun te rp art st o$\mu$ an d $\Lam bda$ are $ $\b eg in{aligned} \ label{eq:m un}
\wideha t{\ mu}(\, \c dot \ ,)
&=
\fr a c{1}{k }\sum_{i= 1} ^ n \del t a_{ (k/ n ) \w ideh at{{\bm}V}_{i} }( \ ,\ c dot\,), &
\wideh at { \L a m bda }(\ ,\cdot\,) &= \fra c {1 } { k }\s u m_{i=1}^n \de lta_{ (n/k) / \wi dehat{ {\ bm} V }_{i}}(\,\ cd o t\, ),\end{alig ned} $$ respec tively , with $ \ delta_a $ theDir acmeas u re a t t he poi n t$a$ . Re placing$\ La mbda$ by$ \ w i deha t{\ Lamb da}$in the defini tio n of $\e ll_\a lpha$ and $ r_\al pha$ p roduc es the empiricaltail dependen cefu nct io n $$\ b egin{a lig ned }
\wi dehat{\ e ll} _\ a l p ha ( \bm{x} )
&= k^ {- 1 } \ textstyl e\sum_ { i= 1} ^ n {{\mat hb bm{ 1}}} \ { \exis ts j \i n \alpha : n + 1-R_{i,j} { \leqsl an t}\ lfl oor k x_j\rfloo r \} \\ &=k ^{-1} \textsty l e\sum_{i=1}^n {{ \ m at h bbm{ 1}} }\{\existsj \i n \al pha: X _{i , j} {\ geqsl an t }X _ {(n-\lfloor k x_j \ rf loor + 1),j} \} \end{alig ned}$$ and t h e empiri calj oi n t tail functio n $$\ begin{alig n ed}
\lab el{eq :r_alpha :estim}
\ widehat{ r}_ \al pha ( \ b m {x } )
&= k^{- 1 } \te xt style\s um_ {i=1}^n {{ \ma thb bm{ 1} }}\{\fora ll j \in \ al ph a: n + 1- R_{i,j} { \le qs lan t}\lf l oor kx_j \ rflo or \ } \\
\nonum b er
& =k^ {-1} \t ex tstyl e\su m _{i =1}^n { {\mathbbm {1} } }\{\ fo ra ll j \i n \alpha : X_ {i ,j} {\geqs la nt} X | F_j(X_{i,j})\}_= n/(n_+ 1 - R_{i,j})$,_where $R_{i,j}$_is_the rank_of_$X_{i,j}$ among $X_{1,j},_\ldots, X_{n,j}$. The_empirical counterparts to $\mu$_and $\Lambda$ are_$$\begin{aligned}
_ \label{eq:mun}
\widehat{\mu}(\,\cdot\,)
&=
__\frac{1}{k}\sum_{i=1}^n \delta_{_(k/n)\widehat{{\bm}V}_{i}_}(\,\cdot\,),_&
\widehat{\Lambda}(\,\cdot\,)
_ &=
_ _\frac{1}{k}\sum_{i=1}^n \delta_{ (n/k)/\widehat{{\bm}V}_{i} }(\,\cdot\,),\end{aligned}$$ respectively, with $\delta_a$ the_Dirac_measure at the_point $a$. Replacing $\Lambda$ by $\widehat{\Lambda}$ in the definition_of $\ell_\alpha$ and $r_\alpha$ produces the_empirical tail dependence_function_$$\begin{aligned}
_ \widehat{\ell}_\alpha( \bm{x} )
_ &= k^{-1} \textstyle\sum_{i=1}^n {{\mathbbm{1}}}\{\exists j_\in \alpha : n + 1_- R_{i,j} {\leqslant}\lfloor k x_j \rfloor \}_\\
&= k^{-1} \textstyle\sum_{i=1}^n {{\mathbbm{1}}}\{\exists_j \in \alpha : X_{i,j}_{\geqslant}X_{(n-\lfloor k_x_j \rfloor +1),j} \} \end{aligned}$$_and the empirical_joint tail_function $$\begin{aligned}
\label{eq:r_alpha:estim}
_\widehat{r}_\alpha( \bm{x} )
&= k^{-1}_\textstyle\sum_{i=1}^n {{\mathbbm{1}}}\{\forall j_\in \alpha : n + 1_-_R_{i,j} {\leqslant}\lfloor k_x_j_\rfloor_\} \\
\nonumber_
_&=_k^{-1} \textstyle\sum_{i=1}^n_{{\mathbbm{1}}}\{\forall_j \in \alpha : X_{i,j} {\geqslant}X |
(\[eq7.5.1\]) to the RHS of (\[eq46.4.2\]) implies that it is at most $$\biggl(1+\varepsilon+\frac{\varepsilon_{7}}{8} \biggr)|x|_{L} + 2 \biggl(\max_{r}
\frac{w_{r}}{\rho_{r}} \biggr) \Bigl(\max_{r} \nu_{r} \Bigr)|x|_{L}/C_{24}N^{2}.$$ Consequently, for small enough $\varepsilon> 0$, $$\label{eq46.4.4}
|X(t)|_{S} \le(1+\varepsilon_{7}/4)|x|_{L} \qquad\mbox{for all } t \in[0,N^{3}].$$ Together with (\[eq46.4.1\]), this implies (\[eq46.3.8\]).
Decomposition of $|X(N^{3})|_{r,s}-|x|_{r,i_{r}(s)}$ {#decomposition-of-xn3_rs-x_ri_rs.unnumbered}
----------------------------------------------------
In this short subsection, we decompose $|X(N^{3})|_{r,s}-|x|_{r,i_{r}(s)}$ into several parts, one of which contains the factor $Z^{*}_{r}(N^{3},s) - z^{*}_{r}(i_{r}(s))$. In Proposition \[prop44.1.2\], we then obtain upper bounds on this factor. In this and the remaining subsection, the estimates need to be more precise than in previous sections in order to make use of the subcriticality of $X(\cdot)$.
The decomposition that was referred to above is given by $$\begin{aligned}
\label{eq42.1.2}
&&|X(N^{3})|_{r,s} - |x|_{r,i_{r}(s)} \nonumber\\
&&\qquad= \frac
{w_{r}(1+as)(Z^{*}_{r}(N^{3},s) -
z^{*}_{r}(i_{r}(s)))}{\nu_{r}\Gamma(\sigma_{r})} \\
%
&&\qquad\quad{} - |x|_{r,i_{r}(s)}\frac{1+as}{1+ai_{r}(s)} \frac{\Gamma(\sigma
_{r}) - \Gamma
({\sigma | (\[eq7.5.1\ ]) to the RHS of (\[eq46.4.2\ ]) implies that it is at most $ $ \biggl(1+\varepsilon+\frac{\varepsilon_{7}}{8 } \biggr)|x|_{L } + 2 \biggl(\max_{r }
\frac{w_{r}}{\rho_{r } } \biggr) \Bigl(\max_{r } \nu_{r } \Bigr)|x|_{L}/C_{24}N^{2}.$$ Consequently, for small enough $ \varepsilon > 0 $, $ $ \label{eq46.4.4 }
|X(t)|_{S } \le(1+\varepsilon_{7}/4)|x|_{L } \qquad\mbox{for all } t \in[0,N^{3}].$$ in concert with (\[eq46.4.1\ ]), this entail (\[eq46.3.8\ ]).
Decomposition of $ |X(N^{3})|_{r, s}-|x|_{r, i_{r}(s)}$ { # decomposition-of-xn3_rs-x_ri_rs.unnumbered }
----------------------------------------------------
In this short subsection, we disintegrate $ |X(N^{3})|_{r, s}-|x|_{r, i_{r}(s)}$ into several share, one of which contains the factor $ Z^{*}_{r}(N^{3},s) - z^{*}_{r}(i_{r}(s))$. In Proposition \[prop44.1.2\ ], we then receive upper bounds on this factor. In this and the stay subsection, the estimates need to be more accurate than in previous sections in order to make habit of the subcriticality of $ X(\cdot)$.
The decomposition that was referred to above is given by $ $ \begin{aligned }
\label{eq42.1.2 }
& & |X(N^{3})|_{r, s } - |x|_{r, i_{r}(s) } \nonumber\\
& & \qquad= \frac
{ w_{r}(1+as)(Z^{*}_{r}(N^{3},s) -
z^{*}_{r}(i_{r}(s)))}{\nu_{r}\Gamma(\sigma_{r }) } \\
%
& & \qquad\quad { } - |x|_{r, i_{r}(s)}\frac{1+as}{1+ai_{r}(s) } \frac{\Gamma(\sigma
_ { r }) - \Gamma
({ \sigma | (\[eq7.5.1\]) to the RHS of (\[eq46.4.2\]) implier that it is at most $$\bmggl(1+\varspsilon+\ffac{\varepsilon_{7}}{8} \biggr)|x|_{L} + 2 \biggp(\mqx_{r}
\frqc{w_{r}}{\rho_{r}} \biggr) \Bigl(\max_{f} \nu_{r} \Bigg)|x|_{L}/C_{24}N^{2}.$$ Conwequtntly, for small eikugh $\vavzpsilkk> 0$, $$\layeo{eq46.4.4}
|X(t)|_{S} \le(1+\varepxilon_{7}/4)|x|_{L} \qqgad\mbox{for all } g \nn[0,N^{3}].$$ Together with (\[eq46.4.1\]), this implies (\[eq46.3.8\]).
Qecompoxihion of $|X(N^{3})|_{r,s}-|x|_{r,y_{r}(s)}$ {#cqcomlosition-of-xn3_rs-x_ri_rs.unnumbered}
----------------------------------------------------
In thjs shoru subsection, we devompose $|X(N^{3})|_{r,s}-|x|_{r,i_{r}(s)}$ into sevfral parts, one of whicj contains jge sqctor $Z^{*}_{r}(N^{3},s) - x^{*}_{r}(i_{r}(s))$. In Pgmposition \[krop44.1.2\], we then obtain upper bounds on tkis factor. Un thlv and the rxmainigg subsectiok, the evtimatex need to be mpre prwcise than in previouv sections in ordet to make gsz of the subcriticaliry of $X(\wdot)$.
Dhe awcoopoaivioh that waa referred to above iw given by $$\begin{alibnqe}
\label{eq42.1.2}
&&|X(N^{3})|_{r,s} - |x|_{r,i_{r}(s)} \njnumber\\
&&\qquad= \frac
{w_{r}(1+as)(Z^{*}_{r}(N^{3},s) -
z^{*}_{r}(i_{r}(s)))}{\nu_{r}\Gamma(\vigja_{r})} \\
%
&&\qquad\quad{} - |x|_{r,i_{r}(s)}\frax{1+as}{1+ai_{r}(s)} \frac{\Gamma(\sigmw
_{r}) - \Gamma
({\figma | (\[eq7.5.1\]) to the RHS of (\[eq46.4.2\]) implies is most $$\biggl(1+\varepsilon+\frac{\varepsilon_{7}}{8} + 2 \biggl(\max_{r} Consequently, small enough $\varepsilon> $$\label{eq46.4.4} |X(t)|_{S} \le(1+\varepsilon_{7}/4)|x|_{L} all } t \in[0,N^{3}].$$ Together with this implies (\[eq46.3.8\]). Decomposition of $|X(N^{3})|_{r,s}-|x|_{r,i_{r}(s)}$ {#decomposition-of-xn3_rs-x_ri_rs.unnumbered} ---------------------------------------------------- In this short subsection, we $|X(N^{3})|_{r,s}-|x|_{r,i_{r}(s)}$ into several parts, one of which contains the factor $Z^{*}_{r}(N^{3},s) - z^{*}_{r}(i_{r}(s))$. Proposition we obtain bounds on this factor. In this and the remaining subsection, the estimates need to be more than in previous sections in order to make of the subcriticality of The decomposition that was referred above given by \label{eq42.1.2} - \nonumber\\ &&\qquad= \frac - z^{*}_{r}(i_{r}(s)))}{\nu_{r}\Gamma(\sigma_{r})} \\ % &&\qquad\quad{} - |x|_{r,i_{r}(s)}\frac{1+as}{1+ai_{r}(s)} \frac{\Gamma(\sigma _{r}) - \Gamma ({\sigma | (\[eq7.5.1\]) to the RHS of (\[eq46.4.2\]) implies that It is at most $$\Biggl(1+\VarEpsIlOn+\frAc{\vaRepsilon_{7}}{8} \biggr)|x|_{l} + 2 \BiggL(\max_{r}
\frac{w_{r}}{\rho_{r}} \biggr) \BiGl(\max_{R} \nU_{R} \BigR)|X|_{L}/c_{24}N^{2}.$$ ConSequentLY, fOR SmaLl EnOugH $\vARePsiloN> 0$, $$\laBel{eq46.4.4}
|X(t)|_{s} \le(1+\varepsiLon_{7}/4)|X|_{L} \Qquad\mbox{for ALl } T \in[0,N^{3}].$$ TogethEr wIth (\[eq46.4.1\]), this impLieS (\[eq46.3.8\]).
DecOmPosITion oF $|X(N^{3})|_{R,s}-|x|_{r,i_{R}(s)}$ {#decoMPositiOn-of-xn3_rs-x_Ri_RS.unnumBEred}
----------------------------------------------------
In tHIS sHort Subsection, we decomPOsE $|x(N^{3})|_{r,s}-|x|_{r,i_{r}(s)}$ into sEveral PaRTs, ONE of WhiCh contains ThE factOR $Z^{*}_{r}(N^{3},s) - z^{*}_{r}(I_{R}(s))$. iN pRopOSition \[prop44.1.2\], we tHen obtain upPEr bOunds oN tHis FActor. IN this AnD The Remaining suBsecTion, the esTimateS Need to bE More preCise thAn iN prEvioUS sEcTioNs IN orDEr To mAKe uSe of the sUbCrIticaLity OF $x(\CDot)$.
THe dEcomPositIon that was refErrEd to ABovE is giVen by $$\BegiN{aLigneD}
\label{Eq42.1.2}
&&|X(N^{3})|_{r,S} - |x|_{R,i_{r}(s)} \nonumber\\
&&\qquAd= \frAc
{w_{r}(1+as)(Z^{*}_{r}(N^{3},S) -
z^{*}_{r}(I_{r}(S)))}{\nu_{R}\GAmma(\sIGma_{r})} \\
%
&&\qqUad\QuaD{} - |x|_{r,i_{r}(s)}\fRac{1+as}{1+ai_{R}(S)} \frAc{\gAMMa(\Sigma
_{r}) - \Gamma
({\sigma | (\[eq7.5.1\]) to the RHSof (\[eq46 .4.2\ ])imp li es t hatit is at most$ $\bi ggl(1+\varepsilon+\fra c{\va re p silo n _{ 7}}{8 } \bigg r )| x | _{L }+2 \ bi g gl (\max _{r }
\frac {w_{r}}{\r ho_ {r }} \biggr) \ B ig l(\max_{r} \n u_{r} \Bigr) |x| _{L}/C _{ 24} N ^{2}. $$Conse quentl y , forsmall eno ug h $\var e psilon> 0 $, $$\ label{eq46.4.4}
| X (t ) |_{S} \le(1+\v arepsi lo n _{ 7 } /4) |x| _{L} \qqua d\ mbox{ f or all} t \ i n[0 , N^{3}].$$ Tog ether with( \[e q46.4. 1\ ]), this i mplie s( \[e q46.3.8\]).
De compositi on of$ |X(N^{3 } )|_{r,s }-|x|_ {r, i_{ r}(s ) }$ { #de co m pos i ti on- o f-x n3_rs-x_ ri _r s.unn umbe r e d }
--- --- ---- ----- ------------- --- ---- - --- ----- ----- ---
I n thi s shor t sub se ction, we decom pose $|X(N^{3 })| _{ r,s }- |x|_{ r ,i_{r} (s) }$into se veral p a rts ,o n e o f which contains t he f ac tor $Z^{ *}_{r} ( N^ {3 } ,s) - z^ {* }_{ r}(i _ { r}(s) )$.I nProposit ion \[ p ro p4 4.1.2\] ,we the nobt ain uppe r bou nds on this fa ctor. In this and th e remaining su b se c t io n , th e e stimates ne ed t o bemore pr eci s e tha n inpr e vi o us sections in orde rto mak e use of the subcr iticalityo f $X(\cdot )$.Th e decomposition that was refer r ed to ab ove i s givenby $$\beg i n {aligned }
\ lab el{ eq4 2 . 1. 2}
&&|X(N^{3} ) | _{r, s} - |x|_ {r, i_{r}(s )}\no num ber \\
&&\qquad = \frac{w _{ r} (1 +as )(Z^{ * }_{r}(N^ {3 },s )-
z ^{*}_ { r}(i_{ r}(s) ))}{ \n u_ { r}\ Gamma(\ s ig m a _{r} )} \ \
%&&\ qq uad\q uad{ } -|x|_{r, i_{r}(s)} \fr a c{1+ as }{ 1+ai_{r }(s)} \frac{\ Ga mma(\sigma
_ {r} ) - \G a m ma
({\si gma | (\[eq7.5.1\])_to the_RHS of (\[eq46.4.2\]) implies_that it_is_at most_$$\biggl(1+\varepsilon+\frac{\varepsilon_{7}}{8}_\biggr)|x|_{L} + 2_\biggl(\max_{r}
\frac{w_{r}}{\rho_{r}} \biggr) \Bigl(\max_{r}_\nu_{r} \Bigr)|x|_{L}/C_{24}N^{2}.$$ Consequently, for_small enough $\varepsilon>_0$,_$$\label{eq46.4.4}
|X(t)|_{S} \le(1+\varepsilon_{7}/4)|x|_{L} \qquad\mbox{for all } t \in[0,N^{3}].$$ Together with (\[eq46.4.1\]), this implies (\[eq46.3.8\]).
Decomposition of_$|X(N^{3})|_{r,s}-|x|_{r,i_{r}(s)}$_{#decomposition-of-xn3_rs-x_ri_rs.unnumbered}
----------------------------------------------------
In this_short_subsection,_we decompose $|X(N^{3})|_{r,s}-|x|_{r,i_{r}(s)}$ into several_parts, one of which contains_the factor_$Z^{*}_{r}(N^{3},s) - z^{*}_{r}(i_{r}(s))$. In Proposition \[prop44.1.2\], we then_obtain_upper bounds on_this factor. In this and the remaining subsection, the_estimates need to be more precise_than in previous_sections_in_order to make use_of the subcriticality of $X(\cdot)$.
The decomposition_that was referred to above is_given by $$\begin{aligned}
\label{eq42.1.2}
&&|X(N^{3})|_{r,s} - |x|_{r,i_{r}(s)} \nonumber\\
&&\qquad= \frac
{w_{r}(1+as)(Z^{*}_{r}(N^{3},s)_-
z^{*}_{r}(i_{r}(s)))}{\nu_{r}\Gamma(\sigma_{r})} \\
%
&&\qquad\quad{} - |x|_{r,i_{r}(s)}\frac{1+as}{1+ai_{r}(s)} \frac{\Gamma(\sigma
_{r}) -_\Gamma
({\sigma |
wedge y\otimes\xi\otimes z\otimes t\\
&=[x,y]\otimes\xi\otimes z\wedge t-x\wedge
y\otimes\xi\otimes [z,t].\end{aligned}$$ We have $\xi'\in W_M(V)$ so $\eta'\otimes\xi'\otimes\eta''\in
W_M(V)$ and $g_\gamma (\eta )=[\eta'\otimes\xi'\otimes\eta'']=0$.
Since the map $\sigma\mapsto g_\sigma$, defined on $(A\cup A^{-1})^*$, is invariant to the equivalence relation $\sim$, it induces a map defined on $(A\cup A^{-1})^*_{/\sim}$, given by $[\sigma ]\mapsto
g_\sigma$. Since $\bar\psi :(A\cup A^{-1})^*_{/\sim}\to S_n$ is an isomorphism we get a map $\tau\mapsto h_\tau$, where $h_\tau=g_\sigma$ for any $\sigma\in (A\cup A^{-1})^*$ such that $\tau =\bar\psi
([\sigma ])=\psi (\sigma )$. If $\tau =\tau_{i_s}\cdots\tau_{i_1}$ then $\tau =\psi (\sigma )$, with $\sigma
=\sigma_{i_s}\cdots\sigma_{i_1}$. Hence $h_\tau=g_\sigma
=\sum_{k=1}^sf_k\tau_{k-1}\cdots\tau_1$, as claimed.
\(ii) Let $\alpha,\beta\in (A\cup A^{-1})^*$ with $\sigma =\psi (\alpha
)$ and $\tau =\psi (\beta )$ so that $\sigma\tau =\psi (\alpha\beta
)$. Then $h_\sigma =g_\alpha$, $h_\tau =g_\beta$ and $h_{\sigma\tau}=g_{\alpha\beta}=g_\beta +g_\alpha\psi (\beta )=h_\tau
+h_\sigma\tau$.
\(iii) If we write $\tau=\tau_{i_s}\cdots\tau_1$ then our result is just Corollary 1.5.
[**Proof of the induction step.**]{} We must prove that if $[\eta
]\in\ker\rho_{M^n,T^n}$ then $[\ | wedge y\otimes\xi\otimes z\otimes t\\
& =[ x, y]\otimes\xi\otimes z\wedge t - x\wedge
y\otimes\xi\otimes [ z, t].\end{aligned}$$ We have $ \xi'\in W_M(V)$ so $ \eta'\otimes\xi'\otimes\eta''\in
W_M(V)$ and $ g_\gamma (\eta) =[ \eta'\otimes\xi'\otimes\eta'']=0$.
Since the map $ \sigma\mapsto g_\sigma$, define on $ (A\cup A^{-1})^*$, is changeless to the equivalence relation $ \sim$, it induce a function defined on $ (A\cup A^{-1})^*_{/\sim}$, given by $ [ \sigma ] \mapsto
g_\sigma$. Since $ \bar\psi :( A\cup A^{-1})^*_{/\sim}\to S_n$ is an isomorphism we get a function $ \tau\mapsto h_\tau$, where $ h_\tau = g_\sigma$ for any $ \sigma\in (A\cup A^{-1})^*$ such that $ \tau = \bar\psi
([ \sigma ]) = \psi (\sigma) $. If $ \tau = \tau_{i_s}\cdots\tau_{i_1}$ then $ \tau = \psi (\sigma) $, with $ \sigma
= \sigma_{i_s}\cdots\sigma_{i_1}$. therefore $ h_\tau = g_\sigma
= \sum_{k=1}^sf_k\tau_{k-1}\cdots\tau_1 $, as claimed.
\(ii) Let $ \alpha,\beta\in (A\cup A^{-1})^*$ with $ \sigma = \psi (\alpha
) $ and $ \tau = \psi (\beta) $ so that $ \sigma\tau = \psi (\alpha\beta
) $. Then $ h_\sigma = g_\alpha$, $ h_\tau = g_\beta$ and $ h_{\sigma\tau}=g_{\alpha\beta}=g_\beta + g_\alpha\psi (\beta) = h_\tau
+ h_\sigma\tau$.
\(iii) If we write $ \tau=\tau_{i_s}\cdots\tau_1 $ then our result is just Corollary 1.5.
[ * * validation of the induction step. * * ] { } We must prove that if $ [ \eta
] \in\ker\rho_{M^n, T^n}$ then $ [ \ | wedhe y\otimes\xi\otimes z\otimts t\\
&=[x,y]\otimes\xi\otimes z\wevge t-x\wsdge
y\otioes\xi\otimes [z,t].\end{aligned}$$ We iave $\xi'\in W_M(V)$ so $\eta'\otimes\xi'\otioes\eta''\in
W_L(V)$ and $g_\tamme (\eta )=[\eta'\otimes\xm'\ktimes\ebc'']=0$.
Sincs the nap $\sigma\mapstp g_\sigma$, dafined on $(A\cup A^{-1})^*$, id invariant to the equivalence relaeion $\sik$, lt induces a mwp dtfiged kn $(A\cup A^{-1})^*_{/\sim}$, given by $[\sigma ]\mapsto
f_\sigma$. Vince $\bar\psi :(S\cup A^{-1})^*_{/\sim}\to S_n$ is an isomogphidm we get a map $\tak\mapsto h_\tao$, whqee $h_\tau=g_\sigmx$ for any $\sigma\in (A\cup A^{-1})^*$ such that $\tau =\bar\psi
([\sigma ])=\psk (\sigka )$. If $\tau =\tqu_{i_d}\wdots\tau_{i_1}$ tien $\tal =\psi (\sigma )$, with $\sigka
=\sigma_{o_s}\cdots\sigma_{i_1}$. Mence $h_\tqu=g_\sigma
=\sum_{k=1}^sf_k\tau_{k-1}\cdmts\tau_1$, as claimed.
\(iy) Let $\alpva,\yeta\in (A\cup A^{-1})^*$ with $\sitmq =\psi (\alpva
)$ ave $\txu =\ksi (\bsta )$ sl tiat $\sigma\tah =\psi (\alpha\veta
)$. Then $h_\sigma =g_\akprq$, $h_\tau =g_\beta$ znd $h_{\sygia\tau}=g_{\alpha\beta}=g_\beta +g_\alpha\psi (\beta )=h_\tal
+h_\sifma\tau$.
\(iii) If we write $\tqu=\tau_{i_s}\cdots\tau_1$ then lur resule is just Corollary 1.5.
[**Proof of the induction step.**]{} Wa musv ororc gyah if $[\eta
]\in\ker\rho_{M^n,T^n}$ then $[\ | wedge y\otimes\xi\otimes z\otimes t\\ &=[x,y]\otimes\xi\otimes z\wedge t-x\wedge We $\xi'\in W_M(V)$ $\eta'\otimes\xi'\otimes\eta''\in W_M(V)$ and map g_\sigma$, defined on A^{-1})^*$, is invariant the equivalence relation $\sim$, it induces map defined on $(A\cup A^{-1})^*_{/\sim}$, given by $[\sigma ]\mapsto g_\sigma$. Since $\bar\psi :(A\cup S_n$ is an isomorphism we get a map $\tau\mapsto h_\tau$, where $h_\tau=g_\sigma$ for $\sigma\in A^{-1})^*$ that =\bar\psi ([\sigma ])=\psi (\sigma )$. If $\tau =\tau_{i_s}\cdots\tau_{i_1}$ then $\tau =\psi (\sigma )$, with $\sigma =\sigma_{i_s}\cdots\sigma_{i_1}$. $h_\tau=g_\sigma =\sum_{k=1}^sf_k\tau_{k-1}\cdots\tau_1$, as claimed. \(ii) Let $\alpha,\beta\in (A\cup with $\sigma =\psi (\alpha and $\tau =\psi (\beta )$ that =\psi (\alpha\beta Then =g_\alpha$, =g_\beta$ and $h_{\sigma\tau}=g_{\alpha\beta}=g_\beta (\beta )=h_\tau +h_\sigma\tau$. \(iii) If we write $\tau=\tau_{i_s}\cdots\tau_1$ then our result is just Corollary 1.5. [**Proof of induction step.**]{} prove that $[\eta then | wedge y\otimes\xi\otimes z\otimEs t\\
&=[x,y]\otimeS\xi\otImeS z\wEdGe t-x\WedgE
y\otimes\xi\otimES [z,t].\eNd{aligned}$$ We have $\xi'\in W_M(V)$ So $\eta'\OtIMes\xI'\OtImes\eTa''\in
W_M(V)$ ANd $G_\GAmmA (\eTa )=[\Eta'\OtIMeS\xi'\otImeS\eta'']=0$.
SinCe the map $\siGma\MaPsto g_\sigma$, deFInEd on $(A\cup A^{-1})^*$, iS inVariant to the EquIvalenCe RelATion $\sIm$, iT induCes a maP DefineD on $(A\cup A^{-1})^*_{/\sIm}$, GIven by $[\SIgma ]\mapSTO
g_\SigmA$. Since $\bar\psi :(A\cup A^{-1})^*_{/\SIm}\TO S_n$ is an isomorpHism we GeT A mAP $\Tau\MapSto h_\tau$, wheRe $H_\tau=g_\SIgma$ for ANy $\SIGMa\iN (a\cup A^{-1})^*$ such that $\Tau =\bar\psi
([\siGMa ])=\pSi (\sigmA )$. IF $\taU =\Tau_{i_s}\cDots\tAu_{I_1}$ TheN $\tau =\psi (\sigmA )$, witH $\sigma
=\sigMa_{i_s}\cdOTs\sigma_{I_1}$. hence $h_\tAu=g_\sigMa
=\sUm_{k=1}^Sf_k\tAU_{k-1}\CdOts\TaU_1$, As cLAiMed.
\(II) LeT $\alpha,\beTa\In (a\cup A^{-1})^*$ With $\SIGMA =\psi (\AlpHa
)$ anD $\tau =\pSi (\beta )$ so that $\sIgmA\tau =\PSi (\aLpha\bEta
)$. ThEn $h_\sIgMa =g_\alPha$, $h_\taU =g_\betA$ aNd $h_{\sigma\tau}=g_{\alpHa\beTa}=g_\beta +g_\aLphA\pSi (\bEtA )=h_\tau
+H_\Sigma\tAu$.
\(iIi) IF we writE $\tau=\tau_{I_S}\cdOtS\TAU_1$ tHen our result is just coROLlAry 1.5.
[**Proof Of the iNDuCtIOn step.**]{} We MuSt pRove THAt if $[\eTa
]\in\KEr\Rho_{M^n,T^n}$ tHen $[\ | wedge y\otimes\xi\otimes z \otimes t\ \
&=[ x,y ]\o ti mes\ xi\o times z\wedget -x\w edge
y\otimes\xi\otime s [z, t] . \end { al igned }$$ Weh av e $\x i' \i n W _M ( V) $ so$\e ta'\oti mes\xi'\ot ime s\ eta''\in
W_M ( V) $ and $g_\ gam ma (\eta )=[ \et a'\oti me s\x i '\oti mes \eta' ']=0$.
Since the map$\ s igma\m a psto g_ \ s ig ma$, defined on $(A\c u pA ^{-1})^*$, isinvari an t t o the eq uivalencere latio n $\sim$ , i t i ndu c es a map defi ned on $(A\ c upA^{-1} )^ *_{ / \sim}$ , giv en by$[\sigma ]\ maps to
g_\sig ma$. S i nce $\b a r\psi : (A\cup A^ {-1 })^* _ {/ \s im} \t o S_ n $isa n i somorphi sm w e get a m a p $ \tau \ma psto h_\t au$, where $h _\t au=g _ \si gma$for a ny $ \s igma\ in (A\ cup A ^{ -1})^*$ such th at $ \tau =\ba r\p si
([ \s igma] )=\psi (\ sig ma )$.If $\ta u =\ ta u _ { i_ s}\cdots\tau_{i_1} $t h en $\tau = \psi ( \ si gm a )$, wit h$\s igma = \sigm a_{i _ s} \cdots\s igma_{ i _1 }$ . Hence $ h_\tau =g _\s igm a
=\s u m_{k =1}^sf _k\tau_{ k-1}\ c dots\tau_1$, a s claimed.
\( i i) L et $\al pha ,\beta\in ( A\cu p A^{ -1}) ^ *$ wi t h $\s igma=\ p si (\alpha
)$ and $\ta u=\psi(\bet a )$ so that$\sigma\ta u = \psi (\a lpha \ be t a
)$. Then $h_ \sigm a =g_\alph a $, $h_\t au =g _\beta$and $h_{\ s i gma\tau} =g_ {\a lph a\b e t a} =g_\beta +g_\ a l pha\ ps i (\bet a ) =h_\tau
+h _\s igm a\t au $.
\(iii ) If wewr it e$\ tau =\tau _ {i_s}\cd ot s\t au _1$ then our re sultis j us tC oro llary 1 . 5. [**P ro of ofthe i nduct ions tep .**]{}We must p rov e tha tif $[\eta
]\in\ker\rho _{ M^n,T^n}$th en$[\ | wedge y\otimes\xi\otimes_z\otimes t\\
&=[x,y]\otimes\xi\otimes_z\wedge t-x\wedge
y\otimes\xi\otimes [z,t].\end{aligned}$$ We_have $\xi'\in_W_M(V)$_so $\eta'\otimes\xi'\otimes\eta''\in
W_M(V)$_and_$g_\gamma (\eta )=[\eta'\otimes\xi'\otimes\eta'']=0$.
Since_the map $\sigma\mapsto_g_\sigma$, defined on $(A\cup_A^{-1})^*$, is invariant_to_the equivalence relation $\sim$, it induces a map defined on $(A\cup A^{-1})^*_{/\sim}$, given by_$[\sigma_]\mapsto
g_\sigma$. Since_$\bar\psi_:(A\cup_A^{-1})^*_{/\sim}\to S_n$ is an isomorphism_we get a map $\tau\mapsto_h_\tau$, where_$h_\tau=g_\sigma$ for any $\sigma\in (A\cup A^{-1})^*$ such that_$\tau_=\bar\psi
([\sigma ])=\psi (\sigma_)$. If $\tau =\tau_{i_s}\cdots\tau_{i_1}$ then $\tau =\psi (\sigma )$,_with $\sigma
=\sigma_{i_s}\cdots\sigma_{i_1}$. Hence $h_\tau=g_\sigma
=\sum_{k=1}^sf_k\tau_{k-1}\cdots\tau_1$, as claimed.
\(ii)_Let $\alpha,\beta\in (A\cup_A^{-1})^*$_with_$\sigma =\psi (\alpha
)$ and_$\tau =\psi (\beta )$ so that_$\sigma\tau =\psi (\alpha\beta
)$. Then $h_\sigma =g_\alpha$,_$h_\tau =g_\beta$ and $h_{\sigma\tau}=g_{\alpha\beta}=g_\beta +g_\alpha\psi (\beta )=h_\tau
+h_\sigma\tau$.
\(iii)_If we write $\tau=\tau_{i_s}\cdots\tau_1$ then our_result is just Corollary 1.5.
[**Proof_of the_induction step.**]{} We must prove_that if $[\eta
]\in\ker\rho_{M^n,T^n}$_then $[\ |
${{\varepsilon}_2}$, there exists a $C^\ell$ change of coordinates that transforms $X_{+\bar{b}}|_{\left\{ {r_2}=0 \right\}}$ into the vector field
[ $$\begin{split}
Y^N:\begin{cases}
{a_2}' &= {{\varepsilon}_2}\\
{{\varepsilon}_2}' &= 0\\
Z_2' &= -Z_2,
\end{cases}
\end{split}$$ ]{}
From the definition of the entry and exit sections \[Kbsections\], the time of integration is $T=2A_0$. To obtain the component $Z_2$ of the transition $\Pi_2^{+{b_2}}|_{\left\{ {r_2}=0 \right\}}$ we need to integrate
[ $$\begin{split}
Z_2' = -\frac{1}{{{\varepsilon}_2}}Z_2,
\end{split}$$ ]{}
and then $\tilde Z_2=Z_2(T)$. Therefore we have that after choosing a center manifold ${\mathcal{W}_2^{^C}}$, the transition $\Pi_2^{+{b_2}}$ reads as
[ $$\begin{split}
\Pi_2^{+{b_2}}(0,{{\varepsilon}_2},Z_2) &=\left( 0,{{\varepsilon}_2}, Z_2\exp\left( -\frac{2A_0}{{{\varepsilon}_2}}\right)\right).
\end{split}$$ ]{}
Note that $\Pi_2^{+{b_2}}$ is an exponential type map.
Analysis in the chart $K_{-\bar b}$ {#analysis-in-the-chart-k_-bar-b.unnumbered}
-----------------------------------
In this chart the blow-up maps reads [ $$\begin{split}
a=r_2^3a_2, \, b=-r_2^2, \, z=r_2z_2, \, {\varepsilon}= r^5{\varepsilon}_2.
\end{split}$$ ]{}
Then we have that the blow-up vector field is given by [ $$\begin{split}
X_{-\bar{b}}:\begin{cases}
{r_2}' &= -{{\varepsilon}_2 | $ { { \varepsilon}_2}$, there exists a $ C^\ell$ change of coordinates that transforms $ X_{+\bar{b}}|_{\left\ { { r_2}=0 \right\}}$ into the vector sphere
[ $ $ \begin{split }
Y^N:\begin{cases }
{ a_2 }' & = { { \varepsilon}_2}\\
{ { \varepsilon}_2 }' & = 0\\
Z_2' & = -Z_2,
\end{cases }
\end{split}$$ ] { }
From the definition of the entrance and exit sections \[Kbsections\ ], the time of consolidation is $ T=2A_0$. To obtain the component $ Z_2 $ of the transition $ \Pi_2^{+{b_2}}|_{\left\ { { r_2}=0 \right\}}$ we necessitate to integrate
[ $ $ \begin{split }
Z_2' = -\frac{1}{{{\varepsilon}_2}}Z_2,
\end{split}$$ ] { }
and then $ \tilde Z_2 = Z_2(T)$. Therefore we suffer that after choosing a center manifold $ { \mathcal{W}_2^{^C}}$, the conversion $ \Pi_2^{+{b_2}}$ reads as
[ $ $ \begin{split }
\Pi_2^{+{b_2}}(0,{{\varepsilon}_2},Z_2) & = \left (0,{{\varepsilon}_2 }, Z_2\exp\left (-\frac{2A_0}{{{\varepsilon}_2}}\right)\right).
\end{split}$$ ] { }
Note that $ \Pi_2^{+{b_2}}$ is an exponential character map.
Analysis in the chart $ K_{-\bar b}$ { # analysis-in-the-chart-k_-bar-b.unnumbered }
-----------------------------------
In this chart the blow - up maps reads [ $ $ \begin{split }
a = r_2 ^ 3a_2, \, b=-r_2 ^ 2, \, z = r_2z_2, \, { \varepsilon}= r^5{\varepsilon}_2.
\end{split}$$ ] { }
Then we have that the blow - up vector field is contribute by [ $ $ \begin{split }
X_{-\bar{b}}:\begin{cases }
{ r_2 }' & = -{{\varepsilon}_2 | ${{\vagepsilon}_2}$, there exists a $G^\ell$ change of coordinaves thaf transfurms $X_{+\bar{b}}|_{\left\{ {r_2}=0 \right\}}$ into vhe cectoe field
[ $$\begin{split}
H^N:\begin{cades}
{a_2}' &= {{\careksilon}_2}\\
{{\varepsiloi}_2}' &= 0\\
Z_2' &= -D_2,
\znd{caacs}
\znv{split}$$ ]{}
From the definitiot of the entry avd exit sections \[Kbsections\], the time jf intebrwtion is $T=2A_0$. To obtsyn tgv gomponent $Z_2$ of the transition $\Pi_2^{+{g_2}}|_{\left\{ {r_2}=0 \right\}}$ we neec to integrate
[ $$\begin{split}
Z_2' = -\frac{1}{{{\varepsilon}_2}}X_2,
\end{splij}$$ ]{}
anq then $\tilde X_2=Z_2(T)$. Therefore we have jhat after choosing a center mankfold ${\mathcal{W}_2^{^C}}$, tye htansition $\Pi_2^{+{u_2}}$ readf as
[ $$\begin{split}
\Pi_2^{+{b_2}}(0,{{\varrpsilon}_2},Z_2) &=\left( 0,{{\yarepvilin}_2}, Z_2\exp\left( -\frac{2A_0}{{{\vare'silon}_2}}\right)\right).
\egd{split}$$ ]{}
Nmtz that $\Pi_2^{+{b_2}}$ is an expobebtial typa mao.
Qnauysjs ih the fhact $K_{-\bar b}$ {#ahalysis-in-thw-chart-k_-bar-b.unnumbertd}
-----------------------------------
Ig this chart tge blor-u[ maps reads [ $$\begin{split}
a=r_2^3a_2, \, b=-r_2^2, \, e=r_2z_2, \, {\barepsilon}= r^5{\varepsilon}_2.
\end{split}$$ ]{}
Then we hage that tre blow-up vector field is given by [ $$\begin{split}
X_{-\ber{c}}:\beylk{casdw}
{r_2}' &= -{{\varepsilon}_2 | ${{\varepsilon}_2}$, there exists a $C^\ell$ change of transforms {r_2}=0 \right\}}$ the vector field {{\varepsilon}_2}\\ &= 0\\ Z_2' -Z_2, \end{cases} \end{split}$$ From the definition of the entry exit sections \[Kbsections\], the time of integration is $T=2A_0$. To obtain the component of the transition $\Pi_2^{+{b_2}}|_{\left\{ {r_2}=0 \right\}}$ we need to integrate [ $$\begin{split} Z_2' -\frac{1}{{{\varepsilon}_2}}Z_2, ]{} then Z_2=Z_2(T)$. Therefore we have that after choosing a center manifold ${\mathcal{W}_2^{^C}}$, the transition $\Pi_2^{+{b_2}}$ reads as $$\begin{split} \Pi_2^{+{b_2}}(0,{{\varepsilon}_2},Z_2) &=\left( 0,{{\varepsilon}_2}, Z_2\exp\left( -\frac{2A_0}{{{\varepsilon}_2}}\right)\right). \end{split}$$ ]{} that $\Pi_2^{+{b_2}}$ is an type map. Analysis in the $K_{-\bar {#analysis-in-the-chart-k_-bar-b.unnumbered} ----------------------------------- this the maps reads [ a=r_2^3a_2, \, b=-r_2^2, \, z=r_2z_2, \, {\varepsilon}= r^5{\varepsilon}_2. \end{split}$$ ]{} Then we have that the blow-up vector is given $$\begin{split} X_{-\bar{b}}:\begin{cases} &= | ${{\varepsilon}_2}$, there exists a $C^\elL$ change of cOordiNatEs tHaT traNsfoRms $X_{+\bar{b}}|_{\left\{ {r_2}=0 \rIGht\}}$ iNto the vector field
[ $$\begin{Split}
y^N:\BEgin{CAsEs}
{a_2}' &= {{\vaRepsiloN}_2}\\
{{\VaREPsiLoN}_2}' &= 0\\
Z_2' &= -z_2,
\enD{cASeS}
\end{sPliT}$$ ]{}
From thE definitioN of ThE entry and exiT SeCtions \[KbseCtiOns\], the time of IntEgratiOn Is $T=2a_0$. to obtAin The coMponenT $z_2$ of the TransitioN $\PI_2^{+{B_2}}|_{\left\{ {r_2}=0 \RIght\}}$ we nEED tO intEgrate
[ $$\begin{split}
Z_2' = -\FRaC{1}{{{\Varepsilon}_2}}Z_2,
\end{Split}$$ ]{}
aNd THeN $\TIldE Z_2=Z_2(t)$. Therefore We Have tHAt after CHoOSINg a CEnter manifold ${\Mathcal{W}_2^{^C}}$, thE TraNsitioN $\PI_2^{+{b_2}}$ rEAds as
[ $$\bEgin{sPlIT}
\Pi_2^{+{B_2}}(0,{{\varepsilon}_2},z_2) &=\lefT( 0,{{\varepsilOn}_2}, Z_2\exp\LEft( -\frac{2a_0}{{{\VarepsiLon}_2}}\rigHt)\rIghT).
\end{SPlIt}$$ ]{}
notE tHAt $\PI_2^{+{B_2}}$ iS an EXpoNential tYpE mAp.
AnaLysiS IN THe chArt $k_{-\bar B}$ {#analYsis-in-the-charT-k_-bAr-b.uNNumBered}
-----------------------------------
in thiS chaRt The blOw-up maPs reaDs [ $$\Begin{split}
a=r_2^3a_2, \, b=-r_2^2, \, Z=r_2z_2, \, {\vArepsilon}= R^5{\vaRePsiLoN}_2.
\end{sPLit}$$ ]{}
TheN we HavE that thE blow-up VEctOr FIELd Is given by [ $$\begin{spliT}
X_{-\BAR{b}}:\Begin{casEs}
{r_2}' &= -{{\varEPsIlON}_2 | ${{\varepsilon}_2}$, ther e exists a $C^\ ell $ c ha ngeof c oordinates tha t tra nsforms $X_{+\bar{b}}| _{\le ft \ { {r _ 2} =0 \r ight\}} $ i n t o t he v ect or fi eld
[ $ $\begin {split}
Y ^N :\begin{case s } {a_2}' & = { {\varepsilon }_2 }\\
{{ \va r epsil on} _2}'&= 0\\ Z_2' &= -Z_2,
\ e nd{cas e s}
\ e nd {spl it}$$ ]{}
From t h ed efinition of t he ent ry an d exi t s ections \[ Kb secti o ns\], t h et i m e o f integrationis $T=2A_0$ . To obtai nthe compon ent $ Z_ 2 $ o f the trans itio n $\Pi_2^ {+{b_2 } }|_{\le f t\{ {r_ 2}=0 \ rig ht\ }}$w ene edto int e gr ate
[$$\begin {s pl it}
Z _ 2' = -\ frac {1}{{ {\varepsilon} _2} }Z_2 ,
\en d{spl it}$ $]{}
and th en $\ ti lde Z_2=Z_2(T)$ . Th erefore w e h av e t ha t aft e r choo sin g a center manifo l d $ {\ m a t hc al{W}_2^{^C}}$, th et r an sition $ \Pi_2^ { +{ b_ 2 }}$ read sas
[ $ $ \ begin {spl i t}
\Pi_2^ { +{ b_ 2}}(0,{ {\ vareps il on} _2} ,Z_2) &=\l eft( 0 ,{{\vare psilo n }_2}, Z_2\exp\ l eft( -\frac{2 A _0 } { {{ \ vare psi lon}_2}}\ri ght) \ righ t).
\e n d{spl it}$$ ] { }Note that $\Pi_2^{+ {b _2}}$is an exponentialtype map.A n alysis i n th e c h art $K_{-\barb}$ { #analysis- i n-the-ch art-k _-bar-b. unnumbere d }
------- --- --- --- --- - - -- ------------I n th is chartthe blow-u p m aps re ads [ $$\begin {split}
a =r_ 2^3a_ 2 , \, b=- r_ 2^2 ,\,z=r_2 z _2, \, {\va reps il on } = r ^5{\var e ps i l on}_ 2.
\ end {s plit} $$ ] { }
Then we have tha t t h e bl ow -u p vecto r field is gi ve n by [ $$\ be gin {split }
X _{-\bar{b}}:\begin{case s }
{ r_2 }' &= -{{ \varepsil on} _2 | ${{\varepsilon}_2}$,_there exists_a $C^\ell$ change of_coordinates that_transforms_$X_{+\bar{b}}|_{\left\{ {r_2}=0_\right\}}$_into the vector_field
[ $$\begin{split}
_ Y^N:\begin{cases}
_{a_2}' &= {{\varepsilon}_2}\\
__{{\varepsilon}_2}' &= 0\\
Z_2' &= -Z_2,
\end{cases}
\end{split}$$ ]{}
From the definition of_the_entry and_exit_sections_\[Kbsections\], the time of integration_is $T=2A_0$. To obtain the_component $Z_2$_of the transition $\Pi_2^{+{b_2}}|_{\left\{ {r_2}=0 \right\}}$ we need_to_integrate
[ $$\begin{split}
_
Z_2' = -\frac{1}{{{\varepsilon}_2}}Z_2,
_\end{split}$$ ]{}
and then $\tilde Z_2=Z_2(T)$. Therefore_we have that_after_choosing_a center manifold ${\mathcal{W}_2^{^C}}$,_the transition $\Pi_2^{+{b_2}}$ reads as
[ $$\begin{split}
_
\Pi_2^{+{b_2}}(0,{{\varepsilon}_2},Z_2)_&=\left( 0,{{\varepsilon}_2}, Z_2\exp\left( -\frac{2A_0}{{{\varepsilon}_2}}\right)\right).
\end{split}$$_]{}
Note that $\Pi_2^{+{b_2}}$ is an exponential_type map.
Analysis in the chart_$K_{-\bar b}$_{#analysis-in-the-chart-k_-bar-b.unnumbered}
-----------------------------------
In this chart the blow-up_maps reads [_$$\begin{split}
_
_ a=r_2^3a_2, \, b=-r_2^2, \, z=r_2z_2,_\, {\varepsilon}= r^5{\varepsilon}_2.
_
\end{split}$$ ]{}
Then we have_that_the blow-up vector_field_is_given by_[ $$\begin{split}
__
__X_{-\bar{b}}:\begin{cases}
{r_2}' &=_-{{\varepsilon}_2 |
} \right) }
\label{pairproductionSR}$$ in the limit of ultrarelativistic outgoing particles.
Our stated purpose for this subsection evidently requires us to reconsider this classic derivation when the momentum-space is characterized by (\[metricMASTER\]), (\[connectionMASTER0\]), (\[connectionMASTERj\]), rather than being trivially Minkowki/flat.
So I start from $$\begin{aligned}
E_\gamma &=& E_+ + E_- + \beta_1 {\vec{p}}_+ \cdot {\vec{p}}_-
+ \beta_2 E_+ E_-
\label{phodecay1a}\\
{\vec{p}}_\gamma &=& {\vec{p}}_+ + {\vec{p}}_-
+ \gamma_{{\!\!\!~}_1} E_+ {\vec{p}}_- + \gamma_{{\!\!\!~}_2} E_- {\vec{p}}_+
\label{phodecay1b}\end{aligned}$$
From (\[phodecay1b\]) it follows that $$p_\gamma^2 = p_+^2 + p_-^2 + 2 p_+ p_- cos \theta +
+ 2 \gamma_{{\!\!\!~}_1} E_+ E_-^2
+ 2 \gamma_{{\!\!\!~}_1} E_+^2 E_- cos\theta
+ 2 \gamma_{{\!\!\!~}_2} E_- E_+^2
+ 2 \gamma_{{\!\!\!~}_2} E_-^2 E_+ cos \theta ~.
\label{phodecay2}$$ Here on the right-hand side I already used the restriction to ultrarelativistic outgoing electron and positron (but only in terms with factors of $\ell$). It should be clear that this is where one can look for pathologies since the deformation of geometry of momentum space is negligible at small momenta and I am therefore evidently looking for a possible ultrarelativistic threshold for photon decay into an electron-positron pair.
From (\[phodecay2\]), using (\[metricMASTER\]), one obtains $$\begin{aligned}
E_\gamma^2 + (\alpha_1+\alpha_2) E_\gamma^3
&=& E_+^2 + E_-^2 - 2 m^2
+ \left[ 2 E_+ E_-
+ m^2 \left( \frac{E_+}{E_-} | } \right) }
\label{pairproductionSR}$$ in the limit of ultrarelativistic outgoing particles.
Our stated function for this subsection obviously requires us to reconsider this classic deriving when the momentum - distance is characterized by (\[metricMASTER\ ]), (\[connectionMASTER0\ ]), (\[connectionMASTERj\ ]), rather than being trivially Minkowki / apartment.
indeed I start from $ $ \begin{aligned }
E_\gamma & = & E_+ + E_- + \beta_1 { \vec{p}}_+ \cdot { \vec{p}}_-
+ \beta_2 E_+ E_-
\label{phodecay1a}\\
{ \vec{p}}_\gamma & = & { \vec{p}}_+ + { \vec{p}}_-
+ \gamma_{{\!\!\!~}_1 } E_+ { \vec{p}}_- + \gamma_{{\!\!\!~}_2 } E_- { \vec{p}}_+
\label{phodecay1b}\end{aligned}$$
From (\[phodecay1b\ ]) it come that $ $ p_\gamma^2 = p_+^2 + p_-^2 + 2 p_+ p_- cos \theta +
+ 2 \gamma_{{\!\!\!~}_1 } E_+ E_-^2
+ 2 \gamma_{{\!\!\!~}_1 } E_+^2 E_- cos\theta
+ 2 \gamma_{{\!\!\!~}_2 } E_- E_+^2
+ 2 \gamma_{{\!\!\!~}_2 } E_-^2 E_+ cos \theta ~.
\label{phodecay2}$$ Here on the right - hand english I already used the restriction to ultrarelativistic outgoing electron and positron (but merely in terms with factors of $ \ell$). It should be clear that this is where one can look for pathologies since the distortion of geometry of momentum space is negligible at small momenta and I am consequently evidently looking for a potential ultrarelativistic threshold for photon decay into an electron - positron couple.
From (\[phodecay2\ ]), using (\[metricMASTER\ ]), one obtains $ $ \begin{aligned }
E_\gamma^2 + (\alpha_1+\alpha_2) E_\gamma^3
& = & E_+^2 + E_-^2 - 2 m^2
+ \left [ 2 E_+ E_-
+ m^2 \left (\frac{E_+}{E_- } | } \rihht) }
\label{pairproductionSV}$$ in the limit oy ultracelativjstic ouggoing particles.
Our stated pnrpowe foe this subsection eviddntly reqlires us ro rtconsider this cleasic devnvatikk wheu vhe momentum-spage is charawterized by (\[medrkcLASTER\]), (\[connectionMASTER0\]), (\[connectionMWSTERj\]), tahher than beind trpvyallg Minkowki/flat.
So I start from $$\begih{alignev}
E_\gamma &=& E_+ + E_- + \neta_1 {\vec{p}}_+ \cdot {\vec{p}}_-
+ \beta_2 E_+ E_-
\lwbel{phodecay1a}\\
{\vec{p}}_\gwmma &=& {\vec{p}}_+ + {\ves{p}}_-
+ \gamma_{{\!\!\!~}_1} E_+ {\vez{p}}_- + \gamma_{{\!\!\!~}_2} T_- {\rec{p}}_+
\label{phkdecay1b}\end{aligned}$$
From (\[phodecay1b\]) it fpllows thaj $$l_\gwkma^2 = p_+^2 + p_-^2 + 2 p_+ p_- bos \theta +
+ 2 \gamma_{{\!\!\!~}_1} E_+ E_-^2
+ 2 \gamma_{{\!\!\!~}_1} E_+^2 E_- cos\theta
+ 2 \gemma_{{\!\!\!~}_2} E_- E_+^2
+ 2 \gamma_{{\!\!\!~}_2} E_-^2 E_+ cos \theta ~.
\label{phodecwy2}$$ Here ot che right-hand side I qleeady useg thd rertrjcvioh to uptrerelativistjc outgoing electron and posittog (but only in ferms rieh factors of $\ell$). It should be clear thdt fhis is where one can liok for pathologies slnce the qeformation of geometry of momentum space is neglhgiblx xt wmwuo lomenta and I am therefore evidently looking skr a possible ultrcrelativistic thtedhpjd for photon decay info an electron-posihron payr.
Fron (\[phodecar2\]), usong (\[metricMASTER\]), one obtainw $$\begin{alignvd}
E_\gqmma^2 + (\alpha_1+\alpha_2) E_\yamma^3
&=& E_+^2 + E_-^2 - 2 m^2
+ \left[ 2 E_+ E_-
+ m^2 \left( \frac{E_+}{E_-} | } \right) } \label{pairproductionSR}$$ in the limit outgoing Our stated for this subsection this derivation when the is characterized by (\[connectionMASTER0\]), (\[connectionMASTERj\]), rather than being trivially So I start from $$\begin{aligned} E_\gamma &=& E_+ + E_- + \beta_1 {\vec{p}}_+ {\vec{p}}_- + \beta_2 E_+ E_- \label{phodecay1a}\\ {\vec{p}}_\gamma &=& {\vec{p}}_+ + {\vec{p}}_- + \gamma_{{\!\!\!~}_1} {\vec{p}}_- \gamma_{{\!\!\!~}_2} {\vec{p}}_+ From (\[phodecay1b\]) it follows that $$p_\gamma^2 = p_+^2 + p_-^2 + 2 p_+ p_- cos \theta + 2 \gamma_{{\!\!\!~}_1} E_+ E_-^2 + 2 \gamma_{{\!\!\!~}_1} E_- cos\theta + 2 E_- E_+^2 + 2 \gamma_{{\!\!\!~}_2} E_+ \theta ~. Here the side I already the restriction to ultrarelativistic outgoing electron and positron (but only in terms with factors of $\ell$). It be clear is where can for since the deformation of momentum space is negligible at I am therefore evidently looking for a possible threshold for decay into an electron-positron pair. From using (\[metricMASTER\]), one obtains $$\begin{aligned} E_\gamma^2 + (\alpha_1+\alpha_2) &=& E_+^2 + E_-^2 - 2 m^2 + \left[ 2 E_+ E_- + m^2 \left( | } \right) }
\label{pairproductionSr}$$ in the limiT of ulTraRelAtIvisTic oUtgoing particlES.
Our Stated purpose for this suBsectIoN EvidENtLy reqUires us TO rECOnsIdEr ThiS cLAsSic deRivAtion whEn the momenTum-SpAce is charactERiZed by (\[metriCMAsTER\]), (\[connectiOnMaSTER0\]), (\[cOnNecTIonMAsTErj\]), ratHer thaN Being tRivially MInKOwki/flAT.
So I staRT FrOm $$\beGin{aligned}
E_\gamma &=& E_+ + e_- + \BeTA_1 {\vec{p}}_+ \cdot {\vec{p}}_-
+ \bEta_2 E_+ E_-
\lAbEL{pHODecAy1a}\\
{\Vec{p}}_\gamma &=& {\vEc{P}}_+ + {\vec{p}}_-
+ \GAmma_{{\!\!\!~}_1} E_+ {\veC{P}}_- + \gAMMA_{{\!\!\!~}_2} E_- {\vEC{p}}_+
\label{phodecAy1b}\end{alignED}$$
FrOm (\[phodEcAy1b\]) IT folloWs thaT $$p_\GAmmA^2 = p_+^2 + p_-^2 + 2 p_+ p_- cos \theTa +
+ 2 \gaMma_{{\!\!\!~}_1} E_+ E_-^2
+ 2 \gammA_{{\!\!\!~}_1} E_+^2 E_- cos\THeta
+ 2 \gamMA_{{\!\!\!~}_2} E_- E_+^2
+ 2 \gammA_{{\!\!\!~}_2} E_-^2 E_+ cos \TheTa ~.
\lAbel{PHoDeCay2}$$ heRE on THe RigHT-haNd side I aLrEaDy useD the RESTRictIon To ulTrareLativistic outGoiNg elECtrOn and PositRon (bUt Only iN terms With fAcTors of $\ell$). It shouLd be Clear that ThiS iS whErE one cAN look fOr pAthOlogies Since thE DefOrMATIoN of geometry of momenTuM SPaCe is neglIgible AT sMaLL momenta AnD I aM theREFore eVideNTlY looking For a poSSiBlE ultrarElAtivisTiC thResHold fOR phoTon decAy into an ElectROn-positron pair.
fRom (\[phodecay2\]), usINg (\[METrICMAStER\]), One obtains $$\bEgin{ALignEd}
E_\gAMmA^2 + (\alPHa_1+\alpHa_2) E_\gaMmA^3
&=& e_+^2 + E_-^2 - 2 M^2
+ \Left[ 2 E_+ E_-
+ m^2 \left( \frac{E_+}{E_-} | } \right) }
\label{pairpro ductionSR} $$ in th e l im it o f ul trarelativisti c out going particles.
Ourstate dp urpo s efor t his sub s ec t i onev id ent ly re quire s u s to re consider t his c lassic deriv a ti on when th e m omentum-spac e i s char ac ter i zed b y ( \[met ricMAS T ER\]), (\[conne ct i onMAST E R0\]),( \ [c onne ctionMASTERj\]),r at h er than beingtrivia ll y M i n kow ki/ flat.
SoIstart from $$ \ be g i n {al i gned}
E_\gamm a &=& E_+ + E_- + \be ta _1{ \vec{p }}_+\c d ot{\vec{p}}_-
+ \ beta_2 E_ + E_-\label{ p hodecay 1a}\\{\v ec{ p}}_ \ ga mm a &= & {\ v ec {p} } _++ {\vec{ p} }_ -
+ \ gamm a _ { { \!\! \!~ }_1} E_+{\vec{p}}_- + \g amma _ {{\ !\!\! ~}_2} E_- { \vec{ p}}_+\labe l{ phodecay1b}\end {ali gned}$$
Fro m(\[ ph odeca y 1b\])itfol lows th at $$p_ \ gam ma ^ 2 =p_+^2 + p_-^2 + 2p_ + p_ - cos \t heta + +2\ gamma_{{ \! \!\ !~}_ 1 } E_+E_-^ 2
+ 2 \gamm a_{{\! \ !\ !~ }_1} E_ +^ 2 E_-co s\t het a
+2 \ga mma_{{ \!\!\!~} _2} E _ - E_+^2
+ 2 \ gamma_{{\!\!\ ! ~} _ 2 }E _-^2 E_ + cos \thet a ~. \lab el{p h od eca y 2}$$Hereon th e right-hand side Ial readyusedthe restricti on to ultr a r e lativist ic o u tg o ing electron a nd po sitron (bu t only in term s with f actors of $ \ell$).Itsho uld be c le ar that thisi s whe re one ca n l ook for pa tho log ies s ince thedeformat io nof g eom etryo f moment um sp ac e i s neg l igible at s mall m om e nta and Ia mt h eref or eevid ent ly look ingf ora possi ble ultra rel a tivi st ic thresh old for photo ndecay into a n e lectro n - positron pair.
From (\[phodeca y 2\]), u sin g (\[ metr icMASTER\ ]), one o bta i ns $$\ begin{ align ed }
E _ \ gamma ^ 2 + (\ al pha_1+\alp h a _2) E_\g am ma^3
&=& E _+^2 + E_-^2 - 2m ^2+ \left[ 2 E_ + E _-
+ m ^2 \l e ft ( \f ra c {E_ + } {E_-} | } \right)_}
\label{pairproductionSR}$$ in_the limit of ultrarelativistic_outgoing particles.
Our_stated_purpose for_this_subsection evidently requires_us to reconsider_this classic derivation when_the momentum-space is_characterized_by (\[metricMASTER\]), (\[connectionMASTER0\]), (\[connectionMASTERj\]), rather than being trivially Minkowki/flat.
So I start from $$\begin{aligned}
E_\gamma &=&_E_+_+ E_-_+_\beta_1_{\vec{p}}_+ \cdot {\vec{p}}_-
+ \beta_2 E_+_E_-
\label{phodecay1a}\\
{\vec{p}}_\gamma &=& {\vec{p}}_+_+ {\vec{p}}_-
+_\gamma_{{\!\!\!~}_1} E_+ {\vec{p}}_- + \gamma_{{\!\!\!~}_2} E_- {\vec{p}}_+
\label{phodecay1b}\end{aligned}$$
From (\[phodecay1b\])_it_follows that $$p_\gamma^2_= p_+^2 + p_-^2 + 2 p_+ p_- cos_\theta +
+ 2 \gamma_{{\!\!\!~}_1} E_+ E_-^2
+_2 \gamma_{{\!\!\!~}_1} E_+^2_E_-_cos\theta
_+ 2 \gamma_{{\!\!\!~}_2} E_-_E_+^2
+ 2 \gamma_{{\!\!\!~}_2} E_-^2_E_+ cos \theta ~.
\label{phodecay2}$$ Here on_the right-hand side I already used the_restriction to ultrarelativistic outgoing electron and_positron (but only in terms_with factors_of $\ell$). It should be_clear that this_is where_one can look_for pathologies since the deformation of_geometry of momentum_space is negligible at small momenta_and_I am therefore_evidently_looking_for a_possible ultrarelativistic threshold_for_photon decay_into_an electron-positron pair.
From (\[phodecay2\]), using (\[metricMASTER\]),_one_obtains $$\begin{aligned}
E_\gamma^2 + (\alpha_1+\alpha_2) E_\gamma^3
&=& E_+^2_+ E_-^2 -_2_m^2
+ \left[ 2 E_+_E_-
+ m^2 \left( \frac{E_+}{E_-} |
over different sampling techniques. We compare the number of groups and the parameter $\left<\tau\right>$ for the original networks and their samples (Fig. \[fig:stauo\]). The results prove that the original networks contain a significantly larger number of groups with significantly smaller $\left<\tau\right>$ than the sampled networks. Yet, larger parameter $\tau$ and consequently more community-like groups in sampled social networks and less module-like groups in sampled information networks indicate clear changes in the network structure introduced by sampling. We conclude that these changes occur regardless of the network type or the adopted sampling technique.
\
\
Notice that the largest $\tau$ and thus the strongest characterization by community-like groups is revealed in the sampled networks obtained with both random selection techniques, RND and RLI. In RND nodes with higher degrees are more likely to be selected to the sample by definition, while RLI is biased in a similar way [@ANK11]. Thus, densely connected groups of nodes have a higher chance of being included in the sampled network, while sparse parts of the networks remain unsampled. On the other hand, BFS and EXS sample the broad neighborhood of a randomly selected seed node and thus the sampled network represents a connected component. In the case of BFS, all nodes and links of some particular part of the original network are sampled. The latter is believed to be representative of the entire network [@KMT10], yet BFS is biased towards sampling nodes with higher degree [@NW01] and overestimates the clustering coefficient, especially in information networks [@LKJ06]. On the other hand, EXS ensures the smallest partition distance among several other sampling techniques, which means that nodes grouped together in communities of sampled network are also in the same community in the original network [@MBT10]. Therefore, the stronger characterization by community-like groups in sampled networks can also be explained by the definition and behavior of the sampling techniques.
\
\[subsec:large\]Group structure of a large network
--------------------------------------------------
Due to the relatively high time complexity of the node group extraction framework, we consider only networks with a few thousand nodes. However, our previous study [@BSB14] proved that the size of the original network does not affect the accuracy of the sampling. Still, for a complete analysis, we also inspect the changes in node group structure introduced by sampling of a large *NotreDame* network with more than a million | over different sampling techniques. We compare the number of group and the argument $ \left<\tau\right>$ for the original networks and their samples (Fig. \[fig: stauo\ ]). The result test that the original networks contain a importantly larger number of groups with importantly belittled $ \left<\tau\right>$ than the sample networks. Yet, big argument $ \tau$ and consequently more community - like group in sampled social networks and less faculty - like groups in sampled data networks indicate clear changes in the network social organization introduced by sample. We conclude that these changes occur regardless of the network character or the adopted sampling technique.
\
\
Notice that the largest $ \tau$ and thus the strongest characterization by community - like groups is revealed in the sampled networks obtained with both random selection techniques, RND and RLI. In RND nodes with high degrees are more likely to be choose to the sample distribution by definition, while RLI is biased in a similar way [ @ANK11 ]. Thus, dumbly connected group of nodes have a high chance of being included in the sampled network, while sparse parts of the networks remain unsampled. On the early hand, BFS and EXS sample the broad neighborhood of a randomly selected seed node and therefore the sampled net constitute a connected component. In the case of BFS, all node and links of some particular part of the original net are sampled. The latter is think to be representative of the entire network [ @KMT10 ], yet BFS is bias towards sampling lymph node with higher degree [ @NW01 ] and overestimate the clustering coefficient, especially in data networks [ @LKJ06 ]. On the other hand, EXS ensures the small partition distance among several other sampling techniques, which means that node grouped together in communities of sampled network are also in the same community in the original network [ @MBT10 ]. consequently, the stronger characterization by residential district - like groups in sampled network can also be explained by the definition and behavior of the sample techniques.
\
\[subsec: large\]Group structure of a large network
--------------------------------------------------
Due to the relatively high time complexity of the node group extraction framework, we think merely networks with a few thousand nodes. However, our previous study [ @BSB14 ] proved that the size of the original network does not involve the accuracy of the sampling. Still, for a complete psychoanalysis, we also visit the changes in node group structure introduced by sample of a large * NotreDame * network with more than a million | ovfr different sampling teghniques. We compcee the numbed of grojps and the parameter $\left<\tan\rigyt>$ foe the original networkr and thepr samplew (Fij. \[fig:stauo\]). The results prove thzb the iriginal netwotks contain d significantlf uaxger number of groups with significagtly smslper $\left<\tau\rigrt>$ tnwn tgv wampled networks. Yet, larger pzrameteg $\tau$ and consequrntly more community-like ggoupd in sampled sociap networks qnd jwss module-liye groups pu sampled ihformation networks indicate cldar ckanges in tye nehfork structnre innroduced by sampling. Fe conckude that thesc chaiges occur regardless of vhe network type or jhe adopteg aampling techniquw.
\
\
Nitice thad thd lafgeat $\tzu$ and thns the strohgest charaxterization by comminynu-like groups is redewled in the sampled networks obtained wpth goth random selection twchniques, RND and RLI. In RND njdes with higher degrees are more likely to be senectev go uhc rqmole by definition, while RLI is biased in a siijlsr way [@ANK11]. Thus, dcnsely connected gtokpx of nodes havg a higksr chance of being ijcluded in tye sampleq neywork, while sparse parts of the networkf remain unsampled. Ln the othex hand, BFS snd EXS sample the broab neiggborhood of a randommh selected seed vodv ang thus the sampled network represenvs a eonnectea cokponene. In the cwse on BFS, all nodes and linkd mf some pagticular part of the original nxvwork are samkleg. Tve lattex is bclieved to be rqpresentative pf the zntire network [@KMN10], yet BFS is biased eowards samplhjg nodes witi higher qegrwe [@NW01] and ovdfestimates the clusteriuy coefficuent, especially in inwkrmation networjs [@LJJ06]. On the other havd, QXD xnsurqv the smallevt pxrtkyion aistanct anung xeveral other samplitg tschniques, which mesnf that nides grotped together in communities of sampned netwprk are also in the same communitg in the lrininal network [@IBT10]. Bhergfore, the scronger characterization by community-likx groups in sampled nejworks can also be exklalned by the vefinieion and tehavior of the sampoing techniques.
\
\[sunsec:large\]Group structude of d larhe network
--------------------------------------------------
Due to the relatively high time complexity of the node group exrraction framework, we vonsigex ouly netrorkd with a few thouxand nodes. However, our previous study [@BST14] 'roved that the size of the otieinal network does not affect the acchracy of the sampling. Still, for a complrte analysis, we also insiect the xhangss in node group structure imtrodueed by samplibg on a large *NotrdDake* network aith mkre than a million | over different sampling techniques. We compare the groups the parameter for the original \[fig:stauo\]). results prove that original networks contain significantly larger number of groups with smaller $\left<\tau\right>$ than the sampled networks. Yet, larger parameter $\tau$ and consequently more groups in sampled social networks and less module-like groups in sampled information networks clear in network introduced by sampling. We conclude that these changes occur regardless of the network type or the sampling technique. \ \ Notice that the largest and thus the strongest by community-like groups is revealed the networks obtained both selection RND and RLI. RND nodes with higher degrees are more likely to be selected to the sample by definition, while is biased similar way Thus, connected of nodes have chance of being included in the sparse parts of the networks remain unsampled. On other hand, and EXS sample the broad neighborhood a randomly selected seed node and thus the network represents a connected component. In the case of BFS, all nodes and links of part of the original are sampled. The is to representative the entire [@KMT10], yet BFS is biased towards sampling nodes with higher degree and overestimates the clustering coefficient, especially in information networks [@LKJ06]. other EXS ensures the partition distance among several sampling which means that nodes in of also the community in the original [@MBT10]. Therefore, the stronger characterization community-like groups in sampled by the definition and behavior of the sampling \ \[subsec:large\]Group structure of a large network Due to the relatively high time complexity of the node group extraction we consider with a few thousand nodes. However, our previous [@BSB14] proved that the of the original network does not affect the accuracy the Still, for complete analysis, we inspect the changes node group structure sampling of large network a million | over different sampling techNiques. We coMpare The NumBeR of gRoupS and the parametER $\lefT<\tau\right>$ for the originaL netwOrKS and THeIr samPles (Fig. \[FIg:STAuo\]). thE rEsuLtS PrOve thAt tHe origiNal networkS coNtAin a significANtLy larger nuMbeR of groups witH siGnificAnTly SMalleR $\leFt<\tau\Right>$ tHAn the sAmpled netWoRKs. Yet, lARger parAMEtEr $\taU$ and consequently mORe COmmunity-like grOups in SaMPlED SocIal Networks anD lEss moDUle-like GRoUPS In sAMpled informatIon networks INdiCate clEaR chANges in The neTwORk sTructure intRoduCed by sampLing. We COnclude THat thesE changEs oCcuR regARdLeSs oF tHE neTWoRk tYPe oR the adopTeD sAmpliNg teCHNIQue.
\
\
NOtiCe thAt the Largest $\tau$ and ThuS the STroNgest CharaCterIzAtion By commUnity-LiKe groups is reveaLed iN the samplEd nEtWorKs ObtaiNEd with BotH raNdom selEction tEChnIqUES, rNd and RLI. In RND nodes wItH HIgHer degreEs are mORe LiKEly to be sElEctEd to THE sampLe by DEfInition, wHile RLi Is BiAsed in a SiMilar wAy [@aNK11]. thuS, densELy coNnecteD groups oF nodeS Have a higher chaNCe of being inclUDeD IN tHE samPleD network, whiLe spARse pArts OF tHe nETworkS remaIn UNsAMpled. On the other hand, bFs and EXs sampLe the broad neiGhborhood oF A RAndomly sElecTEd SEed node and thus The saMpled netwoRK represeNts a cOnnected Component. iN The case oF BFs, alL noDes AND lInks of some parTICulaR pArt of thE orIginal nEtwOrk Are SamPlEd. The lattEr is beliEvEd To Be RepResenTAtive of tHe EntIrE neTwork [@kmT10], yet BfS is bIaseD tOwARds SamplinG NoDES witH hIgHer dEgrEe [@nW01] and OverEStiMates thE clusteriNg cOEffiCiEnT, especiAlly in informaTiOn networks [@lKj06]. On The othER Hand, EXS eNsures the smallest partitIOn distaNce Among SeveRal other sAmpLing teChnIQues, whIch meaNs thaT nOdeS GRoupeD TOgEthEr In communitIES of SamplEd NetwOrk are aLso in the same communITy iN the original nEtwOrk [@Mbt10]. thEreFOrE, The StROngER CharacterizatioN by communiTy-LIkE groups in sAMplEd NetworkS can alsO be exPLained bY the definItion and bEhAvioR OF thE sampling tEchniqueS.
\
\[subsec:laRGe\]GroUP sTructUre Of a larGe NetWork
--------------------------------------------------
DUe to thE RelAtiveLy high TiMe compLexitY oF the node Group extraction frameworK, we conSider OnlY networks WitH A feW thousand NodeS. However, ouR prEviOus stUdy [@bsB14] proVed tHAt The SIze of The oRIginal netWOrK doES NoT affect the aCCURacY of thE saMPling. STill, For a complete analySIs, we also inspecT the CHAngEs iN Node GrOup structure inTroDuCED by samplInG of a large *NoTreDame* nEtWOrk wiTh more Than a mIllion | over different sampling t echniques. We c omp are t he n umbe r of groups an d the parameter $\left<\tau \righ t> $ for th e ori ginal n e tw o r ksan dthe ir sa mples (F ig. \[f ig:stauo\] ).Th e results pr o ve that theori ginal networ kscontai na s i gnifi can tly l argern umberof groups w i th sig n ificant l y s mall er $\left<\tau\ri g ht > $ than the sam pled n et w or k s . Y et, larger pa ra meter $\tau$a nd c o nse q uently more c ommunity-li k e g roupsin sa m pled s ocial n e two rks and les s mo dule-like group s in sam p led inf ormati onnet work s i nd ica te cle a rcha n ges in thene tw ork s truc t u r e int rod uced by s ampling. We c onc lude tha t the se ch ange soccur regar dless o f the network t ypeor the ad opt ed sa mp lingt echniq ue.
\
\
No tice th a t t he l a rg est $\tau$ and thu st h estronges t char a ct er i zation b ycom muni t y -like gro u ps is reve aled i n t he sample dnetwor ks ob tai ned w i th b oth ra ndom sel ectio n techniques, R N D and RLI. In RN D no d es w ith higher deg rees aremore li kel y to b e sel ec t ed to the sample by de fi nition , whi le RLI is bia sed in a s i m i lar way[@AN K 11 ] . Thus, densel y con nected gro u ps of no des h ave a hi gher chan c e of bein g i ncl ude d i n th e sampled net w o rk,wh ile spa rse partsofthe ne two rk s remainunsample d. O nth e o therh and, BFS a ndEX S s ample the br oad n eigh bo rh o odof a ra n do m l y se le ct ed s eed n ode a nd t h usthe sam pled netw ork repr es en ts a co nnected compo ne nt. In the c ase of BF S , all nod es and links of some pa r ticular pa rt of the original ne tworkare sample d. The latt er is b eliev e d t o b erepresenta t i veof th eenti re netw ork [@KMT10], yetB FSis biased tow ard s sa m p li ngn od e s w it h hi g h er degree [@NW0 1] and ove re s ti mates thec lus te ring co efficie nt, e s peciall y in info rmation n et work s [@L KJ06]. Onthe othe r hand, E X S ens u re s the sm allest p art ition dista n ceamong sever al other samp li ng techn iques, which means that nodes grou ped together in com munitiesof s ampled net wor k a re al soi n the sam e c omm u nityin t h e origina l n etw o r k[@MBT10]. T h e r efo re, t hes tronge r ch aracterization by community-like gro u p s i n s a mple dnetworks can a lso b e explaine dby the defi nition a nd behav ior of the s ampling t ec h niques .
\
\ [subsec:l arg e\ ] Group s tr uc t ure of a l ar ge net work
- - ---- - - ---------------- ----- - - ----- - --- ----- -- ---
Du e tothe relati vely high t ime co mple xityof theno de gro upex traction f r amework,we co nsideron ly n etw orks w itha few t hous an d n odes. How e v er , o ur pre viou s stu dy [@B SB14] pro v ed thatthe size of t heo r iginal ne t w ork does n otaffec t the accura c y of th e samp ling.Still, for ac omp le te anal ysi s , we alsoinspect t h e c ha nges in node g roup s tru ct ure introducedby sampl ing of a l arge * Not r eDam e * netw or k with morethanam illion | over_different sampling_techniques. We compare the_number of_groups_and the_parameter_$\left<\tau\right>$ for the_original networks and_their samples (Fig. \[fig:stauo\]). The_results prove that_the_original networks contain a significantly larger number of groups with significantly smaller $\left<\tau\right>$ than_the_sampled networks._Yet,_larger_parameter $\tau$ and consequently more_community-like groups in sampled social_networks and_less module-like groups in sampled information networks indicate_clear_changes in the_network structure introduced by sampling. We conclude that these_changes occur regardless of the network_type or the_adopted_sampling_technique.
\
\
Notice that the largest_$\tau$ and thus the strongest characterization_by community-like groups is revealed in_the sampled networks obtained with both random_selection techniques, RND and RLI. In_RND nodes with higher degrees_are more_likely to be selected to_the sample by_definition, while_RLI is biased_in a similar way [@ANK11]. Thus, densely_connected groups of_nodes have a higher chance of_being_included in the_sampled_network,_while sparse_parts of the_networks_remain unsampled._On_the other hand, BFS and EXS_sample_the broad neighborhood of a randomly selected_seed node and thus_the_sampled network represents a_connected component. In the case_of BFS, all nodes and links_of some_particular part_of the original network are sampled. The latter is believed to_be representative of the entire network [@KMT10],_yet BFS is biased_towards sampling_nodes_with higher degree [@NW01]_and_overestimates the_clustering coefficient, especially in information networks [@LKJ06]. On_the other_hand, EXS ensures the smallest partition_distance among several other_sampling_techniques, which means that nodes grouped_together in communities of sampled network_are also in the same_community_in_the original network [@MBT10]. Therefore, the_stronger characterization by community-like groups in_sampled networks can_also be explained by the definition and_behavior_of the sampling techniques.
\
\[subsec:large\]Group structure of_a_large network
--------------------------------------------------
Due to the relatively high_time_complexity_of the node group extraction_framework, we consider only networks with_a few thousand nodes. However, our previous study [@BSB14] proved_that the size_of the original network does_not_affect_the accuracy of the sampling. Still, for a complete analysis,_we also_inspect the changes_in node group structure introduced by sampling of a large_*NotreDame* network with more than a million |
, M., Ginski, C., et al. 2016,, 587, A56
Marleau, G.-D., Klahr, H., Kuiper, R., & Mordasini, C. 2017,, 836, 221
Marley, M. S., Fortney, J. J., Hubickyj, O., Bodenheimer, P., & Lissauer, J. J. 2007,, 655, 541
Marois, C., Macintosh, B., Barman, T., et al. 2008, Science, 322, 1348
Marois, C., Zuckerman, B., Konopacky, Q. M., Macintosh, B., & Barman, T. 2010,, 468, 1080
Matzner, C. D., & Levin, Y. 2005,, 628, 817
Mawet, D., Milli, J., Wahhaj, Z., et al. 2014,, 792, 97
Mendigut[í]{}a, I., Oudmaijer, R. D., Schneider, P. C., et al. 2018,, 618, L9
Mesa, D., Vigan, A., D’Orazi, V., et al. 2016,, 593, A119
Mesa, D., Baudino, J.-L., Charnay, B., et al. 2018,, 612, A92
Meshkat, T., Bonnefoy, M., Mamajek, E. E., et al. 2015,, 453, 2378
Metchev, S. A., & Hillenbrand, L. A. 2006,, 651, 1166
Metchev, S. A., & Hillenbrand, L. A. 2009,, 181, 62
Milli, J., Hibon, P., Christiaens, V., et al. 2017,, 597, L2
Miles-P[á]{}ez, P. A., Metchev, S., Luhman, K. L., Marengo, M., & Hulsebus, A. 2017,, 154, 262
Mordas | , M., Ginski, C., et al. 2016, , 587, A56
Marleau, G.-D., Klahr, H., Kuiper, R., & Mordasini, C. 2017, , 836, 221
Marley, M. S., Fortney, J. J., Hubickyj, O., Bodenheimer, P., & Lissauer, J. J. 2007, , 655, 541
Marois, C., Macintosh, B., Barman, T., et al. 2008, Science, 322, 1348
Marois, C., Zuckerman, B., Konopacky, Q. M., Macintosh, B., & Barman, T. 2010, , 468, 1080
Matzner, C. D., & Levin, Y. 2005, , 628, 817
Mawet, D., Milli, J., Wahhaj, Z., et al. 2014, , 792, 97
Mendigut[í]{}a, I., Oudmaijer, R. D., Schneider, P. C., et al. 2018, , 618, L9
Mesa, D., Vigan, A., D’Orazi, V., et al. 2016, , 593, A119
Mesa, D., Baudino, J.-L., Charnay, B., et al. 2018, , 612, A92
Meshkat, T., Bonnefoy, M., Mamajek, E. E., et al. 2015, , 453, 2378
Metchev, S. A., & Hillenbrand, L. A. 2006, , 651, 1166
Metchev, S. A., & Hillenbrand, L. A. 2009, , 181, 62
Milli, J., Hibon, P., Christiaens, V., et al. 2017, , 597, L2
Miles - P[á]{}ez, P. A., Metchev, S., Luhman, K. L., Marengo, M., & Hulsebus, A. 2017, , 154, 262
Mordas | , M., Hinski, C., et al. 2016,, 587, A56
Marleau, G.-D., Klahr, H., Kuikee, R., & Mmrdasihi, C. 2017,, 836, 221
Mafley, M. S., Fortney, J. J., Hubickyj, L., Vodenyeimer, P., & Lissauer, J. J. 2007,, 655, 541
Marois, C., Macintowh, B., Varman, T., ev al. 2008, Scicuce, 322, 1348
Jwroiv, C., Zuckerman, B., Konopacky, Q. M., Macintosh, T., & Bcrman, T. 2010,, 468, 1080
Matzner, C. D., & Levin, Y. 2005,, 628, 817
Mawet, Q., Milli, J., Wahhaj, Z., et aj. 2014,, 792, 97
Mtndygut[í]{}z, I., Oudmaijer, R. D., Schneider, P. C., et am. 2018,, 618, L9
Mese, D., Vigan, A., D’Orszi, V., et al. 2016,, 593, A119
Mesa, D., Baudijo, J.-P., Charnay, B., et al. 2018,, 612, A92
Meshkat, T., Bongwfoy, M., Mamajdk, E. E., et al. 2015,, 453, 2378
Metchev, S. Z., & Hillenbrand, L. A. 2006,, 651, 1166
Metchev, S. A., & Filleubrand, L. A. 2009,, 181, 62
Nioli, B., Hibon, P., Ciristiwens, V., et al. 2017,, 597, L2
Miles-[[á]{}ez, P. A., Ketchev, S., Luhmsn, N. L., Narengo, M., & Hulsebus, A. 2017,, 154, 262
Mordas | , M., Ginski, C., et al. 2016,, Marleau, Klahr, H., R., & Mordasini, M. Fortney, J. J., O., Bodenheimer, P., Lissauer, J. J. 2007,, 655, 541 C., Macintosh, B., Barman, T., et al. 2008, Science, 322, 1348 Marois, C., B., Konopacky, Q. M., Macintosh, B., & Barman, T. 2010,, 468, 1080 Matzner, D., Levin, 2005,, 817 Mawet, D., Milli, J., Wahhaj, Z., et al. 2014,, 792, 97 Mendigut[í]{}a, I., Oudmaijer, R. Schneider, P. C., et al. 2018,, 618, L9 D., Vigan, A., D’Orazi, et al. 2016,, 593, A119 D., J.-L., Charnay, et 2018,, A92 Meshkat, T., M., Mamajek, E. E., et al. 2015,, 453, 2378 Metchev, S. A., & Hillenbrand, L. A. 2006,, 1166 Metchev, & Hillenbrand, A. 181, Milli, J., Hibon, V., et al. 2017,, 597, L2 Metchev, S., Luhman, K. L., Marengo, M., & A. 2017,, 262 Mordas | , M., Ginski, C., et al. 2016,, 587, A56
Marleau, G.-D., KlaHr, H., Kuiper, R., & mordaSinI, C. 2017,, 836, 221
MArLey, M. s., ForTney, J. J., Hubickyj, o., bodeNheimer, P., & Lissauer, J. J. 2007,, 655, 541
MaroIs, C., MaCiNTosh, b., baRman, T., Et al. 2008, SciENcE, 322, 1348
mAroIs, c., ZUckErMAn, b., KonoPacKy, Q. M., MacIntosh, B., & BarMan, t. 2010,, 468, 1080
MAtzner, C. D., & LeviN, y. 2005,, 628, 817
MAwet, D., Milli, j., WaHhaj, Z., et al. 2014,, 792, 97
MenDigUt[í]{}a, I., OUdMaiJEr, R. D., SChnEider, p. C., et al. 2018,, 618, l9
mesa, D., VIgan, A., D’OraZi, v., Et al. 2016,, 593, A119
MESa, D., BaudINO, J.-l., ChaRnay, B., et al. 2018,, 612, A92
Meshkat, t., boNNefoy, M., Mamajek, E. e., et al. 2015,, 453, 2378
MEtCHeV, s. a., & HiLleNbrand, L. A. 2006,, 651, 1166
MeTcHev, S. A., & hIllenbrANd, l. a. 2009,, 181, 62
mIllI, j., Hibon, P., ChristIaens, V., et al. 2017,, 597, L2
mIleS-P[á]{}ez, P. a., MEtcHEv, S., LuhMan, K. L., maREngO, M., & Hulsebus, A. 2017,, 154, 262
mordAs | , M., Ginski, C., et al. 2 016,, 587, A56
Ma rle au , G. -D., Klahr, H., Ku i per, R., & Mordasini, C. 2 017,, 8 3 6, 2 2 1
Marl ey, M.S ., F ort ne y, J. J . ,Hubic kyj , O., B odenheimer , P ., & Lissauer, J. J. 2007,, 65 5, 541
Maro is, C., M ac int o sh, B .,Barma n, T., et al. 2008, Sc ie n ce, 32 2 , 1348M ar ois, C., Zuckerman, B . ,K onopacky, Q. M ., Mac in t os h , B. , & Barman, T .2010, , 468, 1 0 80 M atz n er, C. D., &Levin, Y. 2 0 05, , 628, 8 17Mawet, D.,Mi l li, J., Wahhaj , Z. , et al.2014,, 792, 97
Mendig ut[í]{ }a, I. , Ou d ma ij er, R . D. , S chn e ide r, P. C. ,et al.2018 , , 6 18,L9
Mes a, D. , Vigan, A.,D’O razi , V. , etal. 2 016, ,593,A119
Mesa, D ., Baudino, J.- L.,Charnay,B., e t a l. 2018 , , 612, A9 2
Meshkat , T., B o nne fo y , M. , Mamajek, E. E.,et a l. 2015,,453, 2 3 78
M etchev,S. A. , &H i llenb rand , L . A. 200 6,, 65 1 ,11 66
Met ch ev, S. A .,& H illen b rand , L. A . 2009,, 181, 62
Milli, J., Hibon, P., Ch r is t i ae n s, V .,et al. 2017 ,, 5 9 7, L 2
M i le s-P [ á]{}e z, P. A . ,M etchev, S., Luhman, K . L.,Maren go, M., & Hul sebus, A.2 0 1 7,, 154, 262
M o rdas | , M.,_Ginski, C.,_et al. 2016,, 587, A56
Marleau,_G.-D., Klahr,_H.,_Kuiper, R.,_&_Mordasini, C. 2017,, 836,_221
Marley, M. S., Fortney,_J. J., Hubickyj, O., Bodenheimer,_P., & Lissauer,_J. J. 2007,,_655, 541
Marois, C., Macintosh, B., Barman, T., et al. 2008, Science, 322, 1348
Marois, C., Zuckerman,_B.,_Konopacky, Q. M.,_Macintosh,_B.,_& Barman, T. 2010,, 468, 1080
Matzner,_C. D., & Levin, Y. 2005,, 628,_817
Mawet, D.,_Milli, J., Wahhaj, Z., et al. 2014,, 792, 97
Mendigut[í]{}a,_I.,_Oudmaijer, R. D., Schneider,_P. C., et al. 2018,, 618, L9
Mesa, D., Vigan, A., D’Orazi,_V., et al. 2016,, 593, A119
Mesa, D.,_Baudino, J.-L., Charnay,_B.,_et_al. 2018,, 612, A92
Meshkat, T.,_Bonnefoy, M., Mamajek, E. E., et al. 2015,,_453, 2378
Metchev, S. A., & Hillenbrand, L. A. 2006,,_651, 1166
Metchev, S. A., & Hillenbrand, L. A. 2009,, 181,_62
Milli, J., Hibon, P., Christiaens, V.,_et al. 2017,, 597, L2
Miles-P[á]{}ez, P. A.,_Metchev, S.,_Luhman, K. L., Marengo, M., &_Hulsebus, A. 2017,, 154,_262
Mordas |
and $c$ as in Eq. , but excluding the parameter $\psi$. Once the significant effects of memory length, level of conscious arousal, and channel location are characterised with our statistical analysis, we follow with post-hoc, one-sample two-tailed $t$-tests given by $$\begin{gathered}
t = \frac{\langle \Delta C_{\mu}\rangle_f - \mu_0}{s_f / \sqrt{|f|}},
\label{eq:ttest}\end{gathered}$$ in order to examine the nature of interactions between $\lambda$ and $c$ on [$\Delta C_{\mu}$]{}. We set $\mu_0$ to be the value of $\langle \Delta C_{\mu} \rangle_f$ under the null hypothesis $(\mu_0 = 0)$, $s_f$ is the standard deviation of $\langle \Delta C_{\mu} \rangle_f$, and $|f| = 13$ is the sample size. We present the results of these analyses in the following sections.
Results {#sec:results}
-------
In order to observe the effects of isoflurane on neural complexity, we begin by visually inspecting the structure of the reconstructed [$\epsilon\text{-machines}$]{} for the two levels of conscious arousal. We take special interest in observing the differences in the characteristics of the two groups of [$\epsilon\text{-machines}$]{} heralding the two levels of conscious arousal. Here, memory length $\lambda$ plays an important role. At a given $\lambda$, the maximum number of causal states that may be generated scales according to $|\mathcal{A}|^{\lambda}$ [@CSSR2]. In our case, the alphabet is binary, $\mathcal{A} = \{0,1\}$. This greatly restricts the space of [$\epsilon\text{-machine}$]{} configurations available for short history lengths [@Johnson2010]. For $\lambda = 2$ we can observe up to four distinct configurations, which is unlikely to reveal the difference based on conscious states. Given the previous findings [@CohenEneuro], we generally expected that the data from the wakeful state would present more complexity than those from the anaesthetised state.
Visual inspection of the directed graphs indeed suggests higher [$\epsilon\text{-machine}$]{} complexity during the wakeful state compared to the anaesthetised state, at a given set of parameters $\{f,c,\lambda\}$. In particular, the data from the anaesthetised | and $ c$ as in Eq. , but excluding the parameter $ \psi$. Once the significant effects of memory duration, degree of conscious arousal, and channel location are qualify with our statistical analysis, we follow with post - hoc, one - sample distribution two - tailed $ t$-tests given by $ $ \begin{gathered }
t = \frac{\langle \Delta C_{\mu}\rangle_f - \mu_0}{s_f / \sqrt{|f| } },
\label{eq: ttest}\end{gathered}$$ in ordering to examine the nature of interactions between $ \lambda$ and $ c$ on [ $ \Delta C_{\mu}$ ] { }. We set up $ \mu_0 $ to be the value of $ \langle \Delta C_{\mu } \rangle_f$ under the null hypothesis $ (\mu_0 = 0)$, $ s_f$ is the standard deviation of $ \langle \Delta C_{\mu } \rangle_f$, and $ |f| = 13 $ is the sample distribution size. We present the results of these analyses in the following sections.
Results { # sec: results }
-------
In ordering to observe the effects of isoflurane on neural complexity, we get down by visually inspecting the structure of the reconstructed [ $ \epsilon\text{-machines}$ ] { } for the two level of conscious arousal. We carry special interest in observing the differences in the characteristics of the two group of [ $ \epsilon\text{-machines}$ ] { } heralding the two levels of conscious arousal. Here, memory length $ \lambda$ plays an important role. At a given $ \lambda$, the maximum number of causal states that may be generated scales according to $ |\mathcal{A}|^{\lambda}$ [ @CSSR2 ]. In our case, the alphabet is binary, $ \mathcal{A } = \{0,1\}$. This greatly restricts the outer space of [ $ \epsilon\text{-machine}$ ] { } configurations available for short history length [ @Johnson2010 ]. For $ \lambda = 2 $ we cannote up to four distinct configurations, which is unlikely to reveal the difference based on conscious states. move over the previous findings [ @CohenEneuro ], we generally expected that the data from the wakeful state would present more complexity than those from the anaesthetised state.
Visual inspection of the directed graph indeed suggests high [ $ \epsilon\text{-machine}$ ] { } complexity during the waking state compared to the anaesthetised state, at a given hardening of parameters $ \{f, c,\lambda\}$. In particular, the data from the anaesthetised | anf $c$ as in Eq. , but excludikg the parameter $\psi$. Onre the aignificxnt effects of memory length, lwvel if conscious arousal, avd channep locatiin ace characterised with ouv staflsticcl analysis, we fpllow with post-hoc, one-sakpue two-tailed $t$-tests given by $$\begin{gatrered}
h = \frac{\langle \Qelts C_{\mu}\dangle_f - \mu_0}{s_f / \sqrt{|f|}},
\label{eq:tteat}\end{gauhered}$$ in order to examine the nature of intfrachions between $\lambfa$ and $c$ on [$\Deleq C_{\mu}$]{}. We set $\mu_0$ to be uhz value of $\mangle \Delta C_{\mu} \rangle_f$ under ghe nbll hypothewiw $(\mk_0 = 0)$, $s_f$ is thx stanqard deviation of $\latgle \Dekta C_{\mu} \rangle_n$, and $|f| = 13$ is the sample size. Xe present the resuljs of thesa cnalyses in the folloqibg sewtiots.
Reryltr {#stc:rxsumts}
-------
In lrdxr to obserbe the effexts of isoflurane om gviral complexjty, we bqgin by visually inspecting the structuge or the reconstructed [$\epsulon\text{-machines}$]{} for jhe two ledels of conscious arousal. We take special interesd in kcsexylng gye differences in the characteristics of the twj gtolps of [$\epsilon\texb{-machines}$]{} heraldinb hhr two levels ow consejohs arousal. Here, melory leggth $\oambda$ plwys sn important role. At a giveb $\lambda$, the naximum number of eausal stater thst mau be generated scales aecordihg to $|\mathcwl{A}|^{\lambda}$ [@DRSR2]. In our case, ghe anphabet is binary, $\mathcal{A} = \{0,1\}$. This gceatlv restrizts jhe spase of [$\epsipon\tewd{-machine}$]{} configurahions aeailable flr short history lengths [@Johnson2010]. For $\lambda = 2$ ee cat observz up tp four distinst configuratipns, whieh is jnlikely tk revean the diffewence based ot conscious svates. Givqn tye peevious windings [@CohenEmeuro], we ytnerally ezpected that the deta wdom the wakeful stqte would presemt oorq bom'lexief than those froo tfr anadsthetised wgate.
Fisual inspection of the directed graphs imdced suggewts highqr [$\epsilon\texy{-machine}$]{} complexitj durmng thx wakegul state compared to the anaesthstised stwte, at a given fet if parameterx $\{f,c,\lambda\}$. In particular, the data from vhe anaesthetised | and $c$ as in Eq. , but parameter Once the effects of memory and location are characterised our statistical analysis, follow with post-hoc, one-sample two-tailed $t$-tests by $$\begin{gathered} t = \frac{\langle \Delta C_{\mu}\rangle_f - \mu_0}{s_f / \sqrt{|f|}}, \label{eq:ttest}\end{gathered}$$ in to examine the nature of interactions between $\lambda$ and $c$ on [$\Delta C_{\mu}$]{}. set to the of $\langle \Delta C_{\mu} \rangle_f$ under the null hypothesis $(\mu_0 = 0)$, $s_f$ is the standard of $\langle \Delta C_{\mu} \rangle_f$, and $|f| = is the sample size. present the results of these in following sections. {#sec:results} In to observe the of isoflurane on neural complexity, we begin by visually inspecting the structure of the reconstructed [$\epsilon\text{-machines}$]{} for two levels arousal. We special in the differences in of the two groups of [$\epsilon\text{-machines}$]{} levels of conscious arousal. Here, memory length $\lambda$ an important At a given $\lambda$, the maximum of causal states that may be generated scales to $|\mathcal{A}|^{\lambda}$ [@CSSR2]. In our case, the alphabet is binary, $\mathcal{A} = \{0,1\}$. This greatly space of [$\epsilon\text{-machine}$]{} configurations for short history [@Johnson2010]. $\lambda 2$ can observe to four distinct configurations, which is unlikely to reveal the difference on conscious states. Given the previous findings [@CohenEneuro], we generally the from the wakeful would present more complexity those the anaesthetised state. Visual the graphs [$\epsilon\text{-machine}$]{} during wakeful state compared to anaesthetised state, at a given of parameters $\{f,c,\lambda\}$. In anaesthetised | and $c$ as in Eq. , but excluding the Parameter $\pSi$. OncE thE siGnIficAnt eFfects of memory LEngtH, level of conscious arousAl, and ChANnel LOcAtion Are charACtERIseD wItH ouR sTAtIsticAl aNalysis, We follow wiTh pOsT-hoc, one-samplE TwO-tailed $t$-teSts Given by $$\begin{GatHered}
t = \FrAc{\lANgle \DEltA C_{\mu}\rAngle_f - \MU_0}{s_f / \sqrT{|f|}},
\label{eq:TtESt}\end{gAThered}$$ iN ORdEr to Examine the nature oF InTEractions betweEn $\lambDa$ ANd $C$ ON [$\DeLta c_{\mu}$]{}. We set $\mu_0$ To Be the VAlue of $\lANgLE \dEltA c_{\mu} \rangle_f$ undEr the null hyPOthEsis $(\mu_0 = 0)$, $S_f$ Is tHE standArd deViATioN of $\langle \DeLta C_{\Mu} \rangle_f$, And $|f| = 13$ is THe samplE Size. We pResent The ResUlts OF tHeSe aNaLYseS In The FOllOwing secTiOnS.
ResuLts {#sEC:RESultS}
-------
In OrdeR to obServe the effecTs oF isoFLurAne on NeuraL comPlExity, We begiN by viSuAlly inspecting tHe stRucture of The ReConStRucteD [$\EpsiloN\teXt{-mAchines}$]{} For the tWO leVeLS OF cOnscious arousal. We tAkE SPeCial inteRest in OBsErVIng the diFfEreNces IN The chAracTErIstics of The two GRoUpS of [$\epsiLoN\text{-mAcHinEs}$]{} hEraldINg thE two leVels of coNscioUS arousal. Here, meMOry length $\lambDA$ pLAYs AN impOrtAnt role. At a gIven $\LAmbdA$, the MAxImuM NumbeR of caUsAL sTAtes that may be generaTeD scaleS accoRding to $|\mathcaL{A}|^{\lambda}$ [@CSsr2]. iN our case, The aLPhABet is binary, $\matHcal{A} = \{0,1\}$. this greatlY RestrictS the sPace of [$\epSilon\text{-MAChine}$]{} conFigUraTioNs aVAIlAble for short hISTory LeNgths [@JoHnsOn2010]. For $\laMbdA = 2$ we Can ObsErVe up to fouR distincT cOnFiGuRatIons, wHIch is unlIkEly To RevEal thE DifferEnce bAsed On CoNSciOus statES. GIVEn thE pReViouS fiNdIngs [@COheneNeuRo], we genErally expEctED thaT tHe Data froM the wakeful stAtE would presEnT moRe compLEXity than Those from the anaesthetisED state.
VIsuAl insPectIon of the dIreCted grAphS Indeed SuggesTs higHeR [$\epSILon\teXT{-MaChiNe}$]{} Complexity DURinG the wAkEful State coMpared to the anaesthETisEd state, at a givEn sEt of PARaMetERs $\{F,C,\laMbDA\}$. In PARticular, the data From the anaEsTHeTised | and $c$ as in Eq. , but e xcluding t he pa ram ete r$\ps i$.Once the signi f ican t effects of memory le ngth, l e velo fconsc ious ar o us a l , a nd c han ne l l ocati onare cha racterised wi th our statist i ca l analysis , w e follow wit h p ost-ho c, on e -samp letwo-t ailed$ t$-tes ts givenby $$\beg i n{gathe r e d}
t = \frac{\langl e \ D elta C_{\mu}\r angle_ f- \ m u _0} {s_ f / \sqrt{ |f |}}, \lab e l{ e q : tte s t}\end{gather ed}$$ in or d erto exa mi net he nat ure o fi nte ractions be twee n $\lambd a$ and $c$ on[ $\Delta C_{\m u}$ ]{} . We se t$\m u_ 0 $ t o b e t h e v alue of$\ la ngle\Del t a C _{\m u}\ran gle_f $ under the n ull hyp o the sis $ (\mu_ 0 =0) $, $s _f$ is thest andard deviatio n of $\langle \D el taC_ {\mu} \rangl e_f $,and $|f | = 13$ isth e s am ple size. We prese nt t he results of th e se a n alyses i nthe fol l o wingsect i on s.
Resu lts {# s ec :r esults}
- ------
Inord er to obse rve th e effect s ofi soflurane on n e ural complexi t y, w eb egin by visually i nspe c ting the st ruc t ure o f the r e co n structed [$\epsilon \t ext{-m achin es}$]{} for t he two lev e l s of cons ciou s a r ousal. We take spec ial intere s t in obs ervin g the di fferences i n the ch ara cte ris tic s of the two grou p s of[$ \epsilo n\t ext{-ma chi nes }$] {}he ralding t he two l ev el sof co nscio u s arousa l. He re , m emory length $\la mbda $pl a ysan impo r ta n t rol e. A t agiv en $\la mbda $ , t he maxi mum numbe r o f cau sa lstatesthat may be g en erated sca le s a ccordi n g to $|\m athcal{A}|^{\lambda}$ [ @ CSSR2]. In ourcase , the alp hab et isbin a ry, $\ mathca l{A}=\{0 , 1 \}$.T h is gr ea tly restri c t s t he sp ac e of [$\eps ilon\text{-machine } $]{ } configurati ons ava i l ab lef or sho rt his t o ry lengths [@Jo hnson2010] .F or $\lambda= 2$ w e can o bserveup to four di stinct co nfigurati on s, w h i chis unlikel y to rev eal the d i ffere n ce base d o n cons ci ous stat es. Gi v enthe p reviou sfindin gs [@ Co henEneur o], we generally expect ed tha t the da ta from t hew ake ful state wou ld present mo recompl exi t y tha n th o se fr o m the ana e sthetised st ate .
V isual inspe c t i onof th e d i rected gra phs indeed sugges t s higher [$\ep silo n \ tex t{- m achi ne }$]{} complexi tydu r i ng the w ak eful statecompared t o theanaest hetise d state , at a give n se t o f paramet ers $ \ {f,c,\l am bd a \}$. I n pa rt icular , thed ataf r om the anaesthet ised | and_$c$ as_in Eq. , but excluding_the parameter_$\psi$._Once the_significant_effects of memory_length, level of_conscious arousal, and channel_location are characterised_with_our statistical analysis, we follow with post-hoc, one-sample two-tailed $t$-tests given by $$\begin{gathered}
__ t_=_\frac{\langle_\Delta C_{\mu}\rangle_f - \mu_0}{s_f /_\sqrt{|f|}},
\label{eq:ttest}\end{gathered}$$_in order_to examine the nature of interactions between $\lambda$_and_$c$ on [$\Delta_C_{\mu}$]{}. We set $\mu_0$ to be the value of_$\langle \Delta C_{\mu} \rangle_f$ under the_null hypothesis $(\mu_0_=_0)$,_$s_f$ is the standard_deviation of $\langle \Delta C_{\mu} \rangle_f$,_and $|f| = 13$ is the_sample size. We present the results of_these analyses in the following sections.
Results_{#sec:results}
-------
In order to observe the_effects of_isoflurane on neural complexity, we_begin by visually_inspecting the_structure of the_reconstructed [$\epsilon\text{-machines}$]{} for the two levels_of conscious arousal._We take special interest in observing_the_differences in the_characteristics_of_the two_groups of [$\epsilon\text{-machines}$]{}_heralding_the two_levels_of conscious arousal. Here, memory length_$\lambda$_plays an important role. At a given_$\lambda$, the maximum number_of_causal states that may_be generated scales according to_$|\mathcal{A}|^{\lambda}$ [@CSSR2]. In our case, the alphabet_is binary,_$\mathcal{A} =_\{0,1\}$. This greatly restricts the space of [$\epsilon\text{-machine}$]{} configurations available for_short history lengths [@Johnson2010]. For $\lambda =_2$ we can observe_up to_four_distinct configurations, which_is_unlikely to_reveal the difference based on conscious states._Given the_previous findings [@CohenEneuro], we generally expected that_the data from the_wakeful_state would present more complexity than_those from the anaesthetised state.
Visual inspection_of the directed graphs indeed_suggests_higher_[$\epsilon\text{-machine}$]{} complexity during the wakeful_state compared to the anaesthetised state,_at a given_set of parameters $\{f,c,\lambda\}$. In particular, the_data_from the anaesthetised |
t)}{\lambda_{0}(t)} e'(t) = \frac{3 G(t,\lambda_{0}(t)\left(1+e(t)\right))}{2}\left(1+e(t)\right)$$ where $$G(t,\lambda(t)) = v_{sip}(t,\lambda(t))+lin_{ip}(t,\lambda(t))+w_{c,ip}(t,\lambda(t))-E_{v_{1},ip}(t,\lambda(t))$$ Let $\mathcal{B} :=\overline{B}_{1}(0) \subset X$. Our goal is to solve for $e \in \mathcal{B}$ using a fixed point argument. So, we define $T$ on $\mathcal{B}$ by $$T(e)(t) = \int_{t}^{\infty} \frac{dx}{\lambda_{0}(x)^{2}} \int_{x}^{\infty} \frac{3}{2} \lambda_{0}(s)^{2} G(s,\lambda_{0}(s)\left(1+e(s)\right))\left(1+e(s)\right) ds$$ Combining our estimates above, we get $$|G(t,\lambda_{0}(t)\left(1+e(t)\right))| \leq \frac{C}{t^{2} \log^{1+\delta}(t)}$$ where we recall the definition of $\delta$ in. This gives $$|\int_{x}^{\infty} \frac{3}{2}\lambda_{0}(s)^{2} G(s,\lambda_{0}(s)\left(1+e(s)\right))\left(1+e(s)\right) ds| \leq C \int_{x}^{\infty} \frac{\lambda_{0}(s)^{2} ds}{s^{2} \log^{1+\delta}(s)}$$ Then, we integrate by parts to get $$\int_{x}^{\infty} \frac{\lambda_{0}(s)^{2} ds}{s^{2} \log^{1+\delta}(s)} = \frac{\lambda_{0}(x)^{2}}{x \log^{1+\delta}(x)} + \int_{x}^{\infty}\frac{ds}{s}\left(\frac{2 \lambda_{0}(s)\lambda_{0}'(s)}{\log^{1+\delta}(s)}-\frac{(\delta+ | t)}{\lambda_{0}(t) } e'(t) = \frac{3 G(t,\lambda_{0}(t)\left(1+e(t)\right))}{2}\left(1+e(t)\right)$$ where $ $ G(t,\lambda(t) ) = v_{sip}(t,\lambda(t))+lin_{ip}(t,\lambda(t))+w_{c, ip}(t,\lambda(t))-E_{v_{1},ip}(t,\lambda(t))$$ Let $ \mathcal{B }: = \overline{B}_{1}(0) \subset X$. Our goal is to solve for $ e \in \mathcal{B}$ use a situate point argument. So, we specify $ T$ on $ \mathcal{B}$ by $ $ T(e)(t) = \int_{t}^{\infty } \frac{dx}{\lambda_{0}(x)^{2 } } \int_{x}^{\infty } \frac{3}{2 } \lambda_{0}(s)^{2 } G(s,\lambda_{0}(s)\left(1+e(s)\right))\left(1+e(s)\right) ds$$ Combining our estimates above, we get down $ $ |G(t,\lambda_{0}(t)\left(1+e(t)\right))| \leq \frac{C}{t^{2 } \log^{1+\delta}(t)}$$ where we recall the definition of $ \delta$ in. This gives $ $ |\int_{x}^{\infty } \frac{3}{2}\lambda_{0}(s)^{2 } G(s,\lambda_{0}(s)\left(1+e(s)\right))\left(1+e(s)\right) ds| \leq C \int_{x}^{\infty } \frac{\lambda_{0}(s)^{2 } ds}{s^{2 } \log^{1+\delta}(s)}$$ Then, we desegregate by part to get $ $ \int_{x}^{\infty } \frac{\lambda_{0}(s)^{2 } ds}{s^{2 } \log^{1+\delta}(s) } = \frac{\lambda_{0}(x)^{2}}{x \log^{1+\delta}(x) } + \int_{x}^{\infty}\frac{ds}{s}\left(\frac{2 \lambda_{0}(s)\lambda_{0}'(s)}{\log^{1+\delta}(s)}-\frac{(\delta+ | t)}{\lalbda_{0}(t)} e'(t) = \frac{3 G(t,\lambda_{0}(t)\ueft(1+e(t)\right))}{2}\left(1+g(t)\eight)$$ xhere $$G(f,\lambda(t)) = v_{sip}(t,\lambda(t))+lin_{ip}(t,\lambda(t))+w_{r,ip}(t,\oambdq(t))-E_{v_{1},ip}(t,\lambda(t))$$ Let $\matfcal{B} :=\ovegline{B}_{1}(0) \suvset Z$. Our goal is to solve fkv $e \iu \nathcal{B}$ using a fixed pmint argument. Vo, wz define $T$ on $\mathcal{B}$ by $$T(e)(t) = \int_{t}^{\igfty} \frsc{fx}{\lambda_{0}(x)^{2}} \int_{x}^{\ynftj} \srac{3}{2} \lambda_{0}(s)^{2} G(s,\lambda_{0}(s)\left(1+e(s)\right))\left(1+e(a)\right) vs$$ Combining out estimates above, we get $$|G(h,\lamhda_{0}(t)\left(1+e(t)\right))| \lee \frac{C}{t^{2} \lot^{1+\deleq}(t)}$$ where we fecall the definition kf $\delta$ in. This gives $$|\int_{x}^{\infth} \frae{3}{2}\lambda_{0}(s)^{2} G(s,\oanbdw_{0}(v)\left(1+e(s)\righv))\left(1+e(f)\right) ds| \leq C \int_{x}^{\hnfty} \ftac{\lambda_{0}(s)^{2} ds}{s^{2} \loj^{1+\delra}(s)}$$ Then, we integrate by parts to get $$\igt_{x}^{\infty} \xrcc{\lambda_{0}(s)^{2} ds}{s^{2} \log^{1+\deltq}(s)} = \fraw{\lamtda_{0}(x)^{2}}{b \loe^{1+\demte}(x)} + \int_{x}^{\ijftb}\frac{ds}{s}\leff(\frac{2 \lambdq_{0}(s)\lambda_{0}'(s)}{\log^{1+\delta}(s)}-\ftas{(\eelta+ | t)}{\lambda_{0}(t)} e'(t) = \frac{3 G(t,\lambda_{0}(t)\left(1+e(t)\right))}{2}\left(1+e(t)\right)$$ where $$G(t,\lambda(t)) Let :=\overline{B}_{1}(0) \subset Our goal is \mathcal{B}$ a fixed point So, we define on $\mathcal{B}$ by $$T(e)(t) = \int_{t}^{\infty} \int_{x}^{\infty} \frac{3}{2} \lambda_{0}(s)^{2} G(s,\lambda_{0}(s)\left(1+e(s)\right))\left(1+e(s)\right) ds$$ Combining our estimates above, we get $$|G(t,\lambda_{0}(t)\left(1+e(t)\right))| \leq \log^{1+\delta}(t)}$$ where we recall the definition of $\delta$ in. This gives $$|\int_{x}^{\infty} \frac{3}{2}\lambda_{0}(s)^{2} ds| C \frac{\lambda_{0}(s)^{2} \log^{1+\delta}(s)}$$ Then, we integrate by parts to get $$\int_{x}^{\infty} \frac{\lambda_{0}(s)^{2} ds}{s^{2} \log^{1+\delta}(s)} = \frac{\lambda_{0}(x)^{2}}{x \log^{1+\delta}(x)} + \lambda_{0}(s)\lambda_{0}'(s)}{\log^{1+\delta}(s)}-\frac{(\delta+ | t)}{\lambda_{0}(t)} e'(t) = \frac{3 G(t,\lambda_{0}(t)\leFt(1+e(t)\right))}{2}\lEft(1+e(t)\RigHt)$$ wHeRe $$G(t,\LambDa(t)) = v_{sip}(t,\lambda(T))+Lin_{iP}(t,\lambda(t))+w_{c,ip}(t,\lambda(t))-E_{V_{1},ip}(t,\lAmBDa(t))$$ LET $\mAthcaL{B} :=\overlINe{b}_{1}(0) \SUbsEt x$. OUr gOaL Is To solVe fOr $e \in \maThcal{B}$ usinG a fIxEd point argumENt. so, we define $t$ on $\Mathcal{B}$ by $$T(e)(T) = \inT_{t}^{\inftY} \fRac{DX}{\lambDa_{0}(x)^{2}} \Int_{x}^{\iNfty} \frAC{3}{2} \lambdA_{0}(s)^{2} G(s,\lambdA_{0}(s)\LEft(1+e(s)\rIGht))\left(1+E(S)\RiGht) dS$$ Combining our estiMAtES above, we get $$|G(t,\lAmbda_{0}(t)\LeFT(1+e(T)\RIghT))| \leQ \frac{C}{t^{2} \log^{1+\DeLta}(t)}$$ wHEre we reCAlL THE deFInition of $\deltA$ in. This giveS $$|\Int_{X}^{\infty} \FrAc{3}{2}\lAMbda_{0}(s)^{2} G(S,\lambDa_{0}(S)\LefT(1+e(s)\right))\lefT(1+e(s)\rIght) ds| \leq c \int_{x}^{\iNFty} \frac{\LAmbda_{0}(s)^{2} dS}{s^{2} \log^{1+\dEltA}(s)}$$ THen, wE InTeGraTe BY paRTs To gET $$\inT_{x}^{\infty} \fRaC{\lAmbda_{0}(S)^{2} ds}{s^{2} \LOG^{1+\DElta}(S)} = \frAc{\laMbda_{0}(x)^{2}}{X \log^{1+\delta}(x)} + \int_{X}^{\inFty}\fRAc{dS}{s}\lefT(\frac{2 \LambDa_{0}(S)\lambDa_{0}'(s)}{\log^{1+\Delta}(S)}-\fRac{(\delta+ | t)}{\lambda_{0}(t)} e'(t)= \frac{3G(t,\ lam bda _{ 0}(t )\le ft(1+e(t)\righ t ))}{ 2}\left(1+e(t)\right)$ $ whe re $$G( t ,\ lambd a(t)) = v_ { s ip} (t ,\ lam bd a (t ))+li n_{ ip}(t,\ lambda(t)) +w_ {c ,ip}(t,\lamb d a( t))-E_{v_{ 1}, ip}(t,\lambd a(t ))$$ L et $\ m athca l{B } :=\ overli n e{B}_{ 1}(0) \su bs e t X$.O ur goal i sto s olve for $e \in \ m at h cal{B}$ usinga fixe dp oi n t ar gum ent. So, w edefin e $T$ on $\ m a t hca l {B}$ by $$T(e )(t) = \int _ {t} ^{\inf ty } \ f rac{dx }{\la mb d a_{ 0}(x)^{2}}\int _{x}^{\in fty} \ f rac{3}{ 2 } \lamb da_{0} (s) ^{2 } G( s ,\ la mbd a_ { 0}( s )\ lef t (1+ e(s)\rig ht )) \left (1+e ( s ) \ righ t)ds$$ Comb ining our est ima tesa bov e, we get$$|G (t ,\lam bda_{0 }(t)\ le ft(1+e(t)\right ))|\leq \fra c{C }{ t^{ 2} \log ^ {1+\de lta }(t )}$$ wh ere wer eca ll t h edefinition of $\de lt a $ i n. Thisgives$ $| \i n t_{x}^{\ in fty } \f r a c{3}{ 2}\l a mb da_{0}(s )^{2}G (s ,\ lambda_ {0 }(s)\l ef t(1 +e( s)\ri g ht)) \left( 1+e(s)\r ight) ds| \leq C \in t _{x}^{\infty} \f r a c{ \ lamb da_ {0}(s)^{2}ds}{ s ^{2} \lo g ^{ 1+\ d elta} (s)}$ $T he n , we integrate by p ar ts toget $ $\int_{x}^{\i nfty} \fra c { \ lambda_{ 0}(s ) ^{ 2 } ds}{s^{2} \l og^{1 +\delta}(s ) } = \fra c{\la mbda_{0} (x)^{2}}{ x \log^{1+ \de lta }(x )}+ \i nt_{x}^{\inft y } \fra c{ ds}{s}\ lef t(\frac {2\la mbd a_{ 0} (s)\lambd a_{0}'(s )} {\ lo g^ {1+ \delt a }(s)}-\f ra c{( \d elt a+ | t)}{\lambda_{0}(t)} e'(t)_= \frac{3_G(t,\lambda_{0}(t)\left(1+e(t)\right))}{2}\left(1+e(t)\right)$$ where $$G(t,\lambda(t)) =_v_{sip}(t,\lambda(t))+lin_{ip}(t,\lambda(t))+w_{c,ip}(t,\lambda(t))-E_{v_{1},ip}(t,\lambda(t))$$ Let_$\mathcal{B}_:=\overline{B}_{1}(0) \subset_X$._Our goal is_to solve for_$e \in \mathcal{B}$ using_a fixed point_argument._So, we define $T$ on $\mathcal{B}$ by $$T(e)(t) = \int_{t}^{\infty} \frac{dx}{\lambda_{0}(x)^{2}} \int_{x}^{\infty} \frac{3}{2} \lambda_{0}(s)^{2}_G(s,\lambda_{0}(s)\left(1+e(s)\right))\left(1+e(s)\right)_ds$$ Combining_our_estimates_above, we get $$|G(t,\lambda_{0}(t)\left(1+e(t)\right))| \leq_\frac{C}{t^{2} \log^{1+\delta}(t)}$$ where we recall_the definition_of $\delta$ in. This gives $$|\int_{x}^{\infty} \frac{3}{2}\lambda_{0}(s)^{2} G(s,\lambda_{0}(s)\left(1+e(s)\right))\left(1+e(s)\right)_ds|_\leq C \int_{x}^{\infty}_\frac{\lambda_{0}(s)^{2} ds}{s^{2} \log^{1+\delta}(s)}$$ Then, we integrate by parts to_get $$\int_{x}^{\infty} \frac{\lambda_{0}(s)^{2} ds}{s^{2} \log^{1+\delta}(s)} =_\frac{\lambda_{0}(x)^{2}}{x \log^{1+\delta}(x)} +_\int_{x}^{\infty}\frac{ds}{s}\left(\frac{2_\lambda_{0}(s)\lambda_{0}'(s)}{\log^{1+\delta}(s)}-\frac{(\delta+ |
of $f_1,\ldots,f_t$, and therefore the relaxation is well-defined.
\[pps:OMDrelaxadmin\] Suppose that the learning rate $\rho$ is such that $\rho \| f_t \|^*_{\nabla F(\nu_t)} \le 1/2$ for all $t = 1,\ldots,T$. Assume that $\| f_t \|^*_{\nabla F(\nu_t)} \le 1$, for all $t = 1,\ldots,T$. The relaxation is admissible, and the algorithm that generates the sequence $\{\nu_t\}$ is also admissible: $$\wh{V}_T({\cal G}'|f_1,\ldots,f_{t-1}) \ge \sup_{f \in \cF}\left\{ \ave{\nu_t,f} + \wh{V}_T({\mathcal G}' | f_1, \ldots,f_{t-1},f)\right\}$$
See Appendix \[app:OMDrelaxadmin\].
Here we impose the assumption that $\rho \| f_t \|^*_{\nabla F(\nu_t)} \le 1/2$ for all $t = 1,\ldots,T$. A restriction of this kind is necessary when using interior-point methods to construct online optimization schemes — see, for example, the condition of Theorem 4.1 and 4.2 in [@AbernethyIPM]. The boundedness of the dual local norm $\| f_t \|^*_{\nabla F(\nu_t)}$ is a reasonable assumption as well. In particular, Abernethy et al. [@AbernethyIPM] points out that if a large number of the points $\nu_t$ are close to the boundary of ${\cal G}'$, then the eigenvalues of the Hessian of $F$ at those points will be large due to the large curvature of the barrier near the boundary of ${\cal G}'$. This will imply, in turn, that the dual local norm $\| f_t \|^*_{\nabla F(\nu_t)}$ is expected to be small.
Now we are ready to present the online-learning (i.e., steady-state) part of the regret bound for the above algorithm:
\[thm:OMDbound1\] Let $D_F({\cal G'}) \deq\sup_{\nu \in {\cal G}'} D_{F}(\ | of $ f_1,\ldots, f_t$, and therefore the relaxation is well - defined.
\[pps: OMDrelaxadmin\ ] Suppose that the eruditeness pace $ \rho$ is such that $ \rho \| f_t \|^*_{\nabla F(\nu_t) } \le 1/2 $ for all $ thyroxine = 1,\ldots, T$. Assume that $ \| f_t \|^*_{\nabla F(\nu_t) } \le 1 $, for all $ t = 1,\ldots, T$. The rest is admissible, and the algorithm that generates the sequence $ \{\nu_t\}$ is besides admissible: $ $ \wh{V}_T({\cal G}'|f_1,\ldots, f_{t-1 }) \ge \sup_{f \in \cF}\left\ { \ave{\nu_t, f } + \wh{V}_T({\mathcal G }' | f_1, \ldots, f_{t-1},f)\right\}$$
See Appendix \[app: OMDrelaxadmin\ ].
Here we impose the premise that $ \rho \| f_t \|^*_{\nabla F(\nu_t) } \le 1/2 $ for all $ t = 1,\ldots, T$. A restriction of this kind is necessary when use interior - point method acting to manufacture online optimization schemes — see, for example, the condition of Theorem 4.1 and 4.2 in [ @AbernethyIPM ]. The finiteness of the dual local norm $ \| f_t \|^*_{\nabla F(\nu_t)}$ is a reasonable presumption as well. In particular, Abernethy et al. [ @AbernethyIPM ] sharpen out that if a large act of the points $ \nu_t$ are close to the boundary of $ { \cal G}'$, then the eigenvalues of the Hessian of $ F$ at those points will be big due to the large curvature of the barrier near the boundary of $ { \cal G}'$. This will imply, in turn, that the dual local norm $ \| f_t \|^*_{\nabla F(\nu_t)}$ is expected to be small.
Now we are ready to present the online - learning (i.e., steady - state) part of the regret bound for the above algorithm:
\[thm: OMDbound1\ ] permit $ D_F({\cal G' }) \deq\sup_{\nu \in { \cal G }' } D_{F}(\ | of $f_1,\ldots,f_t$, and therefore uhe relaxation is well-dehined.
\[ppa:OMDrelabadmin\] Suppose that the leariing rate $\rho$ is such that $\rho \| f_t \|^*_{\nabla F(\nu_t)} \le 1/2$ foc all $t = 1,\ldots,T$. Easume tmct $\| f_f \|^*_{\nabne F(\nu_t)} \le 1$, for sll $t = 1,\ldods,T$. The relaxadiun is admissible, and the algorithm thwt genetahes the sequense $\{\nl_t\}$ is zlso admissible: $$\wh{V}_T({\cal G}'|f_1,\ldots,f_{t-1}) \fe \sup_{f \in \cF}\left\{ \avr{\nu_t,f} + \wh{V}_T({\mathcal G}' | f_1, \ldlts,f_{h-1},f)\right\}$$
See Appendid \[app:OMDrelazadmyb\].
Here we impuse the assumption thaj $\rho \| f_t \|^*_{\nabla F(\nu_t)} \le 1/2$ for all $t = 1,\lbots,T$. A resjxuctlmn of this jind ps necessary when usitg intetior-point methpds to construct online optmmization schemes — sge, for exakpme, the condition if Theotem 4.1 atd 4.2 kb [@AcerhevhyJPM]. Thf bkundedness of the duao local norm $\| f_t \|^*_{\nanlw F(\nu_t)}$ is a rezsonabje assumption as well. In particular, Aberntthy st al. [@AbernethyIPM] poinrs out that if a largg number os the points $\nu_t$ are close to the boundary of ${\cal G}'$, thxn tht clgenxqlkes of the Hessian of $F$ at those points will fs kagge due to the lavge curvature of tne bstrier near the boundcdy of ${\cal G}'$. This wilp imply, in tyrn, that uhe dial local norm $\| f_t \|^*_{\nabla F(\ny_t)}$ is expectvd ti be small.
Now we axe ready to 'resenj the pnline-learning (i.e., steadv-state) part of thf regret guund for the aboxe slcorithm:
\[thm:OMDbound1\] Let $D_F({\cwl G'}) \deq\snp_{\nu \nn {\cal G}'} D_{F}(\ | of $f_1,\ldots,f_t$, and therefore the relaxation is Suppose the learning $\rho$ is such F(\nu_t)} 1/2$ for all = 1,\ldots,T$. Assume $\| f_t \|^*_{\nabla F(\nu_t)} \le 1$, all $t = 1,\ldots,T$. The relaxation is admissible, and the algorithm that generates sequence $\{\nu_t\}$ is also admissible: $$\wh{V}_T({\cal G}'|f_1,\ldots,f_{t-1}) \ge \sup_{f \in \cF}\left\{ \ave{\nu_t,f} + G}' f_1, See \[app:OMDrelaxadmin\]. Here we impose the assumption that $\rho \| f_t \|^*_{\nabla F(\nu_t)} \le 1/2$ for all = 1,\ldots,T$. A restriction of this kind is when using interior-point methods construct online optimization schemes — for the condition Theorem and in [@AbernethyIPM]. The of the dual local norm $\| f_t \|^*_{\nabla F(\nu_t)}$ is a reasonable assumption as well. In particular, et al. out that a number the points $\nu_t$ to the boundary of ${\cal G}'$, of the Hessian of $F$ at those points be large to the large curvature of the near the boundary of ${\cal G}'$. This will in turn, that the dual local norm $\| f_t \|^*_{\nabla F(\nu_t)}$ is expected to be we are ready to the online-learning (i.e., part the bound the above \[thm:OMDbound1\] Let $D_F({\cal G'}) \deq\sup_{\nu \in {\cal G}'} D_{F}(\ | of $f_1,\ldots,f_t$, and therefore the Relaxation Is welL-deFinEd.
\[Pps:OmDreLaxadmin\] SupposE That The learning rate $\rho$ is suCh thaT $\rHO \| f_t \|^*_{\nABlA F(\nu_t)} \Le 1/2$ for alL $T = 1,\lDOTs,T$. asSuMe tHaT $\| F_t \|^*_{\Nabla f(\nu_T)} \le 1$, for aLl $t = 1,\ldots,T$. THe rElAxation is admISsIble, and the AlgOrithm that geNerAtes thE sEquENce $\{\nu_T\}$ is Also aDmissiBLe: $$\wh{V}_T({\Cal G}'|f_1,\ldotS,f_{T-1}) \Ge \sup_{f \IN \cF}\left\{ \AVE{\nU_t,f} + \wH{V}_T({\mathcal G}' | f_1, \ldots,F_{T-1},f)\RIght\}$$
See AppendiX \[app:OMdrELaXADmiN\].
HeRe we impose ThE assuMPtion thAT $\rHO \| F_T \|^*_{\naBLa F(\nu_t)} \le 1/2$ for alL $t = 1,\ldots,T$. A reSTriCtion oF tHis KInd is nEcessArY WheN using interIor-pOint methoDs to coNStruct oNLine optImizatIon SchEmes — SEe, FoR exAmPLe, tHE cOndITioN of TheorEm 4.1 AnD 4.2 in [@AbErneTHYipM]. ThE boUndeDness Of the dual locaL noRm $\| f_t \|^*_{\NAblA F(\nu_t)}$ Is a reAsonAbLe assUmptioN as weLl. in particular, AbeRnetHy et al. [@AbeRneThYIPm] pOints OUt that If a LarGe numbeR of the pOIntS $\nU_T$ ARe Close to the boundary Of ${\CAL G}'$, Then the eIgenvaLUeS oF The HessiAn Of $F$ At thOSE poinTs wiLL bE large duE to the LArGe CurvatuRe Of the bArRieR neAr the BOundAry of ${\cAl G}'$. This wIll imPLy, in turn, that thE Dual local norm $\| F_T \|^*_{\nABLa f(\Nu_t)}$ iS exPected to be sMall.
nOw we Are rEAdY to PResenT the oNlINe-LEarning (i.e., steady-statE) pArt of tHe regRet bound for thE above algoRITHm:
\[thm:OMDBounD1\] leT $d_F({\cal G'}) \deq\sup_{\nu \In {\cal g}'} D_{F}(\ | of $f_1,\ldots,f_t$, andthereforethe r ela xat io n is wel l-defined.
\[ p ps:O MDrelaxadmin\] Suppose that t h e le a rn ing r ate $\r h o$ i s s uc htha t$ \r ho \| f_ t \|^*_ {\nabla F( \nu _t )} \le 1/2$f or all $t =1,\ ldots,T$. As sum e that $ \|f _t \| ^*_ {\nab la F(\ n u_t)}\le 1$, f or all $t = 1,\ld o t s, T$.The relaxation is ad m issible, and t he alg or i th m tha t g enerates t he sequ e nce $\{ \ nu _ t \ }$i s also admiss ible: $$\wh { V}_ T({\ca lG}' | f_1,\l dots, f_ { t-1 }) \ge \sup _{f\in \cF}\ left\{ \ave{\n u _t,f} + \wh{V }_T ({\ math c al G }'|f _1, \l dot s ,f_ {t-1},f) \r ig ht\}$ $
S e e A ppen dix \[a pp:OM Drelaxadmin\] .
Here weimpos e the ass um ption that$\rho \ | f_t \|^*_{\na blaF(\nu_t)} \l e1/2 $for a l l $t = 1, \ld ots,T$. A rest r ict io n o fthis kind is neces sa r y w hen usin g inte r io r- p oint met ho dsto c o n struc t on l in e optimi zation sc he mes — s ee , forex amp le, thec ondi tion o f Theore m 4.1 and 4.2 in [@A b ernethyIPM].T he b ou n dedn ess of the dua l lo c al n orm$ \| f_ t \|^* _{\na bl a F ( \nu_t)}$ is a reaso na ble as sumpt ion as well.In particu l a r , Aberne thye ta l. [@Abernethy IPM]points out that ifa lar ge numbe r of thep o ints $\n u_t $ a reclo s e t o the boundar y of $ {\ cal G}' $,then th e e ige nva lue sof the He ssian of $ F$ a ttho se po i nts will b e l ar gedue t o the l argecurv at ur e of the ba r ri e r nea rth e bo und ar y of${\c a l G }'$. Th is will i mpl y , in t ur n, that the dual loc al norm $\|f_ t \ |^*_{\ n a bla F(\n u_t)}$ is expected to b e small.
N ow we are ready to pr esentthe online -learn ing ( i. e., s teady - s ta te) p art of the r egr et bo un d fo r the a bove algorithm:
\ [ thm :OMDbound1\]Let $D_ F ( {\ cal G' } ) \ de q \su p _ {\nu \in {\calG}'} D_{F} (\ | of_$f_1,\ldots,f_t$, and_therefore the relaxation is_well-defined.
\[pps:OMDrelaxadmin\] Suppose_that_the learning_rate_$\rho$ is such_that $\rho \|_f_t \|^*_{\nabla F(\nu_t)} \le_1/2$ for all_$t_= 1,\ldots,T$. Assume that $\| f_t \|^*_{\nabla F(\nu_t)} \le 1$, for all $t =_1,\ldots,T$._The relaxation_is_admissible,_and the algorithm that generates_the sequence $\{\nu_t\}$ is also_admissible: $$\wh{V}_T({\cal_G}'|f_1,\ldots,f_{t-1}) \ge \sup_{f \in \cF}\left\{ \ave{\nu_t,f} + \wh{V}_T({\mathcal_G}'_| f_1, \ldots,f_{t-1},f)\right\}$$
See_Appendix \[app:OMDrelaxadmin\].
Here we impose the assumption that $\rho \| f_t_\|^*_{\nabla F(\nu_t)} \le 1/2$ for all_$t = 1,\ldots,T$._A_restriction_of this kind is_necessary when using interior-point methods to_construct online optimization schemes — see,_for example, the condition of Theorem 4.1 and_4.2 in [@AbernethyIPM]. The boundedness of_the dual local norm $\|_f_t \|^*_{\nabla_F(\nu_t)}$ is a reasonable assumption_as well. In_particular, Abernethy_et al. [@AbernethyIPM]_points out that if a large_number of the_points $\nu_t$ are close to the_boundary_of ${\cal G}'$,_then_the_eigenvalues of_the Hessian of_$F$_at those_points_will be large due to the_large_curvature of the barrier near the boundary_of ${\cal G}'$. This_will_imply, in turn, that_the dual local norm $\|_f_t \|^*_{\nabla F(\nu_t)}$ is expected to_be small.
Now_we are_ready to present the online-learning (i.e., steady-state) part of the regret_bound for the above algorithm:
\[thm:OMDbound1\] Let_$D_F({\cal G'}) \deq\sup_{\nu \in_{\cal G}'}_D_{F}(\ |
orig) – (2.5,0) \[->\] (2.5,0) |- (removeTwoPod22); (removeOnePod22) – (removePod22); (removeTwoPod22) – (removePod22); (orig) – (2.5,0) \[->\] (2.5,0) |- (removeOnePod1); (removeOnePod1) – (removeOnePod1B);
**Case 3 :** [$S_{1,1}$]{} $\sim_1$ [$S_{2,2,1}$]{}
Figure \[fig:pod11EQUIV1pod221\] shows the possible moves on [$S_{1,1}$]{} or [$S_{2,2,1}$]{}, and the answer leading to a ${\mathcal{P}}$-position (for readability, we write $S$ instead of $\hat{S}$ in the figure).
(orig) at (0,0) [ ]{};
(removeTwoPod1) at (5,0)
\(1) at (-1.1,1) ; (1b) at (-1.1,-1) ; (2) at (-2.1,0) ; (2b) at (-3.1,0) ; (2c) at (-3.65,0) [$S$]{}; (-3,0) – (-3.75,0.5); (-3,0) – (-3.75,-0.5); (-3.75,0.5) arc (135:225:0.7); (1) – (2); (2) – (2b); (2) – (1b);
(-0.6,0) node [+]{};
(1,0) – (0.25,0.5); (1,0) – (0.25,-0.5); (0.25,0.5) arc (135:225:0.7); (3b) at (0.35,0) [$S$]{}; (3) at (1,0) ; (4) at (2,0) ; (6) at (3,1) ; (5) at (3,-1) ; (5b) at (4,-1) ; (7) at (4,1) ; (8) at | orig) – (2.5,0) \[->\ ] (2.5,0) |- (removeTwoPod22); (removeOnePod22) – (removePod22); (removeTwoPod22) – (removePod22); (orig) – (2.5,0) \[->\ ] (2.5,0) |- (removeOnePod1); (removeOnePod1) – (removeOnePod1B);
* * Case 3 :* * [ $ S_{1,1}$ ] { } $ \sim_1 $ [ $ S_{2,2,1}$ ] { }
Figure \[fig: pod11EQUIV1pod221\ ] shows the possible moves on [ $ S_{1,1}$ ] { } or [ $ S_{2,2,1}$ ] { }, and the answer lead to a $ { \mathcal{P}}$-position (for legibility, we write $ S$ instead of $ \hat{S}$ in the name).
(orig) at (0,0) [ ] { };
(removeTwoPod1) at (5,0)
\(1) at (-1.1,1); (1b) at (-1.1,-1); (2) at (-2.1,0); (2b) at (-3.1,0); (2c) at (-3.65,0) [ $ S$ ] { }; (-3,0) – (-3.75,0.5); (-3,0) – (-3.75,-0.5); (-3.75,0.5) bow (135:225:0.7); (1) – (2); (2) – (2b); (2) – (1b);
(-0.6,0) node [ + ] { };
(1,0) – (0.25,0.5); (1,0) – (0.25,-0.5); (0.25,0.5) arc (135:225:0.7); (3b) at (0.35,0) [ $ S$ ] { }; (3) at (1,0); (4) at (2,0); (6) at (3,1); (5) at (3,-1); (5b) at (4,-1); (7) at (4,1); (8) at | orih) – (2.5,0) \[->\] (2.5,0) |- (removeTwoPod22); (reooveOnePod22) – (removePod22); (cemoveTsoPod22) – (rdmovePod22); (orig) – (2.5,0) \[->\] (2.5,0) |- (removeOiePoe1); (remiveOnePod1) – (removeOnePoa1B);
**Case 3 :** [$D_{1,1}$]{} $\sim_1$ [$S_{2,2,1}$]{}
Fugurt \[fig:pod11EQUIV1pod221\] sikws the possigpe mmtes on [$S_{1,1}$]{} or [$S_{2,2,1}$]{}, akd the answar leading to d ${\oachcal{P}}$-position (for readability, we wryte $S$ imshead of $\hat{S}$ ig tht fygurs).
(orig) at (0,0) [ ]{};
(removeTwoPod1) at (5,0)
\(1) at (-1.1,1) ; (1b) at (-1.1,-1) ; (2) et (-2.1,0) ; (2b) at (-3.1,0) ; (2c) ay (-3.65,0) [$S$]{}; (-3,0) – (-3.75,0.5); (-3,0) – (-3.75,-0.5); (-3.75,0.5) arc (135:225:0.7); (1) – (2); (2) – (2b); (2) – (1h);
(-0.6,0) nofe [+]{};
(1,0) – (0.25,0.5); (1,0) – (0.25,-0.5); (0.25,0.5) arc (135:225:0.7); (3b) ah (0.35,0) [$S$]{}; (3) at (1,0) ; (4) at (2,0) ; (6) at (3,1) ; (5) at (3,-1) ; (5b) at (4,-1) ; (7) au (4,1) ; (8) at | orig) – (2.5,0) \[->\] (2.5,0) |- (removeTwoPod22); (removePod22); – (removePod22); – (2.5,0) \[->\] (removeOnePod1B); 3 :** [$S_{1,1}$]{} [$S_{2,2,1}$]{} Figure \[fig:pod11EQUIV1pod221\] the possible moves on [$S_{1,1}$]{} or and the answer leading to a ${\mathcal{P}}$-position (for readability, we write $S$ instead $\hat{S}$ in the figure). (orig) at (0,0) [ ]{}; (removeTwoPod1) at (5,0) \(1) (-1.1,1) (1b) (-1.1,-1) (2) at (-2.1,0) ; (2b) at (-3.1,0) ; (2c) at (-3.65,0) [$S$]{}; (-3,0) – (-3.75,0.5); (-3,0) (-3.75,-0.5); (-3.75,0.5) arc (135:225:0.7); (1) – (2); (2) (2b); (2) – (1b); node [+]{}; (1,0) – (0.25,0.5); – (0.25,0.5) arc (3b) (0.35,0) (3) at (1,0) (4) at (2,0) ; (6) at (3,1) ; (5) at (3,-1) ; (5b) at (4,-1) ; (7) (4,1) ; | orig) – (2.5,0) \[->\] (2.5,0) |- (removeTwoPod22); (removeONePod22) – (removEPod22); (rEmoVeTWopod22) – (rEmovEPod22); (orig) – (2.5,0) \[->\] (2.5,0) |- (remoVEOnepod1); (removeOnePod1) – (removeONePod1b);
**CASe 3 :** [$S_{1,1}$]{} $\sIM_1$ [$S_{2,2,1}$]{}
figurE \[fig:pod11eqUiv1Pod221\] ShOwS thE pOSsIble mOveS on [$S_{1,1}$]{} or [$S_{2,2,1}$]{}, And the answEr lEaDing to a ${\mathcAL{P}}$-Position (foR reAdability, we wRitE $S$ instEaD of $\HAt{S}$ in The FigurE).
(orig) aT (0,0) [ ]{};
(RemovetwoPod1) at (5,0)
\(1) aT (-1.1,1) ; (1b) AT (-1.1,-1) ; (2) at (-2.1,0) ; (2b) at (-3.1,0) ; (2C) At (-3.65,0) [$S$]{}; (-3,0) – (-3.75,0.5); (-3,0) – (-3.75,-0.5); (-3.75,0.5) arc (135:225:0.7); (1) – (2); (2) – (2b); (2) – (1B);
(-0.6,0) NOdE [+]{};
(1,0) – (0.25,0.5); (1,0) – (0.25,-0.5); (0.25,0.5) arc (135:225:0.7); (3B) at (0.35,0) [$S$]{}; (3) at (1,0) ; (4) at (2,0) ; (6) at (3,1) ; (5) at (3,-1) ; (5b) at (4,-1) ; (7) at (4,1) ; (8) AT | orig) – (2.5,0) \[->\](2.5,0) |- (rem ove Two Po d22) ; (r emoveOnePod22) – (r emovePod22); (removeTw oPod2 2) – (r e mo vePod 22); (o r ig ) – ( 2. 5, 0)\[ - &g t;\](2. 5,0) |- (removeOn ePo d1 ); (removeOn e Po d1) – (rem ove OnePod1B);
**C ase 3:* * [ $ S_{1, 1}$ ]{} $ \sim_1 $ [$S_{ 2,2,1}$]{ }Figure \[fig:p o d 11 EQUI V1pod221\] showst he possible moves on [$ S_ { 1, 1 } $]{ } o r [$S_{2,2 ,1 }$]{} , and th e a n s w erl eading to a $ {\mathcal{P } }$- positi on (f o r read abili ty , we write $S$inst ead of $\ hat{S} $ in the figure) .
(or ig) at (0, 0 )[]{} ;(re m ov eTw o Pod 1) at (5 ,0 )
\(1) at( - 1 . 1,1) ;(1b) at ( -1.1,-1) ; (2 ) a t (- 2 .1, 0) ;(2b)at ( -3 .1,0) ; (2c ) at(- 3.65,0) [$S$]{} ; (- 3,0) – (- 3.7 5, 0.5 ); (-3, 0 ) – (- 3.7 5,- 0.5); ( -3.75,0 . 5)ar c ( 13 5:225:0.7); (1) –(2 ) ; ( 2) – (2b ); (2) –(1 b );
(-0. 6, 0)node [ +]{};
(1 , 0) – (0.25 ,0.5); (1 ,0 ) – (0. 25 ,-0.5) ;(0. 25, 0.5)a rc ( 135:22 5:0.7);(3b)a t (0.35,0) [$S $ ]{}; (3) at ( 1 ,0 ) ;( 4) a t ( 2,0) ; (6)at ( 3 ,1); (5 ) a t ( 3 ,-1); (5b )a t( 4,-1) ; (7) at (4,1 ); (8)at | orig) –_(2.5,0) \[->\]_(2.5,0) |- (removeTwoPod22); (removeOnePod22)_– (removePod22);_(removeTwoPod22)_– (removePod22);_(orig)_– (2.5,0) \[->\]_(2.5,0) |- (removeOnePod1);_(removeOnePod1) – (removeOnePod1B);
**Case 3_:** [$S_{1,1}$]{} $\sim_1$_[$S_{2,2,1}$]{}
Figure \[fig:pod11EQUIV1pod221\]_shows the possible moves on [$S_{1,1}$]{} or [$S_{2,2,1}$]{}, and the answer leading to a_${\mathcal{P}}$-position_(for readability,_we_write_$S$ instead of $\hat{S}$ in_the figure).
(orig) at (0,0) [_]{};
(removeTwoPod1) at_(5,0)
\(1) at (-1.1,1) ; (1b) at (-1.1,-1) ;_(2)_at (-2.1,0) ;_(2b) at (-3.1,0) ; (2c) at (-3.65,0) [$S$]{}; (-3,0)_– (-3.75,0.5); (-3,0) – (-3.75,-0.5); (-3.75,0.5)_arc (135:225:0.7); (1)_–_(2);_(2) – (2b); (2)_– (1b);
(-0.6,0) node [+]{};
(1,0) – (0.25,0.5);_(1,0) – (0.25,-0.5); (0.25,0.5) arc (135:225:0.7);_(3b) at (0.35,0) [$S$]{}; (3) at (1,0)_; (4) at (2,0) ; (6)_at (3,1) ; (5) at_(3,-1) ;_(5b) at (4,-1) ; (7)_at (4,1) ;_(8) at |
>\!0$ (resp. $0$) in ${\mathbb{C}}_p$. Observe that $f'(\zeta)\!=\!a_2a_3\zeta^{a_2-1}(\zeta^{a_3-a_2}-1)$.
For $\zeta\!\in\!{\mathbb{C}}_p$ a root of $f$ with valuation $\frac{{{\mathrm{ord}}}_p (a_3-a_2)}{a_2}$ we then obtain\
${{\mathrm{ord}}}_p f'(\zeta)=
\frac{a_2-1}{a_2}{{\mathrm{ord}}}_p(a_3-a_2)=O({\mathrm{size}}(f))$.\
In other words, we can simply apply Hensel’s Lemma to $f$ and use a root of $f$ in $p^{{{\mathrm{ord}}}_p(a_3-a_2)/a_2}({\mathbb{Z}}/p^{2{{\mathrm{ord}}}_p(a_3-a_2)+1}{\mathbb{Z}})$ as a certificate for a non-degenerate root of $f$ in ${\mathbb{Q}}_p$.
For $\zeta\!\in\!{\mathbb{C}}_p$ a root of $f$ with valuation $0$ we then obtain ${{\mathrm{ord}}}_p f'(\zeta)\!\geq\!{{\mathrm{ord}}}_p(a_3-a_2)$, thanks to identity ($\star$) from the non-degenerate case. Note also that $r'(\zeta)\!=\!\frac{f'(\zeta)}{(\zeta-1)^2}-2\frac{f(\zeta)}
{(\zeta-1)^3}\!=\!\frac{f'(\zeta)}{(\zeta-1)^2}$. Employing the product formula from Lemma \[lemma:syl\] we then obtain\
${{\mathrm{ord}}}_p \Delta \!=\!
\left(\sum\limits_{r(\zeta)=0} {{\mathrm{ord}}}_p f'(\zeta)\right)-2{{\mathrm{ord}}}_p
\prod\limits_{r(\zeta)=0} (\zeta-1)\!=\!\left(\sum\limits_{r(\zeta)=0}
{{\mathrm{ord}}}_p f'(\zeta)\right)-2{{\mathrm{ord}}}_p r(1)$\
since. From our proof of Part (4) of Corollary | > \!0 $ (resp. $ 0 $) in $ { \mathbb{C}}_p$. Observe that $ f'(\zeta)\!=\!a_2a_3\zeta^{a_2 - 1}(\zeta^{a_3 - a_2}-1)$.
For $ \zeta\!\in\!{\mathbb{C}}_p$ a root of $ f$ with valuation $ \frac{{{\mathrm{ord}}}_p (a_3 - a_2)}{a_2}$ we then obtain\
$ { { \mathrm{ord}}}_p f'(\zeta)=
\frac{a_2 - 1}{a_2}{{\mathrm{ord}}}_p(a_3 - a_2)=O({\mathrm{size}}(f))$.\
In other actor's line, we can plainly apply Hensel ’s Lemma to $ f$ and use a root of $ f$ in $ p^{{{\mathrm{ord}}}_p(a_3 - a_2)/a_2}({\mathbb{Z}}/p^{2{{\mathrm{ord}}}_p(a_3 - a_2)+1}{\mathbb{Z}})$ as a security for a non - degenerate root of $ f$ in $ { \mathbb{Q}}_p$.
For $ \zeta\!\in\!{\mathbb{C}}_p$ a root of $ f$ with evaluation $ 0 $ we then prevail $ { { \mathrm{ord}}}_p f'(\zeta)\!\geq\!{{\mathrm{ord}}}_p(a_3 - a_2)$, thanks to identity ($ \star$) from the non - degenerate event. Note also that $ r'(\zeta)\!=\!\frac{f'(\zeta)}{(\zeta-1)^2}-2\frac{f(\zeta) }
{ (\zeta-1)^3}\!=\!\frac{f'(\zeta)}{(\zeta-1)^2}$. hire the product formula from Lemma \[lemma: syl\ ] we then obtain\
$ { { \mathrm{ord}}}_p \Delta \!=\!
\left(\sum\limits_{r(\zeta)=0 } { { \mathrm{ord}}}_p f'(\zeta)\right)-2{{\mathrm{ord}}}_p
\prod\limits_{r(\zeta)=0 } (\zeta-1)\!=\!\left(\sum\limits_{r(\zeta)=0 }
{ { \mathrm{ord}}}_p f'(\zeta)\right)-2{{\mathrm{ord}}}_p r(1)$\
since. From our validation of Part (4) of Corollary | >\!0$ (redp. $0$) in ${\mathbb{C}}_p$. Observe tmat $f'(\zeta)\!=\!a_2a_3\zeta^{a_2-1}(\eera^{a_3-a_2}-1)$.
Foc $\zeta\!\ih\!{\mathbb{C}}_o$ a root of $f$ with valuation $\feac{{{\maukrm{ord}}}_p (a_3-a_2)}{a_2}$ we then ubtain\
${{\matjrm{ord}}}_p d'(\zete)=
\frac{a_2-1}{a_2}{{\mathrm{ord}}}_'(z_3-a_2)=O({\mathvi{sizs}}(n))$.\
In ochxr words, we can simply ap[ly Hensel’s Lekmx co $f$ and use a root of $f$ in $p^{{{\mathrm{owd}}}_p(a_3-a_2)/a_2}({\mstjbb{Z}}/p^{2{{\mathrm{ord}}}_k(a_3-a_2)+1}{\mauhbf{Z}})$ aa a certificate for a non-degenerats root mf $f$ in ${\mathbn{Q}}_p$.
For $\zeta\!\in\!{\mathbb{C}}_p$ a rolt ov $f$ with valuation $0$ we then ovtaig ${{\mathrm{ord}}}_p w'(\zeta)\!\geq\!{{\mauhxm{ord}}}_p(a_3-a_2)$, thznks to identity ($\star$) from the von-deyenerate cawe. Nohg also that $c'(\zeta)\!=\!\fgac{f'(\zeta)}{(\zeta-1)^2}-2\fvsc{f(\zetd)}
{(\zeta-1)^3}\!=\!\frsc{f'(\zeta)}{(\zeta-1)^2}$. Emiloyiig tye product formula frmm Lemma \[lemma:syl\] re then ottcin\
${{\mathrm{ord}}}_p \Delta \!=\!
\oedt(\sum\nimids_{r(\zdra)=0} {{\oatgrk{odd}}}_p f'(\zfta)\cight)-2{{\mathrm{krd}}}_p
\prod\linits_{r(\zeta)=0} (\zeta-1)\!=\!\left(\sim\jpkits_{r(\zeta)=0}
{{\mafhrm{orq}}}_p f'(\zeta)\right)-2{{\mathrm{ord}}}_p r(1)$\
since. From our pgoof of Part (4) of Corollary | >\!0$ (resp. $0$) in ${\mathbb{C}}_p$. Observe that $\zeta\!\in\!{\mathbb{C}}_p$ root of with valuation $\frac{{{\mathrm{ord}}}_p f'(\zeta)= In other words, can simply apply Lemma to $f$ and use a of $f$ in $p^{{{\mathrm{ord}}}_p(a_3-a_2)/a_2}({\mathbb{Z}}/p^{2{{\mathrm{ord}}}_p(a_3-a_2)+1}{\mathbb{Z}})$ as a certificate for a non-degenerate root of $f$ ${\mathbb{Q}}_p$. For $\zeta\!\in\!{\mathbb{C}}_p$ a root of $f$ with valuation $0$ we then obtain f'(\zeta)\!\geq\!{{\mathrm{ord}}}_p(a_3-a_2)$, to ($\star$) the non-degenerate case. Note also that $r'(\zeta)\!=\!\frac{f'(\zeta)}{(\zeta-1)^2}-2\frac{f(\zeta)} {(\zeta-1)^3}\!=\!\frac{f'(\zeta)}{(\zeta-1)^2}$. Employing the product formula from Lemma \[lemma:syl\] we obtain\ ${{\mathrm{ord}}}_p \Delta \!=\! \left(\sum\limits_{r(\zeta)=0} {{\mathrm{ord}}}_p f'(\zeta)\right)-2{{\mathrm{ord}}}_p \prod\limits_{r(\zeta)=0} {{\mathrm{ord}}}_p f'(\zeta)\right)-2{{\mathrm{ord}}}_p r(1)$\ since. our proof of Part (4) Corollary | >\!0$ (resp. $0$) in ${\mathbb{C}}_p$. Observe that $F'(\zeta)\!=\!a_2a_3\zetA^{a_2-1}(\zetA^{a_3-a_2}-1)$.
for $\ZeTa\!\in\!{\MathBb{C}}_p$ a root of $f$ wiTH valUation $\frac{{{\mathrm{ord}}}_p (a_3-a_2)}{A_2}$ we thEn OBtaiN\
${{\MaThrm{oRd}}}_p f'(\zetA)=
\FrAC{A_2-1}{a_2}{{\mAtHrM{orD}}}_p(A_3-A_2)=O({\MathrM{siZe}}(f))$.\
In otHer words, we Can SiMply apply HenSEl’S Lemma to $f$ aNd uSe a root of $f$ in $P^{{{\maThrm{orD}}}_p(A_3-a_2)/a_2}({\MAthbb{z}}/p^{2{{\mAthrm{Ord}}}_p(a_3-a_2)+1}{\MAthbb{Z}})$ As a certifIcATe for a NOn-degenERAtE rooT of $f$ in ${\mathbb{Q}}_p$.
For $\ZEtA\!\In\!{\mathbb{C}}_p$ a rooT of $f$ wiTh VAlUATioN $0$ we Then obtain ${{\MaThrm{oRD}}}_p f'(\zeta)\!\GEq\!{{\MATHrm{ORd}}}_p(a_3-a_2)$, thanks to Identity ($\staR$) FroM the noN-dEgeNErate cAse. NoTe ALso That $r'(\zeta)\!=\!\frAc{f'(\zEta)}{(\zeta-1)^2}-2\frAc{f(\zetA)}
{(\Zeta-1)^3}\!=\!\fraC{F'(\zeta)}{(\zeTa-1)^2}$. EmplOyiNg tHe prODuCt ForMuLA frOM LEmmA \[LemMa:syl\] we tHeN oBtain\
${{\MathRM{ORD}}}_p \DeLta \!=\!
\Left(\Sum\liMits_{r(\zeta)=0} {{\mathRm{oRd}}}_p f'(\ZEta)\Right)-2{{\MathrM{ord}}}_P
\pRod\liMits_{r(\zEta)=0} (\zeTa-1)\!=\!\Left(\sum\limits_{r(\zEta)=0}
{{\mAthrm{ord}}}_p F'(\zeTa)\RigHt)-2{{\MathrM{Ord}}}_p r(1)$\
sIncE. FrOm our prOof of PaRT (4) of coROLLaRy | >\!0$ (resp. $0$) in ${\ma thbb{C}}_p $. Ob ser veth at $ f'(\ zeta)\!=\!a_2a _ 3\ze ta^{a_2-1}(\zeta^{a_3- a_2}- 1) $ .
F o r$\zet a\!\in\ ! {\ m a thb b{ C} }_p $a r oot o f $ f$ with valuation $\ fr ac{{{\mathrm { or d}}}_p (a_ 3-a _2)}{a_2}$ w e t hen ob ta in\ ${{\m ath rm{or d}}}_p f'(\ze ta)=
\fra c{ a _2-1}{ a _2}{{\m a t hr m{or d}}}_p(a_3-a_2)=O ( {\ m athrm{size}}(f ))$.\In ot h e r w ord s, we cansi mplya pply He n se l ’ s Le m ma to $f$ and use a root of$f$ in $ p^{ { {\math rm{or d} } }_p (a_3-a_2)/a _2}( {\mathbb{ Z}}/p^ { 2{{\mat h rm{ord} }}_p(a _3- a_2 )+1} { \m at hbb {Z } })$ as ac ert ificatefo ra non -deg e n e r ateroo t of $f$in ${\mathbb{ Q}} _p$.
Fo r $\z eta\! \in\ !{ \math bb{C}} _p$ a r oot of $f$ with val uation $0 $ w ethe nobtai n ${{\m ath rm{ ord}}}_ p f'(\z e ta) \! \ g e q\ !{{\mathrm{ord}}}_ p( a _ 3- a_2)$, t hankst oid e ntity ($ \s tar $) f r o m the non - de generate case. No te also t ha t $r'( \z eta )\! =\!\f r ac{f '(\zet a)}{(\ze ta-1) ^ 2}-2\frac{f(\z e ta)}
{(\zeta- 1 )^ 3 } \! = \!\f rac {f'(\zeta)} {(\z e ta-1 )^2} $ .Emp l oying thepr o du c t formula from Lemm a\[lemm a:syl \] we then ob tain\
${{\ m a t hrm{ord} }}_p \D e lta \!=\!
\le ft(\s um\limits_ { r(\zeta) =0} { {\mathrm {ord}}}_p f '(\zeta) \ri ght )-2 {{\ m a th rm{ord}}}_p \ p rod\ li mits_{r (\z eta)=0} (\ zet a-1 )\! =\ !\left(\s um\limit s_ {r (\ ze ta) =0} { {\mathrm {o rd} }} _pf'(\z e ta)\ri ght)- 2{{\ ma th r m{o rd}}}_p r( 1 ) $\
s in ce . Fr omou r pro of o f Pa rt (4)of Coroll ary | >\!0$ (resp. $0$)_in ${\mathbb{C}}_p$._Observe that $f'(\zeta)\!=\!a_2a_3\zeta^{a_2-1}(\zeta^{a_3-a_2}-1)$.
For $\zeta\!\in\!{\mathbb{C}}_p$_a root_of_$f$ with_valuation_$\frac{{{\mathrm{ord}}}_p (a_3-a_2)}{a_2}$ we_then obtain\
${{\mathrm{ord}}}_p f'(\zeta)=
\frac{a_2-1}{a_2}{{\mathrm{ord}}}_p(a_3-a_2)=O({\mathrm{size}}(f))$.\
In_other words, we can_simply apply Hensel’s_Lemma_to $f$ and use a root of $f$ in $p^{{{\mathrm{ord}}}_p(a_3-a_2)/a_2}({\mathbb{Z}}/p^{2{{\mathrm{ord}}}_p(a_3-a_2)+1}{\mathbb{Z}})$ as a certificate for_a_non-degenerate root_of_$f$_in ${\mathbb{Q}}_p$.
For $\zeta\!\in\!{\mathbb{C}}_p$ a root_of $f$ with valuation $0$_we then_obtain ${{\mathrm{ord}}}_p f'(\zeta)\!\geq\!{{\mathrm{ord}}}_p(a_3-a_2)$, thanks to identity ($\star$) from_the_non-degenerate case. Note_also that $r'(\zeta)\!=\!\frac{f'(\zeta)}{(\zeta-1)^2}-2\frac{f(\zeta)}
{(\zeta-1)^3}\!=\!\frac{f'(\zeta)}{(\zeta-1)^2}$. Employing the product formula from Lemma_\[lemma:syl\] we then obtain\
${{\mathrm{ord}}}_p \Delta \!=\!_
\left(\sum\limits_{r(\zeta)=0} {{\mathrm{ord}}}_p f'(\zeta)\right)-2{{\mathrm{ord}}}_p_
\prod\limits_{r(\zeta)=0}_(\zeta-1)\!=\!\left(\sum\limits_{r(\zeta)=0}_
{{\mathrm{ord}}}_p f'(\zeta)\right)-2{{\mathrm{ord}}}_p r(1)$\
since. From_our proof of Part (4) of_Corollary |
of the two clicked detectors is in an even-parity polarization mode and an even-parity spatial mode, the state of the photon pair $AB$ is projected to the maximally hyperentangled Bell state $|\phi\rangle_{AB}=\frac{1}{2}(|HH\rangle+|VV\rangle)_{AB}
(|a_1b_1\rangle+|a_2b_2\rangle)$. If the outcome of the two clicked detectors is in an odd-parity polarization (spatial) mode, a phase-flip operation $\sigma^P_z$ ($\sigma^S_z$) on photon $B$ is required to obtain the state $|\phi\rangle_{AB}$.
In principle, if Alice and Bob both have only one detector clicked, the maximally hyperentangled Bell states can be obtained with the probability of $P=4|\alpha\beta\gamma\delta|^2$. Otherwise, the hyper-ECP fails. In practical, whether this hyper-ECP succeeds or not can also be read out by postselection, if the efficiencies of the single-photon detectors are lower than 100 %. With this method, the maximally hyperentangled GHZ states can be obtained from the partially hyperentangled GHZ states [@HECP2] in the same way.
### Hyper-ECP with nonlinear optical elements {#sec3.1.3}
Here, we introduce the hyper-ECP with nonlinear optical elements. That is, the polarization parity-check measurement and the spatial-mode parity-check measurement are replaced by the polarization parity-check QND (P-QND) and the spatial-mode parity-check QND (S-QND), respectively, where the P-QND and S-QND are constructed with the cross-Kerr nonlinearity.
![Schematic diagram of the polarization parity-check quantum nondemolition detector (P-QND) with the cross-Kerr nonlinearity [@HECPLixhOE].[]{data-label="figure3.3"}](fig15_P-QND.eps){width="7.2cm"}
*P-QND* — The setup of the P-QND is shown in Fig. \[figure3.3\], which is different from the one in Fig. \[HBSA2\]. In Fig. \[HBSA2\], the spatial-mode state is detected at the | of the two clicked detectors is in an even - parity polarization mood and an tied - parity spatial mode, the department of state of the photon couple $ AB$ is projected to the maximally hyperentangled Bell state $ |\phi\rangle_{AB}=\frac{1}{2}(|HH\rangle+|VV\rangle)_{AB }
(|a_1b_1\rangle+|a_2b_2\rangle)$. If the outcome of the two click detectors is in an odd - parity bit polarization (spatial) mood, a phase - flip mathematical process $ \sigma^P_z$ ($ \sigma^S_z$) on photon $ B$ is required to obtain the country $ |\phi\rangle_{AB}$.
In rationale, if Alice and Bob both have only one detector chatter, the maximally hyperentangled Bell states can be obtained with the probability of $ P=4|\alpha\beta\gamma\delta|^2$. differently, the hyper - ECP fails. In practical, whether this hyper - ECP succeeds or not can besides be read out by postselection, if the efficiencies of the single - photon detectors are lower than 100% . With this method acting, the maximally hyperentangled GHZ states can be receive from the partially hyperentangled GHZ states [ @HECP2 ] in the same way.
# # # Hyper - ECP with nonlinear ocular elements { # sec3.1.3 }
Here, we introduce the hyper - ECP with nonlinear optical elements. That is, the polarization parity - check mark measurement and the spatial - manner parity - check measurement are replaced by the polarization parity - check QND (P - QND) and the spatial - mode parity bit - check QND (S - QND), respectively, where the P - QND and S - QND are construct with the cross - Kerr nonlinearity.
! [ conventional diagram of the polarization parity bit - check quantum nondemolition detector (P - QND) with the cross - Kerr nonlinearity [ @HECPLixhOE].[]{data - label="figure3.3"}](fig15_P - QND.eps){width="7.2 cm " }
* P - QND * — The setup of the P - QND is indicate in Fig. \[figure3.3\ ], which is different from the one in Fig. \[HBSA2\ ]. In Fig. \[HBSA2\ ], the spatial - mode state is detected at the | of the two clicked detectovs is in an even-kaeity pmlarizztion moae and an even-parity spatial mide, tye state of the photon pair $AB$ ps projecred uo the maximally igperentangled Nell vvate $|\phi\rangle_{AN}=\frac{1}{2}(|HH\rangne+|VV\rangle)_{AB}
(|a_1b_1\savgpe+|a_2b_2\rangle)$. If the outcome of the twj clickrd detectors is yn am odd-larity polarization (spatial) mode, a phase-fnip operation $\sigma^P_z$ ($\sigma^S_z$) on photon $B$ id required to obtaln the statg $|\phy\eangle_{AB}$.
In pfinciple, if Alice and Gob both have only one detector clicled, the mazinalpi hyperentanjled Bvll states cak be obdained eith the probanilitb of $P=4|\alpha\beta\gamma\delta|^2$. Otherwise, the hypgr-ECP failv. Nn practical, whether rhus hyker-EC[ suzxeeas kr nkt can alao be read out by posrselection, if the egfyboencies of tge sindlq-photon detectors are lower than 100 %. With thjs method, the maximally hyperentangled GHZ sjates can fe obtained from the partially hyperentangled GHZ statxs [@HEEI2] kb hhe same way.
### Hyper-ECP with nonlinear optical qmekekts {#sec3.1.3}
Here, we iktroduce the hyper-RCO eyth nonlinear opticcm slements. That is, tje polatizatiin parity-shecl measurement and the spatiql-mode paritj-chexk measurement are replaced bv the kolariaation parity-check QND ('-QND) ahd the spatlal-mode pzfity-check QND (S-QVD), gespactively, where the P-QND anq S-QND arx conxtructea wijh the sross-Kerr jonlikaarity.
![Schematic diwgram ox the polagization parity-check quantum noivemolition dejecdor (P-QND) wich the cross-Kerr noglinearity [@HECKLixhOE].[]{daca-labeu="figure3.3"}](fig15_I-QND.eps){wivth="7.2cm"}
*P-QND* — Tre setup of tvg P-QND is shoxn in Fig. \[siguee3.3\], wyich is aifferent from the one pn Fig. \[YBSA2\]. In Fig. \[HBSA2\], thc spajizl-mode state is deuexted at the | of the two clicked detectors is in polarization and an spatial mode, the $AB$ projected to the hyperentangled Bell state (|a_1b_1\rangle+|a_2b_2\rangle)$. If the outcome of the clicked detectors is in an odd-parity polarization (spatial) mode, a phase-flip operation $\sigma^P_z$ on photon $B$ is required to obtain the state $|\phi\rangle_{AB}$. In principle, if and both only detector clicked, the maximally hyperentangled Bell states can be obtained with the probability of $P=4|\alpha\beta\gamma\delta|^2$. Otherwise, hyper-ECP fails. In practical, whether this hyper-ECP succeeds not can also be out by postselection, if the of single-photon detectors lower 100 With this method, maximally hyperentangled GHZ states can be obtained from the partially hyperentangled GHZ states [@HECP2] in the same ### Hyper-ECP optical elements Here, introduce hyper-ECP with nonlinear That is, the polarization parity-check measurement parity-check measurement are replaced by the polarization parity-check (P-QND) and spatial-mode parity-check QND (S-QND), respectively, where P-QND and S-QND are constructed with the cross-Kerr ![Schematic diagram of the polarization parity-check quantum nondemolition detector (P-QND) with the cross-Kerr nonlinearity [@HECPLixhOE].[]{data-label="figure3.3"}](fig15_P-QND.eps){width="7.2cm"} The setup of the is shown in \[figure3.3\], is from one in \[HBSA2\]. In Fig. \[HBSA2\], the spatial-mode state is detected at the | of the two clicked detectors iS in an even-pArity PolAriZaTion Mode And an even-paritY SpatIal mode, the state of the phOton pAiR $aB$ is PRoJecteD to the mAXiMALly HyPeRenTaNGlEd BelL stAte $|\phi\rAngle_{AB}=\fraC{1}{2}(|HH\RaNgle+|VV\rangle)_{ab}
(|a_1B_1\rangle+|a_2b_2\rAngLe)$. If the outcoMe oF the twO cLicKEd detEctOrs is In an odD-Parity PolarizatIoN (SpatiaL) Mode, a phASE-fLip oPeration $\sigma^P_z$ ($\siGMa^s_Z$) on photon $B$ is reQuired To OBtAIN thE stAte $|\phi\rangLe_{aB}$.
In pRInciple, IF ALICE anD bob both have onLy one detectOR clIcked, tHe MaxIMally hYpereNtANglEd Bell stateS can Be obtaineD with tHE probabILity of $P=4|\Alpha\bEta\GamMa\deLTa|^2$. otHerWiSE, thE HyPer-ecP fAils. In prAcTiCal, whEtheR THIS hypEr-EcP suCceedS or not can also Be rEad oUT by PostsElectIon, iF tHe effIciencIes of ThE single-photon deTectOrs are lowEr tHaN 100 %. WiTh This mEThod, thE maXimAlly hypErentanGLed gHz STAtEs can be obtained froM tHE PaRtially hYperenTAnGlED GHZ statEs [@hECp2] in tHE Same wAy.
### HyPEr-eCP with nOnlineAR oPtIcal eleMeNts {#sec3.1.3}
heRe, wE inTroduCE the Hyper-EcP with noNlineAR optical elemenTS. That is, the polARiZATiON parIty-Check measurEmenT And tHe spATiAl-mODe parIty-chEcK MeASurement are replaced By The polArizaTion parity-cheCk QND (P-QND) aND THe spatiaL-modE PaRIty-check QND (S-QNd), respEctively, whERe the P-QNd and S-qND are coNstructed WITh the croSs-KErr NonLinEARiTy.
![Schematic diAGRam oF tHe polarIzaTion parIty-CheCk qUanTuM nondemolItion detEcToR (P-qNd) wiTh the CRoss-Kerr NoNliNeAriTy [@HECplixhOE].[]{Data-lAbel="FiGuRE3.3"}](fiG15_P-QND.epS){WiDTH="7.2cm"}
*P-qNd* — THe seTup Of The P-QnD is SHowN in Fig. \[fIgure3.3\], whicH is DIffeReNt From the One in Fig. \[HBSA2\]. IN FIg. \[HBSA2\], the sPaTiaL-mode sTATe is deteCted at the | of the two clicked detect ors is inan ev en- par it y po lari zation mode an d aneven-parity spatial mo de, t he stat e o f the photon pa i r $A B$ i s p ro j ec ted t o t he maxi mally hype ren ta ngled Bell s t at e $|\phi\r ang le_{AB}=\fra c{1 }{2}(| HH \ra n gle+| VV\ rangl e)_{AB }
(|a_1 b_1\rangl e+ | a_2b_2 \ rangle) $ . I f th e outcome of thet wo clicked detect ors is i n a n odd -pa rity polar iz ation (spatia l )m o d e,a phase-flip o peration $\ s igm a^P_z$ ( $\s i gma^S_ z$) o np hot on $B$ is r equi red to ob tain t h e state $|\phi\ rangle _{A B}$ .
I n p ri nci pl e , i f A lic e an d Bob bo th h ave o nlyo n e dete cto r cl icked , the maximal lyhype r ent angle d Bel l st at es ca n be o btain ed with the proba bili ty of $P= 4|\ al pha \b eta\g a mma\de lta |^2 $. Othe rwise,t hehy p e r -E CP fails. In pract ic a l ,whetherthis h y pe r- E CP succe ed s o r no t can a lsob eread out by po s ts el ection, i f theef fic ien cieso f th e sing le-photo n det e ctors are lowe r than 100 %.W it h th i s me tho d, the maxi mall y hyp eren t an gle d GHZstate sc an be obtained from th epartia lly h yperentangled GHZ state s [ @HECP2]in t h es ame way.
###Hyper -ECP withn onlinear opti cal elem ents {#se c 3 .1.3}
H ere , w e i ntr o d uc e the hyper-E C P wit hnonline aroptical el eme nts . T ha t is, the polariz at io npa rit y-che c k measur em ent a ndthe s p atial- modepari ty -c h eck measur e me n t are r ep lace d b ythe p olar i zat ion par ity-check QN D (P- QN D) and th e spatial-mod eparity-che ck QN D (S-Q N D ), respe ctively, where the P-QN D and S- QND arecons tructed w ith the c ros s -Kerrnonlin earit y.
! [ S chema t i cdia gr am of thep o lar izati on par ity-che ck quantum nondemo l iti on detector ( P-Q ND)w i th th e c r oss -K e rrn o nlinearity [@HE CPLixhOE]. [] { da ta-label=" f igu re 3.3"}]( fig15_P -QND. e ps){wid th="7.2cm "}
*P-QN D* — T h e se tup of the P-QND i s shown i n Fig. \[ figur e3. 3\], w hi chis di fferen t fr om th e onein Fig.\[HBS A2 \]. In F ig. \[HBSA2\], the spat ial-mo de st ate is detec ted atthe | of_the two_clicked detectors is in_an even-parity_polarization_mode and_an_even-parity spatial mode,_the state of_the photon pair $AB$_is projected to_the_maximally hyperentangled Bell state $|\phi\rangle_{AB}=\frac{1}{2}(|HH\rangle+|VV\rangle)_{AB}
(|a_1b_1\rangle+|a_2b_2\rangle)$. If the outcome of the two clicked detectors is_in_an odd-parity_polarization_(spatial)_mode, a phase-flip operation $\sigma^P_z$_($\sigma^S_z$) on photon $B$ is_required to_obtain the state $|\phi\rangle_{AB}$.
In principle, if Alice and_Bob_both have only_one detector clicked, the maximally hyperentangled Bell states can_be obtained with the probability of_$P=4|\alpha\beta\gamma\delta|^2$. Otherwise, the_hyper-ECP_fails._In practical, whether this_hyper-ECP succeeds or not can also_be read out by postselection, if_the efficiencies of the single-photon detectors are_lower than 100 %. With this_method, the maximally hyperentangled GHZ_states can_be obtained from the partially_hyperentangled GHZ states_[@HECP2] in_the same way.
###_Hyper-ECP with nonlinear optical elements {#sec3.1.3}
Here,_we introduce the_hyper-ECP with nonlinear optical elements. That_is,_the polarization parity-check_measurement_and_the spatial-mode_parity-check measurement are_replaced_by the_polarization_parity-check QND (P-QND) and the spatial-mode_parity-check_QND (S-QND), respectively, where the P-QND and_S-QND are constructed with_the_cross-Kerr nonlinearity.
![Schematic diagram of_the polarization parity-check quantum nondemolition_detector (P-QND) with the cross-Kerr nonlinearity_[@HECPLixhOE].[]{data-label="figure3.3"}](fig15_P-QND.eps){width="7.2cm"}
*P-QND* —_The setup_of the P-QND is shown in Fig. \[figure3.3\], which is different from_the one in Fig. \[HBSA2\]. In Fig. \[HBSA2\],_the spatial-mode state is_detected at_the |
\max\left\{ R, (1-R) \right\}^{\gamma/2}, \quad
{e_{\gamma}^{{lb}}(R)}= \min\{ R, (1-R) \}^{\gamma/2}, \end{gathered}$$ or with $${d_{\gamma}^{{lb}}(r)}= r^{\gamma/2} (1-r)^{\gamma/2}, \quad {d_{\gamma}^{{ub}}(r)}=1,\quad
{e_{\gamma}^{{ub}}(R)}= 2^{1-\gamma/2}, \quad
{e_{\gamma}^{{lb}}(R)}= R^{\gamma/2} (1-R)^{\gamma},$$ as shown in \[Sec Model 3 upper\] and \[Sec Model 3 lower\].
This Model 3 for the transition function $\mathcal{B}$ is of particular importance for establishing macroscopic moments models starting from the Boltzmann equation. Namely, in [@MPC-Dj-S] it is shown that the six moments model with this transition function provides a source term which satisfies the macroscopic residual inequality on the whole range of six moments model validity, that provides the total agreement with the extended thermodynamics theory of six moments, as one of the rare systems admitting a non-linear or *far from equilibrium* closure of the governing equations using the entropy principle. Moreover, for this model 3 of the transition function $\mathcal{B}$, the macroscopic fourteen moments model achieves the matching with experimental data as well. More precisely, [@MPC-Dj-S] shows that the model 3 yields transport coefficients (shear and bulk viscosities and heat conductivity) that provide both experimentally measured dependence of the shear viscosity upon temperature and the Prantdtl number which coincides with its theoretical value at a satisfactory level. This remarkable success lies in the new *additive* form of the transition function that we propose in, instead of the multiplicative one used so far in the literature.
Fundamental lemmas {#Sec: fund lemmas}
==================
In this Section we prove two fundamental lemmas, that should be used sequentially, as they are presented.\
All of them are motivated by the search of a proof showing that $k$-th polynomial moment of the solution will satisfy an ODI with a negative superlinear term, that is $$\frac{\mathrm{d} }{\mathrm{d}t} \mathfrak{m}_k[f](t)= \int_{ \Delta} Q(f,f)( | \max\left\ { R, (1 - R) \right\}^{\gamma/2 }, \quad
{ e_{\gamma}^{{lb}}(R)}= \min\ { R, (1 - R) \}^{\gamma/2 }, \end{gathered}$$ or with $ $ { d_{\gamma}^{{lb}}(r)}= r^{\gamma/2 } (1 - r)^{\gamma/2 }, \quad { d_{\gamma}^{{ub}}(r)}=1,\quad
{ e_{\gamma}^{{ub}}(R)}= 2^{1-\gamma/2 }, \quad
{ e_{\gamma}^{{lb}}(R)}= R^{\gamma/2 } (1 - R)^{\gamma},$$ as testify in \[Sec Model 3 upper\ ] and \[Sec Model 3 lower\ ].
This Model 3 for the passage function $ \mathcal{B}$ is of particular importance for establishing macroscopic moment models starting from the Boltzmann equality. Namely, in [ @MPC - Dj - S ] it is shown that the six moments model with this passage function provides a reservoir term which satisfies the macroscopic residual inequality on the hale compass of six moments model validity, that provides the total agreement with the strain thermodynamics theory of six moments, as one of the rare system admitting a non - linear or * far from equilibrium * closure of the government equations use the entropy principle. Moreover, for this model 3 of the passage function $ \mathcal{B}$, the macroscopic fourteen moments model achieves the matching with experimental data as well. More precisely, [ @MPC - Dj - S ] shows that the model 3 yields transport coefficients (shear and bulk viscosities and heat conductivity) that provide both experimentally measured dependence of the shear viscosity upon temperature and the Prantdtl act which coincides with its theoretical value at a satisfactory level. This noteworthy achiever lies in the new * additive * form of the passage function that we propose in, instead of the multiplicative one used therefore far in the literature.
Fundamental lemmas { # Sec: fund lemmas }
= = = = = = = = = = = = = = = = = =
In this Section we prove two cardinal lemmas, that should be used sequentially, as they are presented.\
All of them are motivated by the search of a proof showing that $ k$-th polynomial moment of the solution will meet an ODI with a damaging superlinear term, that is $ $ \frac{\mathrm{d } } { \mathrm{d}t } \mathfrak{m}_k[f](t)= \int _ { \Delta } Q(f, f) ( | \lax\left\{ R, (1-R) \right\}^{\gamma/2}, \qmad
{e_{\gamma}^{{lb}}(R)}= \miu\{ R, (1-R) \}^{\gemma/2}, \ens{gatherea}$$ or with $${d_{\gamma}^{{lb}}(r)}= r^{\gamma/2} (1-r)^{\jammq/2}, \quae {d_{\gamma}^{{ub}}(r)}=1,\quad
{e_{\gamma}^{{jb}}(R)}= 2^{1-\gammw/2}, \quad
{e_{\ganna}^{{lb}}(R)}= R^{\gamma/2} (1-R)^{\gamma},$$ as sglwn nn \[Sec Model 3 upker\] and \[Sec Kodel 3 lower\].
Thhs Mldel 3 for the transition function $\mwthcal{B}$ id of particulat imppwtandv nor establishing macroscopic momsnts movels starting ftom the Boltzmann equation. Namfly, in [@MPC-Dj-S] it id shown thaj thq six moments model with this transjtion function provides a sourcd terk which sajnwfifv the macrowcopib residual incaualitf on thr whole range pf vix moments model validivy, that provides the total agsezment with the extendwd therkodytamizw tfeody or six lomxnts, as one of the rarw systems admitting a bon-linear or *rar frjm equilibrium* closure of the governing exuafions using the entropy principle. Moreover, flr this mjdel 3 of the transition function $\mathcal{B}$, the macsosco'iz fiuvteev mlments model achieves the matching with experyjemtsl data as welj. More precosflu, [@MPC-Dj-S] shows that cge model 3 yields trajsport soeffucients (srear and bulk viscosities and hwat conductinity) that provide both experimentclly mgasurec dependence of the shecr visdosity upon temperathfe and the Prantatl ngmber which coincides with its theoceticcl value at s satiffactory lfvel. Bvis remarkable sucfess pias in the jew *additive* form of the transivmon function jhad wv propose in, ikstead of the mtltiplicative pne useb so fxr in the miteratnre.
Fundamentwl lemmas {#Sec: vund lemmas}
==================
Ii this Sestiob we prove gdo fundamental lemmas, tkct should be used sequentiakly, zs they are prewenuee.\
All of them ate ootyvwtxd by dhe search ox a oroug shoding that $k$-bh oolymomial moment of the solhtion will satisfy ak ODI wity a negaeive superlinrar term, that is $$\fgac{\mavhrm{d} }{\oathrk{d}t} \mathfrak{m}_k[f](t)= \int_{ \Delta} Q(f,f)( | \max\left\{ R, (1-R) \right\}^{\gamma/2}, \quad {e_{\gamma}^{{lb}}(R)}= \min\{ \}^{\gamma/2}, or with r^{\gamma/2} (1-r)^{\gamma/2}, \quad R^{\gamma/2} as shown in Model 3 upper\] \[Sec Model 3 lower\]. This Model for the transition function $\mathcal{B}$ is of particular importance for establishing macroscopic moments starting from the Boltzmann equation. Namely, in [@MPC-Dj-S] it is shown that the moments with transition provides a source term which satisfies the macroscopic residual inequality on the whole range of six model validity, that provides the total agreement with extended thermodynamics theory of moments, as one of the systems a non-linear *far equilibrium* of the governing using the entropy principle. Moreover, for this model 3 of the transition function $\mathcal{B}$, the macroscopic fourteen model achieves with experimental as More [@MPC-Dj-S] shows that 3 yields transport coefficients (shear and heat conductivity) that provide both experimentally measured dependence the shear upon temperature and the Prantdtl number coincides with its theoretical value at a satisfactory This remarkable success lies in the new *additive* form of the transition function that we instead of the multiplicative used so far the Fundamental {#Sec: lemmas} ================== this Section we prove two fundamental lemmas, that should be used as they are presented.\ All of them are motivated by of proof showing that polynomial moment of the will an ODI with a term, is \int_{ Q(f,f)( | \max\left\{ R, (1-R) \right\}^{\gamma/2}, \quad
{e_{\gAmma}^{{lb}}(R)}= \min\{ r, (1-R) \}^{\gamMa/2}, \eNd{gAtHereD}$$ or wIth $${d_{\gamma}^{{lb}}(r)}= r^{\gAMma/2} (1-r)^{\Gamma/2}, \quad {d_{\gamma}^{{ub}}(r)}=1,\quad
{E_{\gammA}^{{uB}}(r)}= 2^{1-\gamMA/2}, \qUad
{e_{\gAmma}^{{lb}}(R)}= r^{\GaMMA/2} (1-R)^{\gAmMa},$$ As sHoWN iN \[Sec MOdeL 3 upper\] aNd \[Sec Model 3 LowEr\].
this Model 3 for THe Transition FunCtion $\mathcal{b}$ is Of partIcUlaR ImporTanCe for EstablIShing mAcroscopiC mOMents mODels staRTInG froM the Boltzmann equaTIoN. namely, in [@MPC-Dj-S] It is shOwN ThAT The Six Moments modEl With tHIs transITiON FUncTIon provides a sOurce term whICh sAtisfiEs The MAcroscOpic rEsIDuaL inequality On thE whole ranGe of siX Moments MOdel valIdity, tHat ProVideS ThE tOtaL aGReeMEnT wiTH thE extendeD tHeRmodyNamiCS THEory Of sIx moMents, As one of the rarE syStemS AdmIttinG a non-LineAr Or *far From eqUilibRiUm* closure of the gOverNing equatIonS uSinG tHe entROpy priNciPle. moreoveR, for thiS ModEl 3 OF THe Transition function $\MaTHCaL{B}$, the macRoscopIC fOuRTeen momeNtS moDel aCHIeves The mATcHing with ExperiMEnTaL data as WeLl. More PrEciSelY, [@MPC-DJ-s] shoWs that The model 3 YieldS Transport coeffICients (shear anD BuLK ViSCosiTieS and heat conDuctIVity) That PRoVidE Both eXperiMeNTaLLy measured dependencE oF the shEar viScosity upon teMperature aND THe PrantdTl nuMBeR Which coincides With iTs theoretiCAl value aT a satIsfactorY level. ThiS REmarkablE suCceSs lIes IN ThE new *additive* fORM of tHe TransitIon FunctioN thAt wE prOpoSe In, instead Of the mulTiPlIcAtIve One usED so far in ThE liTeRatUre.
FuNDamentAl lemMas {#SEc: FuND leMmas}
==================
In tHIs sECtioN wE pRove Two FuNdameNtal LEmmAs, that sHould be usEd sEQuenTiAlLy, as theY are presented.\
alL of them are MoTivAted by THE search oF a proof showing that $k$-th poLYnomial MomEnt of The sOlution wiLl sAtisfy An Odi with a NegatiVe supErLinEAR term, THAt Is $$\fRaC{\mathrm{d} }{\maTHRm{d}T} \mathFrAk{m}_k[F](t)= \int_{ \DeLta} Q(f,f)( | \max\left\{ R, (1-R) \r ight\}^{\g amma/ 2}, \q ua d
{ e_{\ gamma}^{{lb}}( R )}=\min\{ R, (1-R) \}^{\g amma/ 2} , \en d {g ather ed}$$ o r w i t h $ ${ d_ {\g am m a} ^{{lb }}( r)}= r^ {\gamma/2} (1 -r )^{\gamma/2} , \ quad {d_{\ gam ma}^{{ub}}(r )}= 1,\qua d {e _ {\gam ma} ^{{ub }}(R)} = 2^{1 -\gamma/2 }, \quad {e_{\ g a mm a}^{ {lb}}(R)}= R^{\ga m ma / 2} (1-R)^{\gam ma},$$ a s s h o wnin\[Sec Mode l3 upp e r\] and \[ S e c Mo d el 3 lower\].
This Mode l 3for th etra n sition func ti o n $ \mathcal{B} $ is of parti culari mportan c e for e stabli shi ngmacr o sc op icmo m ent s m ode l s s tartingfr om theBolt z m a n n eq uat ion. Name ly, in [@MPC- Dj- S] i t is show n tha t th esix m oments mode lwith this trans itio n functio n p ro vid es a so u rce te rmwhi ch sati sfies t h e m ac r o s co pic residual inequ al i t yon the w hole r a ng eo f six mo me nts mod e l vali dity , t hat prov ides t h eto tal agr ee ment w it h t heexten d ed t hermod ynamicstheor y of six moment s , as one of t h er a re syst ems admittinga no n -lin earo r*fa r from equi li b ri u m* closure of the g ov erning equa tions using t he entropy p r inciple. Mor e ov e r, for this mo del 3 of the tr a nsitionfunct ion $\ma thcal{B}$ , the macr osc opi c f our t e en moments mode l achi ev es themat ching w ith ex per ime nt al data a s well.Mo re p re cis ely,[ @MPC-Dj- S] sh ow s t hat t h e mode l 3 y ield str a nsp ort coe f fi c i ents ( sh earand b ulk v isco s iti es andheat cond uct i vity )th at prov ide both expe ri mentally m ea sur ed dep e n dence of the shear viscosity up o n tempe rat ure a nd t he Prantd tlnumber wh i ch coi ncides with i tst h eoret i c al va lu e at a sat i s fac toryle vel. This r emarkable successl ies in the new * add itiv e * f orm of the t r ans i t ion function th at we prop os e i n, instead ofth e multi plicati ve on e used s o far inthe liter at ure. Fun damental l emmas {# Sec: fund lemma s }===== === ====== == ==
In t his Se c tio n weprovetw o fund ament al lemmas, that should be used se quenti ally, as they are pr e sen ted.\
All ofthem are m oti vat ed by th e sear ch o f a pr o of sh owin g that $k$ - th po l y no mial moment o f th e sol uti o n will sat isfy an ODI witha negative supe rlin e a r t erm , tha tis $$\frac{\ma thr m{ d } }{\math rm {d}t} \math frak{m}_ k[ f ](t)= \in t_{ \D elta} Q ( f ,f ) ( | _ \max\left\{_R, (1-R) \right\}^{\gamma/2}, \quad
_{e_{\gamma}^{{lb}}(R)}= \min\{_R,_(1-R) \}^{\gamma/2},_\end{gathered}$$_or with $${d_{\gamma}^{{lb}}(r)}=_r^{\gamma/2} (1-r)^{\gamma/2}, \quad_{d_{\gamma}^{{ub}}(r)}=1,\quad
{e_{\gamma}^{{ub}}(R)}= 2^{1-\gamma/2},_\quad
_{e_{\gamma}^{{lb}}(R)}=_R^{\gamma/2} (1-R)^{\gamma},$$ as shown in \[Sec Model 3 upper\] and \[Sec Model 3 lower\].
This_Model_3 for_the_transition_function $\mathcal{B}$ is of particular_importance for establishing macroscopic moments_models starting_from the Boltzmann equation. Namely, in [@MPC-Dj-S] it_is_shown that the_six moments model with this transition function provides a_source term which satisfies the macroscopic_residual inequality on_the_whole_range of six moments_model validity, that provides the total_agreement with the extended thermodynamics theory_of six moments, as one of the_rare systems admitting a non-linear or_*far from equilibrium* closure of_the governing_equations using the entropy principle._Moreover, for this_model 3_of the transition_function $\mathcal{B}$, the macroscopic fourteen moments_model achieves the_matching with experimental data as well._More_precisely, [@MPC-Dj-S] shows_that_the_model 3_yields transport coefficients_(shear_and bulk_viscosities_and heat conductivity) that provide both_experimentally_measured dependence of the shear viscosity upon_temperature and the Prantdtl_number_which coincides with its_theoretical value at a satisfactory_level. This remarkable success lies in_the new_*additive* form_of the transition function that we propose in, instead of the_multiplicative one used so far in_the literature.
Fundamental lemmas {#Sec:_fund lemmas}
==================
In_this_Section we prove_two_fundamental lemmas,_that should be used sequentially, as they_are presented.\
All_of them are motivated by the_search of a proof_showing_that $k$-th polynomial moment of the_solution will satisfy an ODI with_a negative superlinear term, that_is_$$\frac{\mathrm{d}_}{\mathrm{d}t} \mathfrak{m}_k[f](t)= \int_{_\Delta} Q(f,f)( |
* model of an activity. We provide both quantitative and qualitative results of our zero-shot framework using these “on the fly” models on the challenging DIVA dataset[^2], which contains fine-grained human-object interactions under a real world video surveillance setting.
In summary, the main contributions of the paper are:
- A zero-shot classification of complex action sequences with dynamic action signatures which establishes a new state-of-the-art on Olympic Sports [@olympic-sports-2010] dataset. We outperform all other methods for the ZSL evaluation regardless of training assumptions (inductive/transductive).
- To the best of our knowledge, we are the first to demonstrate zero-shot decoding of complex action sequences. We present our results on a surgical dataset, JIGSAWS [@jigsaws-2014], to jointly segment and classify fine-grained surgical gestures where we establish an impressive baseline.
- A demonstration of zero-shot classification of fine-grained human-object interactions that requires no supervised training of attributes by leveraging off-the-shelf object detectors in video surveillance.
Related Work
============
Methodology {#sec:methodology}
===========
We first establish a basic hierarchy of concepts. At the highest level we have the *activity*—for example, suturing in robotic surgery. Each activity can be decomposed into a sequence of actions $(y_1, \ldots, y_N)$. Possible examples of actions are “Pushing needle through tissue" in a suturing activity or “throwing javelin" in a sporting event. Zero-shot learning approaches further decompose each action $y$ into a set of $K$ elementary attributes (usually taken to be binary-valued) $y = \{a_1, \ldots, a_K\}$.
Given a video recording of an activity (represented as a sequence of frames $X = (x_1, \ldots, x_T)$), our goal is to map each frame $x_t$ to its corresponding action $y_t$ by detecting the presence or absence of each attribute $\hat{a}(x_t)$ in the video, then choosing the action whose signature $a(y_t)$ best fits those attributes. In other words, we choose the action with highest score: $$\label{eq:decode_cost}
\hat{y}(x) = \arg | * model of an activity. We provide both quantitative and qualitative results of our zero - scene model using these “ on the fly ” model on the ambitious DIVA dataset[^2 ], which contains fine - grained human - aim interaction under a real universe video surveillance setting.
In summary, the independent contributions of the paper are:
- A zero - guess categorization of complex action sequences with active action touch which establishes a new state - of - the - art on Olympic Sports [ @olympic - sports-2010 ] dataset. We surpass all other methods for the ZSL evaluation regardless of education assumptions (inductive / transductive).
- To the good of our knowledge, we are the beginning to demonstrate zero - shot decoding of complex action sequence. We present our results on a surgical dataset, JIGSAWS [ @jigsaws-2014 ], to jointly segment and classify fine - granulate surgical gestures where we establish an impressive baseline.
- A demonstration of zero - shot classification of fine - grained human - object interactions that necessitate no supervised training of attributes by leverage off - the - ledge aim detectors in video surveillance.
Related Work
= = = = = = = = = = = =
Methodology { # sec: methodology }
= = = = = = = = = = =
We first establish a basic hierarchy of concepts. At the highest level we have the * activity*—for case, suturing in robotic surgery. Each activity can be decomposed into a sequence of military action $ (y_1, \ldots, y_N)$. Possible examples of actions are “ Pushing needle through tissue " in a suturing action or “ throwing javelin " in a sporting event. Zero - shot learning approaches far decompose each action $ y$ into a set of $ K$ elementary attributes (usually taken to be binary - valued) $ y = \{a_1, \ldots, a_K\}$.
give a video recording of an bodily process (represented as a sequence of skeleton $ X = (x_1, \ldots, x_T)$), our goal is to map each human body $ x_t$ to its corresponding military action $ y_t$ by detecting the bearing or absence of each attribute $ \hat{a}(x_t)$ in the video, then choosing the legal action whose key signature $ a(y_t)$ best fits those attributes. In other words, we choose the action with highest score: $ $ \label{eq: decode_cost }
\hat{y}(x) = \arg | * mofel of an activity. We pruvide both quanjirative and qhalitatixe results of our zero-shot fcameqork ysing these “on the fly” models oj the chqlleiging DIVA datasxf[^2], which contajks fiue-jrained human-obmect interawtions under a rdap world video surveillance setting.
Ig summaty, the main conttibutpogs or the paper are:
- A zero-shot classjficatimn of complex action sequences with dynwmic action signatures which estavlisrws a new stage-of-the-art on Olympic Aports [@olympic-sports-2010] dataset. We outpzrform all ityer kethods for the DSL evaluatiok regargless og training assmmptimns (inductive/transductivx).
- To the best of oor knowledce, we are the first ti demmnstsate zeru-shkt dscodinh oh complex adtion sequebces. We present our rqwults on a sudgical dwtaset, JIGSAWS [@jigsaws-2014], to jointly segmett znd classify fine-grainee surgical gestures wjere we eftablish an impressive baseline.
- A demonstration of zxru-shit zoadsification of fine-grained human-object interasfipnx that requiref no supervosfd jraining of atjributes bg leveraging off-thf-shelf jbjecr detectows im video surveillance.
Related Work
============
Methodojigy {#sec:methodology}
===========
Ce first estcblish a baxic hierarchy of concepcs. At fhe highest level we fave the *activith*—fog exdmple, suujring in robotic furgery. Eech aetivity zan ne decjmposed inho a sequence of actions $(y_1, \ldltv, y_N)$. Possihle examples of actions are “Pusimng needle thtouch nissue" in a suburing activity or “throwing jcvelin" iu a spurting eveht. Zero-vhot learnigg approaches vurther decokpose eash axtiob $y$ intu a set of $K$ elrmentary cutributes (ysually taken to bc binxdy-valued) $y = \{a_1, \lbuus, a_K\}$.
Given a vidro fecjrfiig of dn activity (seprdsevyed ar a sequencc ow frsmes $X = (x_1, \ldots, x_T)$), ogr gkal is to map each fvame $x_t$ ti its cowresponding avtion $y_t$ by detectlng tie prerence or absence of each attribute $\hat{z}(x_t)$ in thf vldeo, then chojsinn thg action whpse signature $a(y_t)$ best fits those attriuutes. In other words, wg choose the action wntm highest scpre: $$\lwbel{eq:decmde_cost}
\hat{y}(x) = \art | * model of an activity. We provide and results of zero-shot framework using on challenging DIVA dataset[^2], contains fine-grained human-object under a real world video surveillance In summary, the main contributions of the paper are: - A zero-shot classification complex action sequences with dynamic action signatures which establishes a new state-of-the-art on Sports dataset. outperform other methods for the ZSL evaluation regardless of training assumptions (inductive/transductive). - To the best of knowledge, we are the first to demonstrate zero-shot of complex action sequences. present our results on a dataset, [@jigsaws-2014], to segment classify surgical gestures where establish an impressive baseline. - A demonstration of zero-shot classification of fine-grained human-object interactions that requires no training of leveraging off-the-shelf detectors video Related Work ============ =========== We first establish a basic At the highest level we have the *activity*—for suturing in surgery. Each activity can be decomposed a sequence of actions $(y_1, \ldots, y_N)$. Possible of actions are “Pushing needle through tissue" in a suturing activity or “throwing javelin" in event. Zero-shot learning approaches decompose each action into set $K$ attributes (usually to be binary-valued) $y = \{a_1, \ldots, a_K\}$. Given a video of an activity (represented as a sequence of frames $X \ldots, our goal is map each frame $x_t$ its action $y_t$ by detecting or of in video, choosing the action whose $a(y_t)$ best fits those attributes. other words, we choose $$\label{eq:decode_cost} \hat{y}(x) = \arg | * model of an activity. We providE both quantItatiVe aNd qUaLitaTive Results of our zeRO-shoT framework using these “on The flY” mODels ON tHe chaLlenginG dIva DatAsEt[^2], WhiCh COnTains FinE-graineD human-objeCt iNtEractions undER a Real world vIdeO surveillancE seTting.
IN sUmmARy, the MaiN contRibutiONs of thE paper are:
- a zERo-shot CLassifiCATiOn of Complex action sequENcES with dynamic acTion siGnATuRES whIch EstablisheS a New stATe-of-the-ARt ON oLymPIc Sports [@olympIc-sports-2010] datASet. we outpErForM All othEr metHoDS foR the ZSL evalUatiOn regardlEss of tRAining aSSumptioNs (induCtiVe/tRansDUcTiVe).
- TO tHE beST oF ouR KnoWledge, we ArE tHe firSt to DEMONstrAte Zero-Shot dEcoding of compLex ActiON seQuencEs. We pReseNt Our reSults oN a surGiCal dataset, JIGSAwS [@jiGsaws-2014], to joIntLy SegMeNt and CLassifY fiNe-gRained sUrgical GEstUrES WHeRe we establish an impReSSIvE baselinE.
- A demoNStRaTIon of zerO-sHot ClasSIFicatIon oF FiNe-graineD human-OBjEcT interaCtIons thAt ReqUirEs no sUPervIsed trAining of AttriBUtes by leveragiNG off-the-shelf oBJeCT DeTEctoRs iN video surveIllaNCe.
ReLateD woRk
============
METhodoLogy {#sEc:MEtHOdology}
===========
We first estabLiSh a basIc hieRarchy of concePts. At the hiGHESt level wE havE ThE *Activity*—for exaMple, sUturing in rOBotic surGery. EAch activIty can be dECOmposed iNto A seQueNce OF AcTions $(y_1, \ldots, y_N)$. pOSsibLe ExampleS of Actions Are “pusHinG neEdLe through Tissue" in A sUtUrInG acTivitY Or “throwiNg JavElIn" iN a spoRTing evEnt. ZeRo-shOt LeARniNg approAChES FurtHeR dEcomPosE eAch acTion $Y$ IntO a set of $k$ elementaRy aTTribUtEs (Usually Taken to be binaRy-Valued) $y = \{a_1, \ldOtS, a_K\}$.
given a VIDeo recorDing of an activity (represeNTed as a sEquEnce oF fraMes $X = (x_1, \ldotS, x_T)$), Our goaL is TO map eaCh framE $x_t$ to ItS coRREsponDINg ActIoN $y_t$ by detecTINg tHe preSeNce oR absencE of each attribute $\haT{A}(x_t)$ In the video, theN chOosiNG ThE acTIoN WhoSe SIgnATUre $a(y_t)$ best fits tHose attribUtES. IN other wordS, We cHoOse the aCtion wiTh higHEst scorE: $$\label{eq:dEcode_cost}
\HaT{y}(x) = \aRG | * model of an activity. We provide b oth q uan tit at iveandqualitative re s ults of our zero-shot fram ework u s ingt he se “o n the f l y” m ode ls o n t he ch allen gin g DIVAdataset[^2 ],wh ich contains fi ne-grained hu man-object i nte ractio ns un d er area l wor ld vid e o surv eillancese t ting.In summ a r y, the main contributio n so f the paper ar e:
- A z e r o-s hot classific at ion o f comple x a c t i ons equences with dynamic ac t ion signa tu res whichestab li s hes a new stat e-of -the-arton Oly m pic Spo r ts [@ol ympic- spo rts -201 0 ]da tas et . We ou tpe r for m all ot he rmetho ds f o r t he Z SLeval uatio n regardlessoftrai n ing assu mptio ns ( in ducti ve/tra nsduc ti ve).
- To th e be st of our kn ow led ge , wea re the fi rst to dem onstrat e ze ro - s h ot decoding of compl ex a ct ion sequ ences. We p r esent ou rres ults o n a s urgi c al dataset , JIGS A WS [ @jigsaw s- 2014], t o j oin tly s e gmen t andclassify fine - grained surgic a l gestures wh e re w ee stab lis h an impres sive base line .
- A dem onstr at i on of zero-shot classi fi cation of f ine-grained h uman-objec t i nteracti onst ha t requires no s uperv ised train i ng of at tribu tes by l everaging o ff-the-s hel f o bje ctd e te ctors in vide o surv ei llance.
R elatedWor k
= === === == ===
Meth odology{# se c: me tho dolog y }
====== == ===
Wefirst establ ish a bas ic h i era rchy of co n c epts .At the hi gh est l evel wehave th e *activi ty* — forex am ple, su turing in rob ot ic surgery .Eac h acti v i ty can b e decomposed into a seq u ence of ac tions $(y _1, \ldot s,y_N)$. Po s sibleexampl es of a cti o n s are “ Pu shi ng needle th r o ugh tiss ue " in a sutu ring activity or “ t hro wing javelin" in a s p o rt ing ev e nt. Z e ro- s h ot learning app roaches fu rt h er decompose eac haction$y$ int o a s e t of $K $ element ary attri bu tes( u sua lly takento be bi nary-valu e d) $y =\{a_1 , \ ldots, a _K\ }$.
Givena vi deo r ecordi ng of an acti vi ty (repr esented as a sequence o f fram es $X =(x_1, \ld ots , x_ T)$), our goa l is to ma p e ach fram e $ x _t$ t o it s c orr e spond inga ction $y_ t $byd e te cting the p r e s enc e orabs e nce of eac h attribute $\hat { a}(x_t)$ in th e vi d e o,the n cho os ing the action wh os e signatur e$a(y_t)$ be st fitsth o se at tribut es. In otherw o rd s , we c hoos e t he action wi th highest s co r e: $$\ labe l{ eq:dec ode_co s t}
\hat{y}(x) = \a rg | * model_of an_activity. We provide both_quantitative and_qualitative_results of_our_zero-shot framework using_these “on the_fly” models on the_challenging DIVA dataset[^2],_which_contains fine-grained human-object interactions under a real world video surveillance setting.
In summary, the main_contributions_of the_paper_are:
-_ A zero-shot classification_of complex action sequences with_dynamic action_signatures which establishes a new state-of-the-art on Olympic_Sports_[@olympic-sports-2010] dataset. We_outperform all other methods for the ZSL evaluation regardless_of training assumptions (inductive/transductive).
- _To the best_of_our_knowledge, we are the_first to demonstrate zero-shot decoding of_complex action sequences. We present our_results on a surgical dataset, JIGSAWS [@jigsaws-2014],_to jointly segment and classify fine-grained_surgical gestures where we establish_an impressive_baseline.
- A demonstration_of zero-shot classification_of fine-grained_human-object interactions that_requires no supervised training of attributes_by leveraging off-the-shelf_object detectors in video surveillance.
Related Work
============
Methodology_{#sec:methodology}
===========
We_first establish a_basic_hierarchy_of concepts._At the highest_level_we have_the_*activity*—for example, suturing in robotic surgery._Each_activity can be decomposed into a sequence_of actions $(y_1, \ldots,_y_N)$._Possible examples of actions_are “Pushing needle through tissue"_in a suturing activity or “throwing_javelin" in_a sporting_event. Zero-shot learning approaches further decompose each action $y$ into a_set of $K$ elementary attributes (usually_taken to be binary-valued)_$y =_\{a_1,_\ldots, a_K\}$.
Given a_video_recording of_an activity (represented as a sequence of_frames $X_= (x_1, \ldots, x_T)$), our goal_is to map each_frame_$x_t$ to its corresponding action $y_t$_by detecting the presence or absence_of each attribute $\hat{a}(x_t)$ in_the_video,_then choosing the action whose_signature $a(y_t)$ best fits those attributes._In other words,_we choose the action with highest score:_$$\label{eq:decode_cost}
_ \hat{y}(x) = \arg |
Z}}\mathbb{E} \bigl[\Gamma_{w(s)}^2\bigl(1,\{x/w(s)\}
\bigr) \bigr]
= \sum_{x\in\mathbb{Z}}\mathbb{E} \bigl[\tilde{\Gamma}_{w(s)}^2\bigl(1,\{
x/w(s)\}\bigr) \bigr].$$ It follows from the layer cake representation and the remark after the proof of Proposition \[SignedCardTowardMeasureProp\] that $$\begin{aligned}
w(s)\sum_{x\in\mathbb{Z}}\tilde{\Gamma}_{w(s)}^2\bigl(1,\{x/w(s)\}\bigr)
=\frac{1}{w(s)}\int_0^\infty\operatorname{card}
\bigl\{x\in\mathbb{Z}\dvtx
w^2(s)\tilde{\Gamma}^2_{w(s)}\bigl(1,\{x/w(s)\}\bigr)>c \bigr\}\,\mathrm{d}c\end{aligned}$$ converges $ \mathbb{P} \times\tilde{\mathbb{P} } $-almost surely toward $$\int_0^\infty\ell\bigl(x\in\mathbb{R}\dvtx\tilde{L}^2(1,x)>c \bigr)\,\mathrm{d}c
=\int_{\mathbb{R}}\tilde{L}_\ast^2(1,x)\,\mathrm{d}x.$$ Dominated convergence and Fubini’s theorem imply that $$w(s)\sum_{x\in\mathbb{Z}}\mathbb{E} \bigl[\tilde{\Gamma}_{w(s)}^2\bigl(1,\{
x/w(s)\}\bigr)
\bigr]\longrightarrow
\int_{\mathbb{R}}\mathbb{E} [\tilde{L}_\ast^2(1,x) ]\,\mathrm{d}x\qquad
\mbox{as }s\rightarrow\infty.$$ Therefore, $$w(s)s^{-2}\sum_{x\in\mathbb{Z}}\mathbb{E} [\Gamma^2(s,\{x\})
]\longrightarrow
\int_{\mathbb{R}}\mathbb{E} [\tilde{L}_\ast^2(1,x) ]\,\mathrm{d}x\qquad \mbox{as }
| Z}}\mathbb{E } \bigl[\Gamma_{w(s)}^2\bigl(1,\{x / w(s)\ }
\bigr) \bigr ]
= \sum_{x\in\mathbb{Z}}\mathbb{E } \bigl[\tilde{\Gamma}_{w(s)}^2\bigl(1,\ {
x / w(s)\}\bigr) \bigr].$$ It follows from the layer cake representation and the remark after the validation of Proposition \[SignedCardTowardMeasureProp\ ] that $ $ \begin{aligned }
w(s)\sum_{x\in\mathbb{Z}}\tilde{\Gamma}_{w(s)}^2\bigl(1,\{x / w(s)\}\bigr)
= \frac{1}{w(s)}\int_0^\infty\operatorname{card }
\bigl\{x\in\mathbb{Z}\dvtx
w^2(s)\tilde{\Gamma}^2_{w(s)}\bigl(1,\{x / w(s)\}\bigr)>c \bigr\}\,\mathrm{d}c\end{aligned}$$ converge $ \mathbb{P } \times\tilde{\mathbb{P } } $ -almost surely toward $ $ \int_0^\infty\ell\bigl(x\in\mathbb{R}\dvtx\tilde{L}^2(1,x)>c \bigr)\,\mathrm{d}c
= \int_{\mathbb{R}}\tilde{L}_\ast^2(1,x)\,\mathrm{d}x.$$ Dominated convergence and Fubini ’s theorem entail that $ $ w(s)\sum_{x\in\mathbb{Z}}\mathbb{E } \bigl[\tilde{\Gamma}_{w(s)}^2\bigl(1,\ {
x / w(s)\}\bigr)
\bigr]\longrightarrow
\int_{\mathbb{R}}\mathbb{E } [ \tilde{L}_\ast^2(1,x) ] \,\mathrm{d}x\qquad
\mbox{as } s\rightarrow\infty.$$ consequently, $ $ w(s)s^{-2}\sum_{x\in\mathbb{Z}}\mathbb{E } [ \Gamma^2(s,\{x\ })
] \longrightarrow
\int_{\mathbb{R}}\mathbb{E } [ \tilde{L}_\ast^2(1,x) ] \,\mathrm{d}x\qquad \mbox{as } | Z}}\mahhbb{E} \bigl[\Gamma_{w(s)}^2\bigl(1,\{x/w(s)\}
\nigr) \bigr]
= \sum_{x\in\mathbb{Z}}\kathbb{S} \bigl[\tiude{\Gamma}_{w(s)}^2\bigl(1,\{
x/w(s)\}\bigr) \bigr].$$ Iv foolows from the layer cake rdpresentanion and rhe cemark after the proof on Prollsitnoi \[SignedCardTowsrdMeasure[rop\] that $$\begit{auiyned}
w(s)\sum_{x\in\mathbb{Z}}\tilde{\Gamma}_{w(s)}^2\bigl(1,\{x/r(s)\}\bigr)
=\ftaf{1}{w(s)}\int_0^\infty\opetatormwme{czgd}
\nigl\{x\in\mathbb{Z}\dvtx
w^2(s)\tilde{\Gamma}^2_{w(s)}\gigl(1,\{x/w(s)\}\uigr)>c \bigr\}\,\mathrk{d}c\end{aligned}$$ converges $ \mwthbh{P} \times\tilde{\mathbh{P} } $-almost wurejt toward $$\int_0^\knfty\ell\bigl(x\in\mathbb{R}\svtx\tilde{L}^2(1,x)>c \bigr)\,\mathrm{d}c
=\int_{\matfbb{R}}\tnlde{L}_\ast^2(1,x)\,\majkem{d}d.$$ Dominated ronverdence and Funpni’s thaorem ikply that $$w(s)\suk_{x\ii\matybb{Z}}\mathbb{E} \bigl[\tilde{\Jamma}_{w(s)}^2\bigl(1,\{
x/w(s)\}\bigr)
\bidr]\longrigvtcrrow
\int_{\mathbb{R}}\mathbb{W} [\rilde{N}_\ast^2(1,f) ]\,\magyrm{a}x\qsued
\mgox{as }d\rijhtarrow\inffy.$$ Thereforw, $$w(s)s^{-2}\sum_{x\in\mathbb{Z}}\mstrvb{E} [\Gamma^2(s,\{x\})
]\lohgrigheawrow
\int_{\mathbb{R}}\mathbb{E} [\tilde{L}_\ast^2(1,x) ]\,\mathrm{g}x\qsuad \mbox{as }
| Z}}\mathbb{E} \bigl[\Gamma_{w(s)}^2\bigl(1,\{x/w(s)\} \bigr) \bigr] = \sum_{x\in\mathbb{Z}}\mathbb{E} \bigl[\tilde{\Gamma}_{w(s)}^2\bigl(1,\{ It from the cake representation and of \[SignedCardTowardMeasureProp\] that $$\begin{aligned} =\frac{1}{w(s)}\int_0^\infty\operatorname{card} \bigl\{x\in\mathbb{Z}\dvtx w^2(s)\tilde{\Gamma}^2_{w(s)}\bigl(1,\{x/w(s)\}\bigr)>c converges $ \mathbb{P} \times\tilde{\mathbb{P} } $-almost toward $$\int_0^\infty\ell\bigl(x\in\mathbb{R}\dvtx\tilde{L}^2(1,x)>c \bigr)\,\mathrm{d}c =\int_{\mathbb{R}}\tilde{L}_\ast^2(1,x)\,\mathrm{d}x.$$ Dominated convergence and Fubini’s theorem imply that $$w(s)\sum_{x\in\mathbb{Z}}\mathbb{E} \bigl[\tilde{\Gamma}_{w(s)}^2\bigl(1,\{ \bigr]\longrightarrow \int_{\mathbb{R}}\mathbb{E} [\tilde{L}_\ast^2(1,x) ]\,\mathrm{d}x\qquad \mbox{as }s\rightarrow\infty.$$ Therefore, $$w(s)s^{-2}\sum_{x\in\mathbb{Z}}\mathbb{E} [\Gamma^2(s,\{x\}) ]\longrightarrow \int_{\mathbb{R}}\mathbb{E} [\tilde{L}_\ast^2(1,x) ]\,\mathrm{d}x\qquad } | Z}}\mathbb{E} \bigl[\Gamma_{w(s)}^2\bigl(1,\{x/w(S)\}
\bigr) \bigr]
= \sUm_{x\in\MatHbb{z}}\mAthbB{E} \biGl[\tilde{\Gamma}_{w(s)}^2\BIgl(1,\{
x/W(s)\}\bigr) \bigr].$$ It follows froM the lAyER cakE RePreseNtation ANd THE reMaRk AftEr THe Proof Of PRopositIon \[SignedCArdtoWardMeasurePROp\] That $$\begin{aLigNed}
w(s)\sum_{x\in\mAthBb{Z}}\tilDe{\gamMA}_{w(s)}^2\biGl(1,\{x/W(s)\}\bigR)
=\frac{1}{w(S)}\Int_0^\infTy\operatoRnAMe{card}
\BIgl\{x\in\mATHbB{Z}\dvTx
w^2(s)\tilde{\Gamma}^2_{w(s)}\bIGl(1,\{X/W(s)\}\bigr)>c \bigr\}\,\matHrm{d}c\eNd{ALiGNEd}$$ cOnvErges $ \mathbB{P} \Times\TIlde{\matHBb{p} } $-ALMosT Surely toward $$\iNt_0^\infty\ell\bIGl(x\In\mathBb{r}\dvTX\tilde{l}^2(1,x)>c \biGr)\,\MAthRm{d}c
=\int_{\mathBb{R}}\tIlde{L}_\ast^2(1,x)\,\Mathrm{D}X.$$ DominaTEd conveRgence And fubIni’s THeOrEm iMpLY thAT $$w(S)\suM_{X\in\Mathbb{Z}}\mAtHbB{E} \bigL[\tilDE{\gAMma}_{w(S)}^2\biGl(1,\{
x/w(S)\}\bigr)
\Bigr]\longrightArrOw
\inT_{\MatHbb{R}}\mAthbb{e} [\tilDe{l}_\ast^2(1,x) ]\,\Mathrm{D}x\qquAd
\Mbox{as }s\rightarrOw\inFty.$$ TherefOre, $$W(s)S^{-2}\suM_{x\In\matHBb{Z}}\matHbb{e} [\GaMma^2(s,\{x\})
]\loNgrightARroW
\iNT_{\MAtHbb{R}}\mathbb{E} [\tilde{L}_\aSt^2(1,X) ]\,\MAtHrm{d}x\qquAd \mbox{AS }
| Z}}\mathbb{E} \bigl[\Gamma _{w(s)}^2\ bigl( 1,\ {x/ w( s)\}
\bi gr) \bigr]
= \ s um_{ x\in\mathbb{Z}}\mathbb {E} \ bi g l[\t i ld e{\Ga mma}_{w ( s) } ^ 2\b ig l( 1,\ {x /w (s)\} \bi gr) \bi gr].$$ Itfol lo ws from thel ay er cake re pre sentation an d t he rem ar k a f ter t heproof of Pr o positi on \[Sign ed C ardTow a rdMeasu r e Pr op\] that $$\begin{al i gn e d}
w(s)\sum_{x \in\ma th b b{ Z } }\t ild e{\Gamma}_ {w (s)}^ 2 \bigl(1 , \{ x / w (s) \ }\bigr)
=\fra c{1}{w(s)}\ i nt_ 0^\inf ty \op e ratorn ame{c ar d }
\ bigl\{x\in\ math bb{Z}\dvt x
w^2( s )\tilde { \Gamma} ^2_{w( s)} \bi gl(1 , \{ x/ w(s )\ } \bi g r) >c\ big r\}\,\ma th rm {d}c\ end{ a l i g ned} $$conv erges $ \mathbb{P} \t imes \ til de{\m athbb {P}}$-alm ost su relyto ward $$\int_0^\ inft y\ell\big l(x \i n\m at hbb{R } \dvtx\ til de{ L}^2(1, x)>c \b i gr) \, \ m a th rm{d}c
=\int_{\mat hb b { R} }\tilde{ L}_\as t ^2 (1 , x)\,\mat hr m{d }x.$ $ Domin ated co nvergenc e andF ub in i’s the or em imp ly th at$$w(s ) \sum _{x\in \mathbb{ Z}}\m a thbb{E} \bigl[ \ tilde{\Gamma} _ {w ( s )} ^ 2\bi gl( 1,\{
x/w(s) \}\b i gr)\big r ]\ lon g right arrow
\ i nt _ {\mathbb{R}}\mathbb {E } [\ti lde{L }_\ast^2(1,x) ]\,\mathr m { d }x\qquad
\mb o x{ a s }s\rightarro w\inf ty.$$ Ther e fore, $$ w(s)s ^{-2}\su m_{x\in\m a t hbb{Z}}\ mat hbb {E} [\ G a mm a^2(s,\{x\})] \ long ri ghtarro w
\ int_{\m ath bb{ R}} \ma th bb{E} [\t ilde{L}_ \a st ^2 (1 ,x) ]\,\ m athrm{d} x\ qqu ad \m box{a s }
| Z}}\mathbb{E} \bigl[\Gamma_{w(s)}^2\bigl(1,\{x/w(s)\}
\bigr)_\bigr]
= \sum_{x\in\mathbb{Z}}\mathbb{E}_\bigl[\tilde{\Gamma}_{w(s)}^2\bigl(1,\{
x/w(s)\}\bigr) \bigr].$$ It follows_from the_layer_cake representation_and_the remark after_the proof of_Proposition \[SignedCardTowardMeasureProp\] that $$\begin{aligned}
w(s)\sum_{x\in\mathbb{Z}}\tilde{\Gamma}_{w(s)}^2\bigl(1,\{x/w(s)\}\bigr)
=\frac{1}{w(s)}\int_0^\infty\operatorname{card}
\bigl\{x\in\mathbb{Z}\dvtx
w^2(s)\tilde{\Gamma}^2_{w(s)}\bigl(1,\{x/w(s)\}\bigr)>c_\bigr\}\,\mathrm{d}c\end{aligned}$$ converges $_\mathbb{P}_\times\tilde{\mathbb{P} } $-almost surely toward $$\int_0^\infty\ell\bigl(x\in\mathbb{R}\dvtx\tilde{L}^2(1,x)>c \bigr)\,\mathrm{d}c
=\int_{\mathbb{R}}\tilde{L}_\ast^2(1,x)\,\mathrm{d}x.$$ Dominated convergence and Fubini’s theorem imply that_$$w(s)\sum_{x\in\mathbb{Z}}\mathbb{E}_\bigl[\tilde{\Gamma}_{w(s)}^2\bigl(1,\{
x/w(s)\}\bigr)
\bigr]\longrightarrow
\int_{\mathbb{R}}\mathbb{E} [\tilde{L}_\ast^2(1,x)_]\,\mathrm{d}x\qquad
\mbox{as_}s\rightarrow\infty.$$_Therefore, $$w(s)s^{-2}\sum_{x\in\mathbb{Z}}\mathbb{E} [\Gamma^2(s,\{x\})
]\longrightarrow
\int_{\mathbb{R}}\mathbb{E} [\tilde{L}_\ast^2(1,x) ]\,\mathrm{d}x\qquad_\mbox{as }
|
\right) \right](x),$$ and $$\hat q_k (t,\xi) = {\mathcal{F}}[q_k(t,\cdot)](\xi).$$ Then considering the Fourier transform, one can easily check $$\begin{aligned}
\label{pq rel}
\Delta^{\gamma/2}p_k(t,x)= 2^{kd} 2^{k\gamma} q_k (t,2^k x).\end{aligned}$$ Obviously, for all constants $a \geq 0$ and $c>0$, $$\label{ob}
\sup_{x>0} x^{a} e^{-cx}<\infty.$$ Recalling that $\hat \Psi (\xi )$ has the support in $B_2 \setminus B_{1/2}$, differentiating $\hat{q}_k$ for $2d$-times, and using (\[ob\]), Assumptions \[co as\](ii) and \[ini as\](iii), one can find positive constants $N$ and $c$ such that for all $(t,\xi) \in (0,T) \times {\mathbf{R}}^d$, $$\begin{aligned}
\notag
\Big| \hat q_k(t,\xi)\Big|+\Big|\Delta^d_{\xi} \hat q_k(t,\xi)\Big|
&\leq N 1_{(2^{-1},2)}(|\xi|)\exp\left(- 4c \int_0^t \delta(s)ds \cdot |2^k \xi|^2\right).\end{aligned}$$ Thus $$\begin{aligned}
\left|\left( 1+|x|^{2d}\right)q_j(t,x)\right|
&= \left|{\mathcal{F}}^{-1} \left[ \left(1+\Delta^d_\xi \right) \hat q_k(t,\xi)\right](x) \right| \\
&\leq N\sup_{ |\xi| \in (2^{-1},2)}\left| \exp \left(- 4c \int_0^t \delta(s)ds \cdot |2^k \xi|^2 \right) \right| \\
&\leq N\exp \left(- c \int_0^t \delta(s)ds \cdot 2^{2k} \right),\end{aligned}$$ and $$\begin | \right) \right](x),$$ and $ $ \hat q_k (t,\xi) = { \mathcal{F}}[q_k(t,\cdot)](\xi).$$ Then considering the Fourier transform, one can easily check $ $ \begin{aligned }
\label{pq rel }
\Delta^{\gamma/2}p_k(t, x)= 2^{kd } 2^{k\gamma } q_k (t,2^k x).\end{aligned}$$ Obviously, for all constants $ a \geq 0 $ and $ c>0 $, $ $ \label{ob }
\sup_{x>0 } x^{a } e^{-cx}<\infty.$$ echo that $ \hat \Psi (\xi) $ have the support in $ B_2 \setminus B_{1/2}$, differentiating $ \hat{q}_k$ for $ 2d$-times, and using (\[ob\ ]), Assumptions \[co as\](ii) and \[ini as\](iii), one can determine positive constants $ N$ and $ c$ such that for all $ (t,\xi) \in (0,T) \times { \mathbf{R}}^d$, $ $ \begin{aligned }
\notag
\Big| \hat q_k(t,\xi)\Big|+\Big|\Delta^d_{\xi } \hat q_k(t,\xi)\Big|
& \leq N 1_{(2^{-1},2)}(|\xi|)\exp\left(- 4c \int_0^t \delta(s)ds \cdot |2^k \xi|^2\right).\end{aligned}$$ therefore $ $ \begin{aligned }
\left|\left (1+|x|^{2d}\right)q_j(t, x)\right|
& = \left|{\mathcal{F}}^{-1 } \left [ \left(1+\Delta^d_\xi \right) \hat q_k(t,\xi)\right](x) \right| \\
& \leq N\sup _ { |\xi| \in (2^{-1},2)}\left| \exp \left(- 4c \int_0^t \delta(s)ds \cdot |2^k \xi|^2 \right) \right| \\
& \leq N\exp \left(- c \int_0^t \delta(s)ds \cdot 2^{2k } \right),\end{aligned}$$ and $ $ \begin | \rihht) \right](x),$$ and $$\hat q_k (t,\xl) = {\mathcal{F}}[q_k(t,\cdot)](\xi).$$ Thxn consjdering ghe Fourier transform, one cai eawily xheck $$\begin{aligned}
\lqbel{kq rel}
\Delta^{\gamma/2}p_k(t,x)= 2^{kd} 2^{k\ncmma} s_n (t,2^k e).\end{aligned}$$ Obvlously, for dll constants $d \eee 0$ and $c>0$, $$\label{ob}
\sup_{x>0} x^{a} e^{-cx}<\infty.$$ Resalling tjat $\hat \Psi (\xi )$ hax the support in $B_2 \setminus B_{1/2}$, differentjating $\iat{q}_k$ for $2d$-timex, and using (\[ob\]), Assumptions \[co ws\](ii) and \[ini as\](iii), one can fibd pjwitive constxnts $N$ and $c$ such that for all $(t,\xi) \in (0,T) \times {\mathbf{R}}^a$, $$\begnn{aligned}
\notad
\Big| \hat q_k(t,\wp)\Big|+\Big|\Gelta^d_{\xo} \hat q_k(t,\xi)\Big|
&\keq N 1_{(2^{-1},2)}(|\zi|)\exp\left(- 4c \int_0^t \delte(s)ds \cdot |2^k \xi|^2\right).\egd{aligned}$$ Tkus $$\begin{aligned}
\left|\lwfr( 1+|x|^{2d}\rhght)x_j(t,x)\fughg|
&= \ltft|{\kafhcal{F}}^{-1} \leht[ \left(1+\Deltz^d_\xi \right) \yat q_k(t,\xi)\right](x) \rignt| \\
&\oeq N\sup_{ |\xi| \ih (2^{-1},2)}\left| \qxp \left(- 4c \int_0^t \delta(s)ds \cdot |2^k \xi|^2 \righu) \riggt| \\
&\leq N\exp \left(- c \int_0^t \delta(s)ds \cdot 2^{2k} \rigjt),\end{aligged}$$ and $$\begin | \right) \right](x),$$ and $$\hat q_k (t,\xi) = considering Fourier transform, can easily check 2^{k\gamma} (t,2^k x).\end{aligned}$$ Obviously, all constants $a 0$ and $c>0$, $$\label{ob} \sup_{x>0} x^{a} Recalling that $\hat \Psi (\xi )$ has the support in $B_2 \setminus B_{1/2}$, $\hat{q}_k$ for $2d$-times, and using (\[ob\]), Assumptions \[co as\](ii) and \[ini as\](iii), one find constants and such that for all $(t,\xi) \in (0,T) \times {\mathbf{R}}^d$, $$\begin{aligned} \notag \Big| \hat q_k(t,\xi)\Big|+\Big|\Delta^d_{\xi} \hat q_k(t,\xi)\Big| N 1_{(2^{-1},2)}(|\xi|)\exp\left(- 4c \int_0^t \delta(s)ds \cdot |2^k \xi|^2\right).\end{aligned}$$ $$\begin{aligned} \left|\left( 1+|x|^{2d}\right)q_j(t,x)\right| &= \left[ \left(1+\Delta^d_\xi \right) \hat q_k(t,\xi)\right](x) \\ N\sup_{ |\xi| (2^{-1},2)}\left| \left(- \int_0^t \delta(s)ds \cdot \xi|^2 \right) \right| \\ &\leq N\exp \left(- c \int_0^t \delta(s)ds \cdot 2^{2k} \right),\end{aligned}$$ and $$\begin | \right) \right](x),$$ and $$\hat q_k (t,\xi) = {\matHcal{F}}[q_k(t,\cdOt)](\xi).$$ THen ConSiDeriNg thE Fourier transfORm, onE can easily check $$\begin{alIgned}
\LaBEl{pq REl}
\delta^{\Gamma/2}p_k(T,X)= 2^{kD} 2^{K\GamMa} Q_k (T,2^k x).\EnD{AlIgned}$$ obvIously, fOr all constAntS $a \Geq 0$ and $c>0$, $$\label{OB}
\sUp_{x>0} x^{a} e^{-cx}<\inFty.$$ recalling thaT $\haT \Psi (\xi )$ HaS thE SuppoRt iN $B_2 \setMinus B_{1/2}$, DIffereNtiating $\hAt{Q}_K$ for $2d$-tIMes, and uSINg (\[Ob\]), AsSumptions \[co as\](ii) anD \[InI As\](iii), one can finD positIvE CoNSTanTs $N$ And $c$ such thAt For alL $(T,\xi) \in (0,T) \tIMeS {\MAThbF{r}}^d$, $$\begin{aligneD}
\notag
\Big| \haT Q_k(t,\Xi)\Big|+\BIg|\delTA^d_{\xi} \haT q_k(t,\xI)\BIG|
&\leQ N 1_{(2^{-1},2)}(|\xi|)\exp\left(- 4C \int_0^T \delta(s)ds \Cdot |2^k \xI|^2\Right).\enD{Aligned}$$ thus $$\beGin{AliGned}
\LEfT|\lEft( 1+|X|^{2d}\RIghT)Q_j(T,x)\rIGht|
&= \Left|{\mathCaL{F}}^{-1} \Left[ \lEft(1+\DELTA^D_\xi \rIghT) \hat Q_k(t,\xi)\Right](x) \right| \\
&\leQ N\sUp_{ |\xi| \IN (2^{-1},2)}\leFt| \exp \Left(- 4c \Int_0^t \DeLta(s)dS \cdot |2^k \Xi|^2 \rigHt) \Right| \\
&\leq N\exp \lefT(- c \inT_0^t \delta(s)dS \cdOt 2^{2K} \riGhT),\end{aLIgned}$$ aNd $$\bEgiN | \right) \right](x),$$ and $$\hat q_ k (t, \xi ) = { \mat hcal {F}}[q_k(t,\cd o t)]( \xi).$$ Then consideri ng th eF ouri e rtrans form, o n ec a n e as il y c he c k$$\be gin {aligne d}
\lab e l{ pq rel}
\D elt a^{\gamma/2} p_k (t,x)= 2 ^{k d } 2^{ k\g amma} q_k ( t ,2^k x ).\end{al ig n ed}$$O bviousl y , f or a ll constants $a \ g eq 0$ and $c>0$,$$\lab el { ob } \su p_{ x>0} x^{a} e ^{-cx } <\infty . $$ R e cal l ing that $\ha t \Psi (\xi )$has th esup p ort in $B_2 \ s etm inus B_{1/2 }$,different iating $\hat{q } _k$ for $2d$- tim es, and us in g ( \[ o b\] ) ,Ass u mpt ions \[c oas \](ii ) an d \ [ inias\ ](ii i), o ne can find p osi tive con stant s $N$ and $ c$ su ch tha t for a ll $(t,\xi) \in (0, T) \times {\ ma thb f{ R}}^d $ , $$\b egi n{a ligned}
\notag
\Big| \h at q _k (t,\xi)\ Big|+\ B ig |\ D elta^d_{ \x i}\hat q _k(t, \xi) \ Bi g|
&\leq N 1_{ ( 2^ {- 1},2)}( |\ xi|)\e xp \le ft( - 4c\ int_ 0^t \d elta(s)d s \cd o t |2^k \xi|^2\ r ight).\end{al i gn e d }$ $ Thu s $ $\begin{ali gned }
\le ft|\ l ef t(1 +|x|^ {2d}\ ri g ht ) q_j(t,x)\right|
&=\l eft|{\ mathc al{F}}^{-1} \ left[ \lef t ( 1 +\Delta^ d_\x i \ r ight) \hat q_k (t,\x i)\right]( x ) \right | \\&\leq N\ sup_{ |\x i | \in (2^ {-1 },2 )}\ lef t | \exp \left(-4 c \in t_ 0^t \de lta (s)ds \ cdo t | 2^k \x i| ^2 \right ) \right |\\
& \l eqN\exp \left(-c\in t_ 0^t \del t a(s)ds \cdo t 2^ {2 k} \r ight),\ e nd { a lign ed }$ $ an d $ $\ begin | \right)_\right](x),$$ and_$$\hat q_k (t,\xi) =_{\mathcal{F}}[q_k(t,\cdot)](\xi).$$ Then_considering_the Fourier_transform,_one can easily_check $$\begin{aligned}
_ _ __ \label{pq rel}
\Delta^{\gamma/2}p_k(t,x)= 2^{kd} 2^{k\gamma}_q_k_(t,2^k x).\end{aligned}$$_Obviously,_for_all constants $a \geq 0$_and $c>0$, $$\label{ob}
\sup_{x>0} x^{a} e^{-cx}<\infty.$$_Recalling that_$\hat \Psi (\xi )$ has the support in_$B_2_\setminus B_{1/2}$, differentiating_$\hat{q}_k$ for $2d$-times, and using (\[ob\]), Assumptions \[co as\](ii)_and \[ini as\](iii), one can find_positive constants $N$_and_$c$_such that for all_$(t,\xi) \in (0,T) \times {\mathbf{R}}^d$, $$\begin{aligned}
_ _ _ _\notag
\Big| \hat q_k(t,\xi)\Big|+\Big|\Delta^d_{\xi} \hat q_k(t,\xi)\Big|
&\leq_N 1_{(2^{-1},2)}(|\xi|)\exp\left(-_4c \int_0^t \delta(s)ds \cdot |2^k_\xi|^2\right).\end{aligned}$$ Thus $$\begin{aligned}
\left|\left(_1+|x|^{2d}\right)q_j(t,x)\right|
&= \left|{\mathcal{F}}^{-1}_\left[ \left(1+\Delta^d_\xi \right)_\hat q_k(t,\xi)\right](x) \right| \\
&\leq N\sup_{ |\xi|_\in (2^{-1},2)}\left| _\exp \left(- 4c \int_0^t \delta(s)ds \cdot_|2^k_\xi|^2 \right) \right|_\\
&\leq_N\exp_\left(- c_\int_0^t \delta(s)ds \cdot_2^{2k}_ \right),\end{aligned}$$_and_$$\begin |
.073) – (3.667, 0.812) – (3.219, 1.630) – (2.873, 2.419) – (2.655, 3.212); (3.121, 0.164) – (2.576, 0.979) – (2.127, 1.752) – (1.818, 2.542); (2.068, 0.255) – (1.499, 1.004) – (1.090, 1.775); (1.022, 0.199) – (0.482, 0.924);
(+ 1,0) – (.5\*+.5, /2 + /2);
(0,0) – (+,0) – (+, c+ ) – cycle;
(- 1, /2) node\[above\][$\Omega$]{}; (/2 +.5, 0) node\[below\][$\Gamma_1$]{}; (/4 +.25, /4 + /4) node\[anchor = north west\][$\Gamma_2$]{}; (- /4 + 1.75, /4 + /4) node\[anchor = north east\][$\Gamma_3$]{}; (2\*+.5, 0) node\[above\] [$\phi$]{};
(0, 0) – (+ 1, 0) node\[below\] [$e_1$]{}; (0, 0) – (0.5\*+ 0.5, 0.5\*+ 0.5\*) node\[above\] [$R_{\pi/3} e_1$]{}; (+ 1, 0) – (0.5\*+ 0.5, 0.5\*+ 0.5\*);
(0, -0.3) – (1, -0.3) node\[midway, below\][$\varepsilon$]{};
(0, 0.3) – (+ 1, 0.3) node\[midway, anchor = south east\][$1$]{};
(.8, /2) node\[right\][$\mathcal L_\varepsilon \cong \mathcal G_\varepsilon = (\mathcal V_\varepsilon, \mathcal E_\varepsilon, \mathcal T_\varepsilon)$]{}; (+ 1, /2) – (2\*, /2) node\[midway, above\][$u_\varepsilon$]{}; (3\*, /2) node\[right\][$u_\varepsilon (\mathcal L_\varepsilon )$]{};
The | .073) – (3.667, 0.812) – (3.219, 1.630) – (2.873, 2.419) – (2.655, 3.212); (3.121, 0.164) – (2.576, 0.979) – (2.127, 1.752) – (1.818, 2.542); (2.068, 0.255) – (1.499, 1.004) – (1.090, 1.775); (1.022, 0.199) – (0.482, 0.924);
(+ 1,0) – (.5\*+.5, /2 + /2);
(0,0) – (+, 0) – (+, c+) – cycle;
(- 1, /2) node\[above\][$\Omega$ ] { }; (/2 + .5, 0) node\[below\][$\Gamma_1 $ ] { }; (/4 + .25, /4 + /4) node\[anchor = north west\][$\Gamma_2 $ ] { }; (- /4 + 1.75, /4 + /4) node\[anchor = north east\][$\Gamma_3 $ ] { }; (2\*+.5, 0) node\[above\ ] [ $ \phi$ ] { };
(0, 0) – (+ 1, 0) node\[below\ ] [ $ e_1 $ ] { }; (0, 0) – (0.5\*+ 0.5, 0.5\*+ 0.5\ *) node\[above\ ] [ $ R_{\pi/3 } e_1 $ ] { }; (+ 1, 0) – (0.5\*+ 0.5, 0.5\*+ 0.5\ *);
(0, -0.3) – (1, -0.3) node\[midway, below\][$\varepsilon$ ] { };
(0, 0.3) – (+ 1, 0.3) node\[midway, anchor = south east\][$1 $ ] { };
(.8, /2) node\[right\][$\mathcal L_\varepsilon \cong \mathcal G_\varepsilon = (\mathcal V_\varepsilon, \mathcal E_\varepsilon, \mathcal T_\varepsilon)$ ] { }; (+ 1, /2) – (2\ *, /2) node\[midway, above\][$u_\varepsilon$ ] { }; (3\ *, /2) node\[right\][$u_\varepsilon (\mathcal L_\varepsilon) $ ] { };
The | .073) – (3.667, 0.812) – (3.219, 1.630) – (2.873, 2.419) – (2.655, 3.212); (3.121, 0.164) – (2.576, 0.979) – (2.127, 1.752) – (1.818, 2.542); (2.068, 0.255) – (1.499, 1.004) – (1.090, 1.775); (1.022, 0.199) – (0.482, 0.924);
(+ 1,0) – (.5\*+.5, /2 + /2);
(0,0) – (+,0) – (+, c+ ) – cycls;
(- 1, /2) node\[xbove\][$\Omega$]{}; (/2 +.5, 0) node\[below\][$\Gamma_1$]{}; (/4 +.25, /4 + /4) nide\[anchor = north west\][$\Gxmma_2$]{}; (- /4 + 1.75, /4 + /4) node\[abchoc = north east\][$\Gamma_3$]{}; (2\*+.5, 0) nodc\[cbove\] [$\ihi$]{};
(0, 0) – (+ 1, 0) node\[below\] [$e_1$]{}; (0, 0) – (0.5\*+ 0.5, 0.5\*+ 0.5\*) noge\[above\] [$R_{\pi/3} e_1$]{}; (+ 1, 0) – (0.5\*+ 0.5, 0.5\*+ 0.5\*);
(0, -0.3) – (1, -0.3) node\[midway, below\][$\varepsilon$]{};
(0, 0.3) – (+ 1, 0.3) noce\[lidway, anchor = souuh qast\][$1$]{};
(.8, /2) node\[right\][$\mathcal L_\varepsilon \conf \mathcel G_\varepsilon = (\mathcal V_\varepsilon, \mathcwl E_\garepsilon, \mathcal T_\varepsilob)$]{}; (+ 1, /2) – (2\*, /2) node\[midwah, above\][$u_\vagzpsilon$]{}; (3\*, /2) nkde\[right\][$u_\varepsilon (\mathcal L_\vafepsikon )$]{};
The | .073) – (3.667, 0.812) – (3.219, 1.630) 2.419) (2.655, 3.212); 0.164) – (2.576, (1.818, (2.068, 0.255) – 1.004) – (1.090, (1.022, 0.199) – (0.482, 0.924); (+ – (.5\*+.5, /2 + /2); (0,0) – (+,0) – (+, c+ ) – (- 1, /2) node\[above\][$\Omega$]{}; (/2 +.5, 0) node\[below\][$\Gamma_1$]{}; (/4 +.25, /4 + /4) = west\][$\Gamma_2$]{}; /4 1.75, /4 + /4) node\[anchor = north east\][$\Gamma_3$]{}; (2\*+.5, 0) node\[above\] [$\phi$]{}; (0, 0) – (+ 0) node\[below\] [$e_1$]{}; (0, 0) – (0.5\*+ 0.5, 0.5\*) node\[above\] [$R_{\pi/3} e_1$]{}; 1, 0) – (0.5\*+ 0.5, 0.5\*); -0.3) – -0.3) below\][$\varepsilon$]{}; 0.3) – (+ 0.3) node\[midway, anchor = south east\][$1$]{}; (.8, /2) node\[right\][$\mathcal L_\varepsilon \cong \mathcal G_\varepsilon = (\mathcal V_\varepsilon, \mathcal \mathcal T_\varepsilon)$]{}; /2) – /2) above\][$u_\varepsilon$]{}; /2) node\[right\][$u_\varepsilon (\mathcal The | .073) – (3.667, 0.812) – (3.219, 1.630) – (2.873, 2.419) – (2.655, 3.212); (3.121, 0.164) – (2.576, 0.979) – (2.127, 1.752) – (1.818, 2.542); (2.068, 0.255) – (1.499, 1.004) – (1.090, 1.775); (1.022, 0.199) – (0.482, 0.924);
(+ 1,0) – (.5\*+.5, /2 + /2);
(0,0) – (+,0) – (+, c+ ) – cycle;
(- 1, /2) node\[above\][$\Omega$]{}; (/2 +.5, 0) node\[bElow\][$\Gamma_1$]{}; (/4 +.25, /4 + /4) nOde\[anChoR = noRtH wesT\][$\GamMa_2$]{}; (- /4 + 1.75, /4 + /4) node\[anchor = noRTh eaSt\][$\Gamma_3$]{}; (2\*+.5, 0) node\[above\] [$\phi$]{};
(0, 0) – (+ 1, 0) nodE\[beloW\] [$e_1$]{}; (0, 0) – (0.5\*+ 0.5, 0.5\*+ 0.5\*) NOde\[aBOvE\] [$R_{\pi/3} e_1$]{}; (+ 1, 0) – (0.5\*+ 0.5, 0.5\*+ 0.5\*);
(0, -0.3) – (1, -0.3) Node\[midWAy, BELow\][$\VaRePsiLoN$]{};
(0, 0.3) – (+ 1, 0.3) NoDe\[midWay, Anchor = sOuth east\][$1$]{};
(.8, /2) noDe\[rIgHt\][$\mathcal L_\vaREpSilon \cong \mAthCal G_\varepsilOn = (\mAthcal v_\vArePSilon, \MatHcal E_\VarepsILon, \matHcal T_\varePsILon)$]{}; (+ 1, /2) – (2\*, /2) nodE\[Midway, aBOVe\][$U_\varEpsilon$]{}; (3\*, /2) node\[right\][$u_\VArEPsilon (\mathcal L_\VarepsIlON )$]{};
THE | .073) – (3.667, 0.812) – ( 3.219, 1.6 30) – (2 .87 3, 2.4 19)– (2.655, 3.21 2 ); ( 3.121, 0.164) – (2.576 , 0.9 79 ) – ( 2 .1 27, 1 .752) – (1 . 8 18, 2 .5 42) ;( 2. 068,0.2 55) – ( 1.499, 1.0 04) – (1.090, 1.7 7 5) ; (1.022,0.1 99) – (0.482 , 0 .924);
(+1 ,0) – (. 5\*+. 5, /2+ /2);
(0,0) –(+ , 0) – ( + , c+ )– cy cle;
(- 1, /2) node\ [ ab o ve\][$\Omega$] {}; (/ 2+ .5 , 0)nod e\[below\] [$ \Gamm a _1$]{}; (/ 4 + .25 , /4 + /4) nod e\[anchor = nor th wes t\ ][$ \ Gamma_ 2$]{} ;( - / 4 + 1.75, / 4 +/4) node\ [ancho r = nort h east\] [$\Gam ma_ 3$] {};( 2\ *+ .5, 0 ) no d e\ [ab o ve\ ] [$\phi $] {} ;
(0 , 0) – ( + 1, 0) nod e\[be low\] [$e_1$] {}; (0, 0)– (0. 5\*+0.5, 0 .5\*+ 0.5\* ) nod e\ [above\] [$R_{\ pi/3 } e_1$]{} ; ( +1,0) – (0 . 5\*+ 0 .5, 0. 5\*+ 0. 5\*);
( 0,-0 . 3 ) – (1, -0.3) node\[m id w a y, below\] [$\var e ps il o n$]{};
(0 , 0 .3)– (+ 1, 0.3 ) n ode\[mid way, a n ch or = sout heast\] [$ 1$] {};
(.8 , /2) node\ [right\] [$\ma t hcal L_\vareps i lon \cong \ma t hc a l G _ \var eps ilon = (\ma thca l V_\ vare p si lon , \mat hcalE_ \ va r epsilon, \mathcal T _\ vareps ilon) $]{}; (+ 1, / 2) – (2\*, / 2 ) node\[ midw a y, above\][$u_\va repsi lon$]{}; ( 3 \*, /2)node\ [right\] [$u_\vare p s ilon (\m ath cal L_ \va r e ps ilon )$]{};
T h e | .073) –_(3.667, 0.812)_– (3.219, 1.630) –_(2.873, 2.419)_–_(2.655, 3.212);_(3.121,_0.164) – (2.576,_0.979) – (2.127,_1.752) – (1.818, 2.542);_(2.068, 0.255) –_(1.499,_1.004) – (1.090, 1.775); (1.022, 0.199) – (0.482, 0.924);
(+ 1,0) – (.5\*+.5, /2 +_/2);
(0,0)_– (+,0)_–_(+,_c+ ) – cycle;
(- 1,_/2) node\[above\][$\Omega$]{}; (/2 +.5, 0)_node\[below\][$\Gamma_1$]{}; (/4_+.25, /4 + /4) node\[anchor = north west\][$\Gamma_2$]{};_(-_/4 + 1.75,_/4 + /4) node\[anchor = north east\][$\Gamma_3$]{}; (2\*+.5, 0)_node\[above\] [$\phi$]{};
(0, 0) – (+ 1,_0) node\[below\] [$e_1$]{};_(0,_0)_– (0.5\*+ 0.5, 0.5\*+_0.5\*) node\[above\] [$R_{\pi/3} e_1$]{}; (+ 1,_0) – (0.5\*+ 0.5, 0.5\*+ 0.5\*);
(0,_-0.3) – (1, -0.3) node\[midway, below\][$\varepsilon$]{};
(0, 0.3)_– (+ 1, 0.3) node\[midway, anchor_= south east\][$1$]{};
(.8, /2) node\[right\][$\mathcal_L_\varepsilon \cong_\mathcal G_\varepsilon = (\mathcal V_\varepsilon,_\mathcal E_\varepsilon, \mathcal_T_\varepsilon)$]{}; (+_1, /2) –_(2\*, /2) node\[midway, above\][$u_\varepsilon$]{}; (3\*, /2)_node\[right\][$u_\varepsilon (\mathcal L_\varepsilon_)$]{};
The |
t_0) = \psi_id_0(t) + \underline{d}_1(t)$$ where the capture state is zero at $t_0$ and increases in time. It is a ‘ready’ state as prescribed by nRule (1), for the particle goes discontinuously and irreversibility from being spread throughout space in the first component to being inside the detector in the second component – like going discontinuously from the $1^{st}$ to the $2^{nd}$ orbit of an atom.
The state $\psi(t)$ is an incoming free particle plus all of the scattered components that are correlated with recoil states of the ground state detector. The second component is a superposition that includes all of the recoil components of the detector that have captured the particle[^2].
As probability current flows into the second component in Eq. 2, it may take a stochastic hit at some time $t_{sc}$ following nRule (2). In that case the state will collapse following nRule (3) to yield $$\Phi(t \ge t_{sc} > t_0) = d_1(t)$$ where $d_1$ becomes a realized state and the first component disappears. Renormalization is not necessary because the current flow into $\underline{d}_1(t)$ is continuously normalized according to nRule (2). It is explained in Ref. 1 how to treat this case when there is no capture of the particle.
So far there are no difficulties with these nRules; however, difficulties arise when there are three or more components as in the case of a particle counter.
A Counter {#a-counter.unnumbered}
=========
If a particle counter is exposed to a radioactive source its state function would normally be written $$\Phi(t \ge t_0) = C_0(t) + C_1(t) + C_2(t) + C_3(t) + etc.$$ where $C_0$ is the initial state of the counter (registering 0 counts) at time $t_0$. The state $C_1$ registers 1 count, $C_2$ registers 2 counts, etc. The radioactive source and the particle field are not shown in Eq. 3. The components following $C_0$ are zero at $t_0$; and after that, they become non-zero in the form of a pulse that travels from left | t_0) = \psi_id_0(t) + \underline{d}_1(t)$$ where the capture state is zero at $ t_0 $ and increases in time. It is a ‘ ready ’ department of state as order by nRule (1), for the particle goes discontinuously and irreversibility from being go around throughout distance in the first component to being inside the detector in the second part – like going discontinuously from the $ 1^{st}$ to the $ 2^{nd}$ orbit of an atom.
The state $ \psi(t)$ is an incoming complimentary particle plus all of the scattered component that are correlated with recoil states of the flat coat state detector. The second component is a superposition that includes all of the recoil components of the detector that have capture the particle[^2 ].
As probability current flows into the second component in Eq. 2, it may take a stochastic score at some time $ t_{sc}$ following nRule (2). In that case the state will break down following nRule (3) to yield $ $ \Phi(t \ge t_{sc } > t_0) = d_1(t)$$ where $ d_1 $ becomes a realized state and the inaugural component disappears. Renormalization is not necessary because the current flow into $ \underline{d}_1(t)$ is continuously normalized according to nRule (2). It is explained in Ref. 1 how to treat this case when there is no capture of the particle.
So far there are no difficulties with these nRules; however, difficulties arise when there are three or more component as in the case of a particle counterpunch.
A Counter { # a-counter.unnumbered }
= = = = = = = = =
If a particle counter is exposed to a radioactive source its department of state function would normally be written $ $ \Phi(t \ge t_0) = C_0(t) + C_1(t) + C_2(t) + C_3(t) + etc.$$ where $ C_0 $ is the initial state of the counter (register 0 counts) at time $ t_0$. The state $ C_1 $ registers 1 count, $ C_2 $ registers 2 counts, etc. The radioactive generator and the particle field are not shown in Eq. 3. The components following $ C_0 $ are zero at $ t_0 $; and after that, they become non - zero in the form of a pulsation that travel from left | t_0) = \psi_id_0(t) + \underline{d}_1(t)$$ wheve the capture sjare is vero at $t_0$ and ivcreases in time. It is a ‘reavy’ srate qs prescribed by nRule (1), for the particlw gots discontinuouslb and irvzversjnilitv hrom being spresd throughmut space in tve fnrst component to being inside the dqtector ij the second cjmpomqnt – like going discontinuously from tge $1^{st}$ tm the $2^{nd}$ orbiy of an atom.
The state $\psi(t)$ is wn incoming free pwrticle pluw alj of the scatgered components that zre correlated with recoil statds of the groune wtahg detector. Tie secjnd componenb is a vuperpoxition that ingludev aol of the recoil compmnents of the detestor that hcve captured the partucoe[^2].
As krobatiligt cjrrtnt fmows ijto the secons component in Eq. 2, it may take s fnpchastic hit at soie time $t_{sc}$ following nRule (2). In that case ths state will collapse fillowing nRule (3) to yigld $$\Phi(t \gq t_{sc} > t_0) = d_1(t)$$ where $d_1$ becomes a realized state and the hifst cjopojent disappears. Renormalization is not necesswdy bvcause the currenb flow into $\underlonf{d}_1(y)$ is continuourly noxjamized according to nRule (2). It iw explaintd in Ref. 1 how to treat this case when there ps ni capture of the pcrticle.
So fax therg are mo difficulties with thzse nRhles; howeveg, difficumgies arise when ghege ase three or more componentf as in tie caxe of a parjicle cjunter.
A Coknter {#a-counter.unnumbered}
=========
Lf a karticne counter is exposed to a radioactive soncce its state fgncnion woulb normslly be writtqn $$\Phi(t \ge t_0) = E_0(t) + C_1(t) + E_2(t) + C_3(t) + etc.$$ wherv $C_0$ is thx initial stwte of the cogjter (registecing 0 cougts) qt tume $t_0$. Tfd state $C_1$ regixters 1 colnu, $C_2$ registwrs 2 counts, etc. Thc radkkactive source cud the particle foela awe npt vhown in Eq. 3. Dhe zomopnentr following $Z_0$ arr zero at $t_0$; and aftes thzt, they become non-aevo in the form of a pulse that travels from left | t_0) = \psi_id_0(t) + \underline{d}_1(t)$$ where the is at $t_0$ increases in time. as by nRule (1), the particle goes and irreversibility from being spread throughout in the first component to being inside the detector in the second component like going discontinuously from the $1^{st}$ to the $2^{nd}$ orbit of an atom. state is incoming particle plus all of the scattered components that are correlated with recoil states of the ground detector. The second component is a superposition that all of the recoil of the detector that have the As probability flows the component in Eq. it may take a stochastic hit at some time $t_{sc}$ following nRule (2). In that case the will collapse (3) to $$\Phi(t t_{sc} t_0) = d_1(t)$$ becomes a realized state and the Renormalization is not necessary because the current flow $\underline{d}_1(t)$ is normalized according to nRule (2). It explained in Ref. 1 how to treat this when there is no capture of the particle. So far there are no difficulties with however, difficulties arise when are three or components in case a particle A Counter {#a-counter.unnumbered} ========= If a particle counter is exposed to radioactive source its state function would normally be written $$\Phi(t = + C_1(t) + + C_3(t) + etc.$$ $C_0$ the initial state of (registering counts) The $C_1$ 1 count, $C_2$ registers counts, etc. The radioactive source the particle field are The components following $C_0$ are zero at $t_0$; after that, they become non-zero in the of a pulse that travels from left | t_0) = \psi_id_0(t) + \underline{d}_1(t)$$ where thE capture stAte is ZerO at $T_0$ aNd inCreaSes in time. It is a ‘REady’ State as prescribed by nRuLe (1), for ThE PartIClE goes DiscontINuOUSly AnD iRreVeRSiBilitY frOm being Spread throUghOuT space in the fIRsT component To bEing inside thE deTector In The SEcond ComPonenT – like gOIng disContinuouSlY From thE $1^{St}$ to the $2^{ND}$ OrBit oF an atom.
The state $\psI(T)$ iS An incoming free ParticLe PLuS ALl oF thE scattered CoMponeNTs that aRE cORRElaTEd with recoil sTates of the gROunD state DeTecTOr. The sEcond CoMPonEnt is a superPosiTion that iNcludeS All of thE Recoil cOmponeNts Of tHe deTEcToR thAt HAve CApTurED thE particlE[^2].
AS pRobabIlitY CURRent FloWs inTo the Second componeNt iN Eq. 2, iT May Take a StochAstiC hIt at sOme timE $t_{sc}$ fOlLowing nRule (2). In thAt caSe the statE wiLl ColLaPse foLLowing NRuLe (3) tO yield $$\PHi(t \ge t_{sC} > T_0) = d_1(t)$$ WhERE $D_1$ bEcomes a realized staTe AND tHe first cOmponeNT dIsAPpears. ReNoRmaLizaTIOn is nOt neCEsSary becaUse the CUrReNt flow iNtO $\underLiNe{d}_1(T)$ is ContiNUousLy normAlized acCordiNG to nRule (2). It is exPLained in Ref. 1 hoW To TREaT This CasE when there iS no cAPturE of tHE pArtICle.
So Far thErE ArE No difficulties with tHeSe nRulEs; howEver, difficultIes arise whEN THere are tHree OR mORe components as In the Case of a parTIcle counTer.
A COunter {#a-cOunter.unnUMBered}
=========
If a ParTicLe cOunTER iS exposed to a raDIOactIvE source Its State fuNctIon WouLd nOrMally be wrItten $$\Phi(T \gE t_0) = c_0(t) + c_1(t) + c_2(t) + C_3(T) + etc.$$ wHEre $C_0$ is thE iNitIaL stAte of THe counTer (reGistErInG 0 CouNts) at tiME $t_0$. tHE staTe $c_1$ rEgisTerS 1 cOunt, $C_2$ RegiSTerS 2 counts, Etc. The radIoaCTive SoUrCe and thE particle fielD aRe not shown In eq. 3. THe compONEnts follOwing $C_0$ are zero at $t_0$; and afteR That, theY beCome nOn-zeRo in the foRm oF a pulsE thAT traveLs from Left | t_0) = \psi_id_0(t) + \un derline{d} _1(t) $$whe re the cap ture state isz eroat $t_0$ and increases in t im e . It is a ‘r eady’ s t at e aspr es cri be d b y nRu le(1), fo r the part icl egoes discont i nu ously andirr eversibility fr om bei ng sp r ead t hro ughou t spac e in th e first c om p onentt o being i ns idethe detector in t h es econd componen t – li ke go i n g d isc ontinuousl yfromt he $1^{ s t} $ t o t h e $2^{nd}$ or bit of an a t om.
Thest ate $\psi( t)$ i sa n i ncoming fre e pa rticle pl us all of thes cattere d comp one nts tha t a re co rr e lat e dwit h re coil sta te sof th e gr o u n d sta tedete ctor. The second c omp onen t is a su perpo siti on that inclu des a ll of the recoilcomp onents of th edet ec tor t h at hav e c apt ured th e parti c le[ ^2 ] .
A s probability curr en t fl ows into the s e co nd componen tinEq.2 , it m ay t a ke a stoch astich it a t someti me $t_ {s c}$ fo llowi n g nR ule (2 ). In th at ca s e the state wi l l collapse fo l lo w i ng nRul e ( 3) to yield $$\ P hi(t \ge t_ {sc } > t_ 0) =d_ 1 (t ) $$ where $d_1$ beco me s a re alize d state and t he first c o m p onent di sapp e ar s . Renormalizat ion i s not nece s sary bec ausethe curr ent flowi n to $\und erl ine {d} _1( t ) $is continuous l y nor ma lized a cco rding t o n Rul e ( 2). I t is expl ained in R ef .1how to t r eat this c ase w hen ther e is no capt ureof t h e p article .
S o far t he re a reno diff icul t ies with t hese nRul es; howe ve r, diffic ulties arisewh en there a re th ree or m ore comp onents as in the case o f a part icl e cou nter .
A Coun ter {#a-c oun t er.unn umbere d}
== == === = =
Ifa pa rti cl e counteri s ex posed t o aradioac tive source its st a tefunction woul d n orma l l ybew ri t ten $ $ \Ph i ( t \ge t_0) = C_ 0(t) + C_1 (t ) + C_2(t) +C _3( t) + etc. $$ wher e $C_ 0 $ is th e initial state of t he c o u nte r (registe ring 0 c ounts) at time$ t_ 0$. T hestate$C _1$ regi sters1 co unt,$C_2$re gister s 2 c ou nts, etc . The radioactive sourc e andthe p art icle fiel d a r e n ot shownin E q. 3. Thecom pon entsfol l owing $C_ 0 $are zeroat $ t _0$; anda ft ert h at , they beco m e non -zero in the fo rm o f a pulse that tr a vels from left | t_0)_= \psi_id_0(t)_+ \underline{d}_1(t)$$ where the_capture state_is_zero at_$t_0$_and increases in_time. It is_a ‘ready’ state as_prescribed by nRule_(1),_for the particle goes discontinuously and irreversibility from being spread throughout space in the_first_component to_being_inside_the detector in the second_component – like going discontinuously_from the_$1^{st}$ to the $2^{nd}$ orbit of an atom.
The_state_$\psi(t)$ is an_incoming free particle plus all of the scattered components_that are correlated with recoil states_of the ground_state_detector._The second component is_a superposition that includes all of_the recoil components of the detector_that have captured the particle[^2].
As probability current_flows into the second component in_Eq. 2, it may take a_stochastic hit_at some time $t_{sc}$ following_nRule (2). In_that case_the state will_collapse following nRule (3) to yield_$$\Phi(t \ge t_{sc}_> t_0) = d_1(t)$$ where $d_1$_becomes_a realized state_and_the_first component_disappears. Renormalization is_not_necessary because_the_current flow into $\underline{d}_1(t)$ is continuously_normalized_according to nRule (2). It is explained_in Ref. 1 how to_treat_this case when there_is no capture of the_particle.
So far there are no difficulties_with these_nRules; however,_difficulties arise when there are three or more components as in_the case of a particle counter.
A_Counter {#a-counter.unnumbered}
=========
If a particle_counter is_exposed_to a radioactive_source_its state_function would normally be written $$\Phi(t \ge_t_0) =_C_0(t) + C_1(t) + C_2(t) +_C_3(t) + etc.$$ where_$C_0$_is the initial state of the_counter (registering 0 counts) at time_$t_0$. The state $C_1$ registers_1_count,_$C_2$ registers 2 counts, etc._The radioactive source and the particle_field are not_shown in Eq. 3. The components following $C_0$_are_zero at $t_0$; and after that,_they_become non-zero in the form of_a_pulse_that travels from left |
, $1$ for $t>0$ and $0$ for $t<0$. Thus, these are nontrivial and $\theta$-dependent for the finite time interval $t_1 < t < t_2$. With the Hamiltonian in Eq. (\[H\]), the equations of motion are obtained as $$\begin{aligned}
& & \dot\chi = \{ \chi, H \}_{\mathrm{MPB}} = -2\kappa P_\chi, \quad
\dot\Omega = \{ \Omega, H \}_{\mathrm{MPB}} = 2\kappa P_\Omega,
\label{1st:x_non}\\
& & \dot{P}_\chi = \{ P_\chi, H \}_{\mathrm{MPB}} =2\kappa \theta_2
P_\Omega, \quad \dot{P}_\Omega = \{ P_\Omega, H \}_{\mathrm{MPB}}
= 2\kappa\theta_2 P_\chi. \label{1st:p_non}\end{aligned}$$ Note that the momenta are no longer constants of the motion because of the nonvanishing $\theta_2$; hence, a new set of equations of motion are obtained from Eqs. (\[1st:x\_non\]) and (\[1st:p\_non\]): $$\label{eom:non}
\ddot\chi = -2\kappa \theta_2 \dot\Omega, \quad \ddot\Omega =
-2\kappa \theta_2 \dot\chi.$$ The solutions for the above coupled equations of motion are easily obtained as $$\begin{aligned}
\Omega &=& \alpha e^{2\kappa\theta_2 t} + \beta
e^{-2\kappa\theta_2 t} + A, \label{sol_gen:Omega_non} \\
\chi &=& \alpha e^{2\kappa\theta_2 t} - \beta
e^{-2\kappa\theta_2 t} + B, \label{sol_gen:chi_non}\end{aligned}$$ where $\alpha$, $\beta$, $A$, and $B$ are constants, and they should satisfy the constraint equation in Eq. (\[con\]), $$\label{constr_gen:non}
\kappa t_\pm = \kappa^2 \theta_2^2 | , $ 1 $ for $ t>0 $ and $ 0 $ for $ t<0$. Thus, these are nontrivial and $ \theta$-dependent for the finite time interval $ t_1 < t < t_2$. With the Hamiltonian in Eq. (\[H\ ]), the equations of apparent motion are prevail as $ $ \begin{aligned }
& & \dot\chi = \ { \chi, H \}_{\mathrm{MPB } } = -2\kappa P_\chi, \quad
\dot\Omega = \ { \Omega, H \}_{\mathrm{MPB } } = 2\kappa P_\Omega,
\label{1st: x_non}\\
& & \dot{P}_\chi = \ { P_\chi, H \}_{\mathrm{MPB } } = 2\kappa \theta_2
P_\Omega, \quad \dot{P}_\Omega = \ { P_\Omega, H \}_{\mathrm{MPB } }
= 2\kappa\theta_2 P_\chi. \label{1st: p_non}\end{aligned}$$ Note that the momenta are no longer constants of the movement because of the nonvanishing $ \theta_2 $; hence, a new stage set of equations of motion are obtain from Eqs. (\[1st: x\_non\ ]) and (\[1st: p\_non\ ] ): $ $ \label{eom: non }
\ddot\chi = -2\kappa \theta_2 \dot\Omega, \quad \ddot\Omega =
-2\kappa \theta_2 \dot\chi.$$ The solutions for the above coupled equation of motion are easily prevail as $ $ \begin{aligned }
\Omega & = & \alpha e^{2\kappa\theta_2 t } + \beta
e^{-2\kappa\theta_2 t } + A, \label{sol_gen: Omega_non } \\
\chi & = & \alpha e^{2\kappa\theta_2 t } - \beta
e^{-2\kappa\theta_2 t } + B, \label{sol_gen: chi_non}\end{aligned}$$ where $ \alpha$, $ \beta$, $ A$, and $ B$ are constants, and they should satisfy the constraint equation in Eq. (\[con\ ]), $ $ \label{constr_gen: non }
\kappa t_\pm = \kappa^2 \theta_2 ^ 2 | , $1$ flr $t>0$ and $0$ for $t<0$. Thus, there are nontrivico and $\vheta$-delendent wor the finite time interval $t_1 < t < u_2$. With the Hamiltonkan in Eq. (\[J\]), the eqyatiibs of motion are oncaines as $$\yejin{aligned}
& & \dpt\chi = \{ \chh, H \}_{\mathrm{MPB}} = -2\kxp'a P_\chi, \quad
\dot\Omega = \{ \Omega, R \}_{\mathrk{MOB}} = 2\kappa P_\Omeda,
\mabel{1st:x_non}\\
& & \dot{P}_\chi = \{ P_\chi, H \}_{\mzthrm{MPU}} =2\kappa \theta_2
P_\Omega, \quad \dot{P}_\Omega = \{ P_\Olega, H \}_{\mathrm{MPB}}
= 2\kappa\jgetw_2 P_\chi. \label{1sg:p_non}\end{alpyned}$$ Note tgat the momenta are no longer cunstauts of the norioj because of the gonvanishing $\theta_2$; hance, a mew set of equstimns of motion are obtainxd from Eqs. (\[1st:x\_non\]) anq (\[1st:p\_non\]): $$\nayel{eom:non}
\ddot\chi = -2\kqppa \thgta_2 \dmt\Omdta, \duas \vdof\Omega =
-2\iappa \thetz_2 \dot\chi.$$ Thw solutions for the afive coupled esuatiogs of motion are easily obtained as $$\begin{dlifned}
\Omega &=& \alpha e^{2\kappa\theta_2 t} + \beta
e^{-2\ka[pa\theta_2 t} + A, \label{sol_gen:Omega_non} \\
\chi &=& \alpha e^{2\kdppa\tiega_2 u} - \beta
e^{-2\kappa\theta_2 t} + B, \label{sol_gen:chi_non}\enq{zlogked}$$ where $\alpha$, $\neta$, $A$, and $B$ are cpndtsgts, and they rhould safisfy the constraijt equajion ib Eq. (\[con\]), $$\lwbel{vonstr_gen:non}
\kappa t_\pm = \kappa^2 \theta_2^2 | , $1$ for $t>0$ and $0$ for these nontrivial and for the finite < With the Hamiltonian Eq. (\[H\]), the of motion are obtained as $$\begin{aligned} & \dot\chi = \{ \chi, H \}_{\mathrm{MPB}} = -2\kappa P_\chi, \quad \dot\Omega = \Omega, H \}_{\mathrm{MPB}} = 2\kappa P_\Omega, \label{1st:x_non}\\ & & \dot{P}_\chi = \{ P_\chi, \}_{\mathrm{MPB}} \theta_2 \quad = \{ P_\Omega, H \}_{\mathrm{MPB}} = 2\kappa\theta_2 P_\chi. \label{1st:p_non}\end{aligned}$$ Note that the momenta are no longer of the motion because of the nonvanishing $\theta_2$; a new set of of motion are obtained from (\[1st:x\_non\]) (\[1st:p\_non\]): $$\label{eom:non} = \theta_2 \quad \ddot\Omega = \theta_2 \dot\chi.$$ The solutions for the above coupled equations of motion are easily obtained as $$\begin{aligned} \Omega \alpha e^{2\kappa\theta_2 \beta e^{-2\kappa\theta_2 + \label{sol_gen:Omega_non} \chi &=& \alpha - \beta e^{-2\kappa\theta_2 t} + B, $\beta$, $A$, and $B$ are constants, and they satisfy the equation in Eq. (\[con\]), $$\label{constr_gen:non} \kappa = \kappa^2 \theta_2^2 | , $1$ for $t>0$ and $0$ for $t<0$. Thus, these are noNtrivial anD $\thetA$-dePenDeNt foR the Finite time inteRVal $t_1 < T < t_2$. With the Hamiltonian in eq. (\[H\]), thE eQUatiONs Of motIon are oBTaINEd aS $$\bEgIn{aLiGNeD}
& & \dot\cHi = \{ \cHi, H \}_{\mathRm{MPB}} = -2\kappa p_\chI, \qUad
\dot\Omega = \{ \OMEgA, H \}_{\mathrm{MPb}} = 2\kaPpa P_\Omega,
\labEl{1sT:x_non}\\
& & \dOt{p}_\chI = \{ p_\chi, H \}_{\MatHrm{MPb}} =2\kappa \THeta_2
P_\OMega, \quad \dOt{p}_\omega = \{ P_\oMega, H \}_{\maTHRm{mPB}}
= 2\kAppa\theta_2 P_\chi. \labeL{1St:P_Non}\end{aligned}$$ NOte thaT tHE mOMEntA arE no longer cOnStantS Of the moTIoN BECauSE of the nonvaniShing $\theta_2$; hENce, A new seT oF eqUAtions Of motIoN Are Obtained froM Eqs. (\[1St:x\_non\]) and (\[1St:p\_non\]): $$\LAbel{eom:NOn}
\ddot\cHi = -2\kappA \thEta_2 \Dot\OMEgA, \qUad \DdOT\OmEGa =
-2\KapPA \thEta_2 \dot\chI.$$ THe SolutIons FOR THe abOve CoupLed eqUations of motiOn aRe eaSIly ObtaiNed as $$\BegiN{aLigneD}
\Omega &=& \Alpha E^{2\kAppa\theta_2 t} + \beta
e^{-2\KappA\theta_2 t} + A, \lAbeL{sOl_gEn:omega_NOn} \\
\chi &=& \aLphA e^{2\kAppa\theTa_2 t} - \beta
E^{-2\KapPa\THETa_2 T} + B, \label{sol_gen:chi_noN}\eND{AlIgned}$$ wheRe $\alphA$, $\BeTa$, $a$, And $B$ are cOnStaNts, aND They sHoulD SaTisfy the ConstrAInT eQuation In eq. (\[con\]), $$\lAbEl{cOnsTr_gen:NOn}
\kaPpa t_\pm = \Kappa^2 \theTa_2^2 | , $1$ for $t>0$ and $0$ fo r $t<0$. T hus,the sear e no ntri vial and $\the t a$-d ependent for the finit e tim ei nter v al $t_1 < t <t _2 $ . Wi th t heHa m il tonia n i n Eq. ( \[H\]), th e e qu ations of mo t io n are obta ine d as $$\begi n{a ligned } & & \do t\c hi =\{ \ch i , H \} _{\mathrm {M P B}} =- 2\kappa P _\ chi, \quad
\dot \ Om e ga = \{ \Omeg a, H \ }_ { \m a t hrm {MP B}} = 2\ka pp a P_\ O mega, \la b el{1st:x_non} \\
& & \d o t{P }_\chi = \ { P_\ch i, H\} _ {\m athrm{MPB}} =2\ kappa \th eta_2 P_ \ Omega,\quad\do t{P }_\O m eg a= \ {P _\O m eg a,H \} _{\mathr m{ MP B}} = 2\ka ppa \the ta_2P_\chi. \labe l{1 st:p _ non }\end {alig ned} $$ Note thatthe m om enta are no lon gerconstants of t hemo tionb ecause of th e nonva nishing $\t he t a _ 2$ ; hence, a new set o f eq uationsof mot i on a r e obtain ed fr om E q s . (\[ 1st: x \_ non\]) a nd (\[ 1 st :p \_non\] ): $$\la be l{e om: non} \dd ot\chi = -2\ka ppa \ t heta_2 \dot\Om e ga, \quad \dd o t\ O m eg a =
-2 \kappa \the ta_2 \dot \chi . $$ Th e solu tions f o rt he above coupled eq ua tionsof mo tion are easi ly obtaine d a s $$\beg in{a l ig n ed}
\Omega & =& \a lpha e^{2\ k appa\the ta_2t} + \be ta
e^{- 2\k app a\t het a _ 2t} + A, \labe l { sol_ ge n:Omega _no n} \\
\c hi&=& \a lp ha e^{2\k appa\the ta _2 t }- \ beta e^ {- 2\k appa\ t heta_2 t} + B,\l ab e l{s ol_gen: c hi _ n on}\ en d{ alig ned }$ $ whe re $ \ alp ha$, $\ beta$, $A $,a nd $ B$ a re cons tants, and th ey should sa ti sfy the c o n straintequation in Eq. (\[con\ ] ), $$\l abe l{con str_ gen:non} \ kappat_\ p m = \k appa^2 \the ta _2^ 2 | , $1$_for $t>0$_and $0$ for $t<0$._Thus, these_are_nontrivial and_$\theta$-dependent_for the finite_time interval $t_1_< t < t_2$._With the Hamiltonian_in_Eq. (\[H\]), the equations of motion are obtained as $$\begin{aligned}
& & \dot\chi =_\{_\chi, H_\}_{\mathrm{MPB}}_=_-2\kappa P_\chi, \quad
_ \dot\Omega =_ \{_\Omega, H \}_{\mathrm{MPB}} = 2\kappa P_\Omega,
__ _\label{1st:x_non}\\
& & \dot{P}_\chi = \{ P_\chi,_H \}_{\mathrm{MPB}} =2\kappa \theta_2
_ _P_\Omega,_\quad_\dot{P}_\Omega = \{ P_\Omega,_H \}_{\mathrm{MPB}}
_ = 2\kappa\theta_2 P_\chi. \label{1st:p_non}\end{aligned}$$_Note that the momenta are no longer_constants of the motion because of_the nonvanishing $\theta_2$; hence, a_new set_of equations of motion are_obtained from Eqs. (\[1st:x\_non\])_and (\[1st:p\_non\]):_$$\label{eom:non}
\ddot\chi_= -2\kappa \theta_2 \dot\Omega, \quad \ddot\Omega_=
-2\kappa_\theta_2 \dot\chi.$$ The solutions for the_above_coupled equations of_motion_are_easily obtained_as $$\begin{aligned}
_\Omega_&=& \alpha_e^{2\kappa\theta_2_t} + \beta
__ _e^{-2\kappa\theta_2 t} + A,_\label{sol_gen:Omega_non}_\\
\chi &=&_\alpha e^{2\kappa\theta_2 t} - \beta
_ _ _ _ e^{-2\kappa\theta_2 t} + B, \label{sol_gen:chi_non}\end{aligned}$$ where $\alpha$, $\beta$, $A$, and_$B$ are constants, and they should_satisfy the constraint equation_in Eq. (\[con\]),_$$\label{constr_gen:non}
_ \kappa t_\pm_=_\kappa^2 \theta_2^2 |
{F}}$.
The commutativity of the top square is given by [@BK4 Lemma 4.21]. The commutativity of the lower square can be easily verified by diagram chasing.
\[L:factontraces\] The following diagram of vector spaces is commutative. $$\xymatrix{
\operatorname{\mathsf{Hom}}({\mathcal{F}}, {\mathbbm{k}}_x) \otimes \operatorname{\mathsf{Ext}}({\mathbbm{k}}_x, {\mathcal{F}}) \ar[rrr]^-{\overline{\mathbb{S}}} \ar[ddd]_T & & & \operatorname{\mathsf{End}}\bigl(\operatorname{\mathsf{Hom}}({\mathcal{F}}, {\mathbbm{k}}_x)\bigr)
\ar[ddd] \\
& K \ar[r]^-{\overline{\mathbb{S}}} \ar[d]_-{\overline{T}} \ar@{_{(}->}[lu]& \operatorname{\mathfrak{sl}}\bigl(\operatorname{\mathsf{Hom}}({\mathcal{F}}, {\mathbbm{k}}_x)\bigr) \ar@{^{(}->}[ru] \ar[d] & \\
&\operatorname{\mathfrak{sl}}({\mathcal{F}}\big|_x) \ar[r]^-{Y_1} \ar@{^{(}->}[ld] & \operatorname{\mathfrak{sl}}\bigl(\operatorname{\mathsf{Lin}}({\mathcal{F}}\big|_x, {\mathbbm{k}})\bigr) \ar@{^{(}->}[rd]& \\
\operatorname{\mathsf{End}}({\mathcal{F}}\big|_x) \ar[rrr]^-{Y} & & & \operatorname{\mathsf{End}}\bigl(\operatorname{\mathsf{Lin}}({\mathcal{F}}\big|_x, {\mathbbm{k}})\bigr)
}$$ In this diagram, $\overline{\mathbb{S}}$ is the isomorphism induced by the Serre duality (\[E:SerreMap\]), $Y$ and $Y_1$ are canonical isomorphisms from Lemma \[L:LA2\], $K$ is the subspace of $\operatorname{\mathsf{Hom}}({\mathcal{F}}, \otimes {\mathbbm{k}}_x) \otimes \operatorname{\mathsf{Ext}}({\mathbbm{k}}_x, {\mathcal{F}})$ defined in (\[E:linearspaceK\]), $T$ is the composition of $\mathbbm{1} \otimes (\delta_x^{\mathcal{F}})^{-1}$ from Proposition \[P: | { F}}$.
The commutativity of the top square is given by [ @BK4 Lemma 4.21 ]. The commutativity of the lower square can be well verify by diagram chasing.
\[L: factontraces\ ] The following diagram of vector spaces is commutative. $ $ \xymatrix {
\operatorname{\mathsf{Hom}}({\mathcal{F } }, { \mathbbm{k}}_x) \otimes \operatorname{\mathsf{Ext}}({\mathbbm{k}}_x, { \mathcal{F } }) \ar[rrr]^-{\overline{\mathbb{S } } } \ar[ddd]_T & & & \operatorname{\mathsf{End}}\bigl(\operatorname{\mathsf{Hom}}({\mathcal{F } }, { \mathbbm{k}}_x)\bigr)
\ar[ddd ] \\
& K \ar[r]^-{\overline{\mathbb{S } } } \ar[d]_-{\overline{T } } \ar@{_{(}->}[lu ] & \operatorname{\mathfrak{sl}}\bigl(\operatorname{\mathsf{Hom}}({\mathcal{F } }, { \mathbbm{k}}_x)\bigr) \ar@{^{(}->}[ru ] \ar[d ] & \\
& \operatorname{\mathfrak{sl}}({\mathcal{F}}\big|_x) \ar[r]^-{Y_1 } \ar@{^{(}->}[ld ] & \operatorname{\mathfrak{sl}}\bigl(\operatorname{\mathsf{Lin}}({\mathcal{F}}\big|_x, { \mathbbm{k}})\bigr) \ar@{^{(}->}[rd ] & \\
\operatorname{\mathsf{End}}({\mathcal{F}}\big|_x) \ar[rrr]^-{Y } & & & \operatorname{\mathsf{End}}\bigl(\operatorname{\mathsf{Lin}}({\mathcal{F}}\big|_x, { \mathbbm{k}})\bigr)
} $ $ In this diagram, $ \overline{\mathbb{S}}$ is the isomorphism induce by the Serre dichotomy (\[E: SerreMap\ ]), $ Y$ and $ Y_1 $ are canonical isomorphisms from Lemma \[L: LA2\ ], $ K$ is the subspace of $ \operatorname{\mathsf{Hom}}({\mathcal{F } }, \otimes { \mathbbm{k}}_x) \otimes \operatorname{\mathsf{Ext}}({\mathbbm{k}}_x, { \mathcal{F}})$ defined in (\[E: linearspaceK\ ]), $ T$ is the typography of $ \mathbbm{1 } \otimes (\delta_x^{\mathcal{F}})^{-1}$ from Proposition \[P: | {F}}$.
Thf commutativity of the tup square is girwn by [@UK4 Lemmz 4.21]. The cummutativity of the lower sqnare can ve easily verified by aiagram cjasing.
\[L:fqctoitraces\] The following diagram kn veccoc spaces is comkutative. $$\xfmatrix{
\operatosnxmz{\mathsf{Hom}}({\mathcal{F}}, {\mathbbm{k}}_x) \otimes \jperatotnwme{\mathsf{Ext}}({\majhbbm{l}}_v, {\mafhcal{F}}) \ar[rrr]^-{\overline{\mathbb{S}}} \ar[ddd]_T & & & \opegatorname{\mathsf{Enc}}\bigl(\operatorname{\mathsf{Hom}}({\lathfal{F}}, {\mathbbm{k}}_x)\bigr)
\wr[ddd] \\
& K \at[d]^-{\ovqeline{\mathbb{S}}} \ar[d]_-{\overline{T}} \ar@{_{(}->}[lu]& \opetatorname{\mathfrak{sl}}\bigl(\operatornxme{\machsf{Hom}}({\mathxao{F}}, {\ldthbbm{k}}_x)\bigc) \ar@{^{(}->}[ru] \ar[d] & \\
&\operabprname{\kathfral{sl}}({\mathcal{F}}\big|_w) \ar[r]^-{B_1} \ar@{^{(}->}[od] & \operatorname{\mathfcak{sl}}\bigl(\operatornamg{\mathsf{Lin}}({\kachcal{F}}\big|_x, {\mathbbm{k}})\bitr) \ar@{^{(}->}[rd]& \\
\opesatofbamd{\mauhsh{Ens}}({\mathcwl{F}}\uig|_x) \ar[rrr]^-{Y} & & & \operatirname{\mathsf{End}}\bigl(\ppqgstorname{\mathaf{Lin}}({\mwtrcal{F}}\big|_x, {\mathbbm{k}})\bigr)
}$$ In this diagram, $\mvedline{\mathbb{S}}$ is the isonorphism induced by tje Serre quality (\[E:SerreMap\]), $Y$ and $Y_1$ are canonical isomorphivms fcoo Ltmna \[L:LX2\], $K$ is the subspace of $\operatorname{\mathsf{Hom}}({\mathszl{G}}, \ptimes {\mathbbm{h}}_x) \otimes \olegayjrname{\mathsf{Ebt}}({\mathygm{i}}_x, {\mathcal{F}})$ definef in (\[E:lynearwpaceK\]), $T$ ys tne composition of $\mathbbm{1} \orimes (\delta_x^{\iqthcal{F}})^{-1}$ from Propodition \[P: | {F}}$. The commutativity of the top square by Lemma 4.21]. commutativity of the verified diagram chasing. \[L:factontraces\] following diagram of spaces is commutative. $$\xymatrix{ \operatorname{\mathsf{Hom}}({\mathcal{F}}, {\mathbbm{k}}_x) \operatorname{\mathsf{Ext}}({\mathbbm{k}}_x, {\mathcal{F}}) \ar[rrr]^-{\overline{\mathbb{S}}} \ar[ddd]_T & & & \operatorname{\mathsf{End}}\bigl(\operatorname{\mathsf{Hom}}({\mathcal{F}}, {\mathbbm{k}}_x)\bigr) \ar[ddd] \\ & K \ar[d]_-{\overline{T}} \ar@{_{(}->}[lu]& \operatorname{\mathfrak{sl}}\bigl(\operatorname{\mathsf{Hom}}({\mathcal{F}}, {\mathbbm{k}}_x)\bigr) \ar@{^{(}->}[ru] \ar[d] & \\ &\operatorname{\mathfrak{sl}}({\mathcal{F}}\big|_x) \ar[r]^-{Y_1} \ar@{^{(}->}[ld] & \operatorname{\mathfrak{sl}}\bigl(\operatorname{\mathsf{Lin}}({\mathcal{F}}\big|_x, \ar@{^{(}->}[rd]& \operatorname{\mathsf{End}}({\mathcal{F}}\big|_x) & & \operatorname{\mathsf{End}}\bigl(\operatorname{\mathsf{Lin}}({\mathcal{F}}\big|_x, {\mathbbm{k}})\bigr) }$$ In this diagram, $\overline{\mathbb{S}}$ is the isomorphism induced by the Serre duality $Y$ and $Y_1$ are canonical isomorphisms from Lemma $K$ is the subspace $\operatorname{\mathsf{Hom}}({\mathcal{F}}, \otimes {\mathbbm{k}}_x) \otimes \operatorname{\mathsf{Ext}}({\mathbbm{k}}_x, defined (\[E:linearspaceK\]), $T$ the of \otimes (\delta_x^{\mathcal{F}})^{-1}$ from \[P: | {F}}$.
The commutativity of the top Square is giVen by [@bK4 LEmmA 4.21]. THe coMmutAtivity of the loWEr sqUare can be easily verifieD by diAgRAm chASiNg.
\[L:faCtontraCEs\] tHE foLlOwIng DiAGrAm of vEctOr spaceS is commutaTivE. $$\xYmatrix{
\operaTOrName{\mathsf{hom}}({\Mathcal{F}}, {\mathBbm{K}}_x) \otimEs \OpeRAtornAme{\MathsF{Ext}}({\maTHbbm{k}}_x, {\Mathcal{F}}) \aR[rRR]^-{\overlINe{\mathbB{s}}} \Ar[Ddd]_T & & & \Operatorname{\mathsF{enD}}\Bigl(\operatornaMe{\mathSf{hOm}}({\MAThcAl{F}}, {\Mathbbm{k}}_x)\bIgR)
\ar[ddD] \\
& k \ar[r]^-{\oveRLiNE{\MAthBB{S}}} \ar[d]_-{\overline{t}} \ar@{_{(}->}[lu]& \operatORnaMe{\mathFrAk{sL}}\Bigl(\opEratoRnAMe{\mAthsf{Hom}}({\matHcal{f}}, {\mathbbm{k}}_X)\bigr) \aR@{^{(}->}[Ru] \ar[d] & \\
&\opERatornaMe{\mathFraK{sl}}({\MathCAl{f}}\bIg|_x) \Ar[R]^-{y_1} \ar@{^{(}->}[LD] & \oPerATorName{\mathFrAk{Sl}}\bigL(\opeRATORnamE{\maThsf{lin}}({\maThcal{F}}\big|_x, {\matHbbM{k}})\biGR) \ar@{^{(}->}[Rd]& \\
\opeRatorName{\MaThsf{ENd}}({\mathCal{F}}\bIg|_X) \ar[rrr]^-{Y} & & & \operatorName{\Mathsf{End}}\BigL(\oPerAtOrnamE{\Mathsf{lin}}({\MatHcal{F}}\biG|_x, {\mathbBM{k}})\bIgR)
}$$ iN ThIs diagram, $\overline{\mAtHBB{S}}$ Is the isoMorphiSM iNdUCed by the seRre DualITY (\[E:SerReMaP\]), $y$ aNd $Y_1$ are caNonicaL IsOmOrphismS fRom LemMa \[l:LA2\], $k$ is The suBSpacE of $\opeRatornamE{\mathSF{Hom}}({\mathcal{F}}, \otIMes {\mathbbm{k}}_x) \oTImES \OpERatoRnaMe{\mathsf{Ext}}({\MathBBm{k}}_x, {\MathCAl{f}})$ deFIned iN (\[E:linEaRSpACeK\]), $T$ is the compositioN oF $\mathbBm{1} \otiMes (\delta_x^{\mathCal{F}})^{-1}$ from PrOPOSition \[P: | {F}}$.
The commutativityof the top squa reisgi venby [ @BK4 Lemma 4.2 1 ]. T he commutativity of th e low er squa r ecan b e easil y v e r ifi ed b y d ia g ra m cha sin g.
\[L :factontra ces \] The followi n gdiagram of ve ctor spacesiscommut at ive . $$\x yma trix{
\oper a tornam e{\mathsf {H o m}}({\ m athcal{ F } }, {\m athbbm{k}}_x) \ot i me s \operatorname {\math sf { Ex t } }({ \ma thbbm{k}}_ x, {\ma t hcal{F} } )\ a r [rr r ]^-{\overline {\mathbb{S} } } \ ar[ddd ]_ T & & & \o perat or n ame {\mathsf{En d}}\ bigl(\ope ratorn a me{\mat h sf{Hom} }({\ma thc al{ F}}, {\ ma thb bm { k}} _ x) \bi g r)\ar[ddd] \ \ & K\ar[ r ] ^ - {\ov erl ine{ \math bb{S}}} \ar[d ]_- {\ov e rli ne{T} } \ar @{_{ (} ->}[l u]& \o perat or name{\mathfrak{ sl}} \bigl(\op era to rna me {\mat h sf{Hom }}( {\m athcal{ F}}, {\ m ath bb m { k }} _x)\bigr) \ar@{^{( }- > } [r u] \ar[d ] & \\ & \o p eratorna me {\m athf r a k{sl} }({\ m at hcal{F}} \big|_ x )\a r[r]^-{ Y_ 1} \ar @{ ^{( }-> }[ld] & \o perato rname{\m athfr a k{sl}}\bigl(\o p eratorname{\m a th s f {L i n}}( {\m athcal{F}}\ big| _ x, { \mat h bb m{k } })\bi gr) \ ar @ {^ { (}->}[rd]& \\
\oper at orname {\mat hsf{End}}({\m athcal{F}} \ b i g|_x) \a r[rr r ]^ - {Y} & & & \op erato rname{\mat h sf{End}} \bigl (\operat orname{\m a t hsf{Lin} }({ \ma thc al{ F } }\ big|_x, {\mat h b bm{k }} )\bigr)
}$ $ In th isdia gra m,$\ overline{ \mathbb{ S} }$ i sthe isom o rphism i nd uce dbythe S e rre du ality (\[ E: Se r reM ap\]),$ Y$ a nd $ Y_ 1$ are ca no nical iso m orp hisms f rom Lemma \[ L :LA2 \] ,$K$ isthe subspaceof $\operato rn ame {\math s f {Hom}}({ \mathcal{F}}, \otimes { \ mathbbm {k} }_x)\oti mes \oper ato rname{ \ma t hsf{Ex t}}({\ mathb bm {k} } _ x, {\ m a th cal {F }})$ defin e d in (\[E :l inea rspaceK \]), $T$ is the co m pos ition of $\ma thb bm{1 } \o tim e s( \de lt a _x^ { \ mathcal{F}})^{- 1}$ from P ro p os ition \[P: | {F}}$.
The commutativity_of the_top square is given_by [@BK4_Lemma_4.21]. The_commutativity_of the lower_square can be_easily verified by diagram_chasing.
\[L:factontraces\] The following_diagram_of vector spaces is commutative. $$\xymatrix{
\operatorname{\mathsf{Hom}}({\mathcal{F}}, {\mathbbm{k}}_x) \otimes \operatorname{\mathsf{Ext}}({\mathbbm{k}}_x, {\mathcal{F}}) \ar[rrr]^-{\overline{\mathbb{S}}} \ar[ddd]_T & &_&_\operatorname{\mathsf{End}}\bigl(\operatorname{\mathsf{Hom}}({\mathcal{F}}, {\mathbbm{k}}_x)\bigr)
\ar[ddd]_\\
_&_K \ar[r]^-{\overline{\mathbb{S}}} \ar[d]_-{\overline{T}} \ar@{_{(}->}[lu]& \operatorname{\mathfrak{sl}}\bigl(\operatorname{\mathsf{Hom}}({\mathcal{F}},_{\mathbbm{k}}_x)\bigr) \ar@{^{(}->}[ru] \ar[d] & \\
_&\operatorname{\mathfrak{sl}}({\mathcal{F}}\big|_x) \ar[r]^-{Y_1}_\ar@{^{(}->}[ld] & \operatorname{\mathfrak{sl}}\bigl(\operatorname{\mathsf{Lin}}({\mathcal{F}}\big|_x, {\mathbbm{k}})\bigr) \ar@{^{(}->}[rd]& \\
\operatorname{\mathsf{End}}({\mathcal{F}}\big|_x) \ar[rrr]^-{Y} &__& & \operatorname{\mathsf{End}}\bigl(\operatorname{\mathsf{Lin}}({\mathcal{F}}\big|_x,_{\mathbbm{k}})\bigr)
}$$ In this diagram, $\overline{\mathbb{S}}$ is the isomorphism induced_by the Serre duality (\[E:SerreMap\]), $Y$_and $Y_1$ are_canonical_isomorphisms_from Lemma \[L:LA2\], $K$_is the subspace of $\operatorname{\mathsf{Hom}}({\mathcal{F}}, \otimes_{\mathbbm{k}}_x) \otimes \operatorname{\mathsf{Ext}}({\mathbbm{k}}_x, {\mathcal{F}})$ defined in_(\[E:linearspaceK\]), $T$ is the composition of $\mathbbm{1}_\otimes (\delta_x^{\mathcal{F}})^{-1}$ from Proposition \[P: |
satisfying conditions (DLC1)-(DLC5) (or conditions (DLC1), (DLC2), (LC3), (DLC4), (DLC5)). Then $\p$ satisfies conditions (DLC3S) and (LC3S) as well.
Let $f=f_\p$ (see Lemma \[funcon\] for $f_\p$) and $\p\ap=\p_f$ (see Lemma \[phief\] for $\p_f$). Then, by Lemma \[commdiagr\], $\lbg\circ\p=\p\ap\circ\lag$. Since the function $\p\ap$ satisfies conditions (DLC3S) and (LC3S) (by Lemma \[phief\]) and the functions $\lag$ and $\lbg$ are LCA-isomorphisms, we get that the function $\p$ satisfies conditions (DLC3S) and (LC3S) as well.
The above lemma implies the following fact mentioned in the previous section:
\[dlc3sax\] Condition (DLC3) in Definition \[dhc\] can be replaced by any of the conditions (DLC3S), (LC3) and (LC3S) (i.e., we obtain equivalent systems of axioms by these replacements).
\[compos1\] Let $\p:(A,\rho,\BBBB)\lra (B,\eta,\BBBB\ap)$ be a function between CLCAs and let $\p$ satisfy conditions (DLC1)-(DLC4). Then the function $\p\cuk$ (see (\[cukf\])) satisfies conditions (DLC1)-(DLC5).
Obviously, for every $a\in A$, $\p\cuk(a)\le\p(a)$. Hence, $\p\cuk(0)=0$, i.e. (DLC1) is fulfilled. For (DLC2) and (DLC5) see \[pf1\](d).
Let $a\in\BBBB, b\in A$ and $a\llx b$. Then, by (BC1), there exist $c,d\in\BBBB$ such that $a\llx c\llx d\llx b$. Thus $a\llcr c$ and hence $c^*\llcr a^*$. Now, using Lemma \[c | satisfying conditions (DLC1)-(DLC5) (or conditions (DLC1), (DLC2), (LC3), (DLC4), (DLC5) ). Then $ \p$ satisfies conditions (DLC3S) and (LC3S) equally well.
lease $ f = f_\p$ (see Lemma \[funcon\ ] for $ f_\p$) and $ \p\ap=\p_f$ (see Lemma \[phief\ ] for $ \p_f$). Then, by Lemma \[commdiagr\ ], $ \lbg\circ\p=\p\ap\circ\lag$. Since the affair $ \p\ap$ satisfies condition (DLC3S) and (LC3S) (by Lemma \[phief\ ]) and the functions $ \lag$ and $ \lbg$ are LCA - isomorphisms, we get that the routine $ \p$ satisfies condition (DLC3S) and (LC3S) equally well.
The above lemma entail the following fact mention in the previous section:
\[dlc3sax\ ] Condition (DLC3) in Definition \[dhc\ ] can be replace by any of the conditions (DLC3S), (LC3) and (LC3S) (i.e., we obtain equivalent systems of axioms by these replacements).
\[compos1\ ] permit $ \p:(A,\rho,\BBBB)\lra (B,\eta,\BBBB\ap)$ be a function between CLCAs and let $ \p$ satisfy conditions (DLC1)-(DLC4). Then the routine $ \p\cuk$ (see (\[cukf\ ]) ) satisfies conditions (DLC1)-(DLC5).
Obviously, for every $ a\in A$, $ \p\cuk(a)\le\p(a)$. Hence, $ \p\cuk(0)=0 $, i.e. (DLC1) is carry through. For (DLC2) and (DLC5) see \[pf1\](d).
Let $ a\in\BBBB, b\in A$ and $ a\llx b$. Then, by (BC1), there exist $ c, d\in\BBBB$ such that $ a\llx c\llx d\llx b$. Thus $ a\llcr c$ and hence $ c^*\llcr a^*$. nowadays, using Lemma \[c | sahisfying conditions (DLC1)-(DUC5) (or conditions (DLC1), (DNC2), (LC3), (SLC4), (DLC5)). Ghen $\p$ satisfies conditions (VLC3S) and (OC3S) as well.
Let $f=f_\p$ (see Lemma \[fujcon\] for $f_\p$) end $\p\ap=\p_f$ (see Lemma \[phien\] for $\i_f$). Thzn, by Lemma \[commciagr\], $\lbg\chrc\p=\p\ap\circ\lag$. Sknee the function $\p\ap$ satisfies condityons (DLV3S) and (LC3S) (by Leima \[khiqf\]) ahd the functions $\lag$ and $\lbg$ are LDA-isomogphisms, we get thst the function $\p$ satisfied cojditions (DLC3S) and (PC3S) as well.
Jge wvove lemma ioplies the following fzct mentioned in the previous sdctiou:
\[dlc3sax\] Coneirioj (DLC3) in Defmnitiog \[dhc\] can be replaceg by anu of the condibions (DLX3S), (LC3) and (LC3S) (i.e., we outain equivalent sysjems of axhojs by these replaxenents).
\[wompms1\] Ldr $\p:(X,\rhk,\BUBB)\mra (B,\eha,\BUBB\ap)$ be a runction berween CLCAs and let $\p$ watisfy condifions (QLS1)-(DLC4). Then the function $\p\cuk$ (see (\[cukf\])) sauisfiss conditions (DLC1)-(DLC5).
Obvuously, for every $a\in W$, $\p\cuk(a)\le\[(a)$. Hence, $\p\cuk(0)=0$, i.e. (DLC1) is fulfilled. For (DLC2) and (DLC5) see \['f1\](a).
Leu $q\in\BBCV, h\in A$ and $a\llx b$. Then, by (BC1), there exist $c,d\in\BFGB$ slch that $a\llx c\llw d\llx b$. Thus $a\llct f$ sgd hence $c^*\llct a^*$. Now, usjng Lemma \[c | satisfying conditions (DLC1)-(DLC5) (or conditions (DLC1), (DLC2), (DLC5)). $\p$ satisfies (DLC3S) and (LC3S) Lemma for $f_\p$) and (see Lemma \[phief\] $\p_f$). Then, by Lemma \[commdiagr\], $\lbg\circ\p=\p\ap\circ\lag$. the function $\p\ap$ satisfies conditions (DLC3S) and (LC3S) (by Lemma \[phief\]) and the $\lag$ and $\lbg$ are LCA-isomorphisms, we get that the function $\p$ satisfies conditions and as The lemma implies the following fact mentioned in the previous section: \[dlc3sax\] Condition (DLC3) in Definition \[dhc\] be replaced by any of the conditions (DLC3S), and (LC3S) (i.e., we equivalent systems of axioms by replacements). Let $\p:(A,\rho,\BBBB)\lra be function CLCAs and let satisfy conditions (DLC1)-(DLC4). Then the function $\p\cuk$ (see (\[cukf\])) satisfies conditions (DLC1)-(DLC5). Obviously, for every $a\in A$, Hence, $\p\cuk(0)=0$, is fulfilled. (DLC2) (DLC5) \[pf1\](d). Let $a\in\BBBB, and $a\llx b$. Then, by (BC1), such that $a\llx c\llx d\llx b$. Thus $a\llcr and hence a^*$. Now, using Lemma \[c | satisfying conditions (DLC1)-(DLc5) (or conditiOns (DLc1), (DLc2), (LC3), (dLc4), (DLC5)). then $\P$ satisfies condITionS (DLC3S) and (LC3S) as well.
Let $f=f_\P$ (see LEmMA \[funCOn\] For $f_\p$) And $\p\ap=\p_F$ (SeE lEmmA \[pHiEf\] fOr $\P_F$). THen, by lemMa \[commdIagr\], $\lbg\cirC\p=\p\Ap\Circ\lag$. Since THe Function $\p\aP$ saTisfies condiTioNs (DLC3S) AnD (LC3s) (By LemMa \[pHief\]) aNd the fUNctionS $\lag$ and $\lbG$ aRE LCA-isOMorphisMS, We Get tHat the function $\p$ saTIsFIes conditions (DlC3S) and (lC3s) As WELl.
THe aBove lemma iMpLies tHE followINg FACT meNTioned in the prEvious sectiON:
\[dlC3sax\] CoNdItiON (DLC3) in definItIOn \[dHc\] can be replAced By any of thE condiTIons (DLC3s), (lC3) and (LC3s) (i.e., we oBtaIn eQuivALeNt SysTeMS of AXiOms BY thEse replaCeMeNts).
\[coMpos1\] lET $\P:(a,\rho,\bBBb)\lra (b,\eta,\BbBB\ap)$ be a functIon BetwEEn ClCAs aNd let $\P$ satIsFy conDitionS (DLC1)-(DlC4). then the function $\P\cuk$ (See (\[cukf\])) saTisFiEs cOnDitioNS (DLC1)-(DLc5).
ObVioUsly, for Every $a\iN a$, $\p\cUk(A)\LE\P(a)$. hence, $\p\cuk(0)=0$, i.e. (DLC1) is fuLfILLeD. For (DLC2) aNd (DLC5) sEE \[pF1\](d).
lEt $a\in\BBBb, b\In A$ And $a\LLX b$. TheN, by (Bc1), ThEre exist $C,d\in\BBbb$ sUcH that $a\lLx C\llx d\lLx B$. ThUs $a\Llcr c$ ANd heNce $c^*\llCr a^*$. Now, usIng LeMMa \[c | satisfying conditions (DL C1)-(DLC5) (orcon dit io ns ( DLC1 ), (DLC2), (LC 3 ), ( DLC4), (DLC5)). Then $ \p$ s at i sfie s c ondit ions (D L C3 S ) an d(L C3S )a swell.
L et $f=f _\p$ (seeLem ma \[funcon\]f or $f_\p$) a nd$\p\ap=\p_f$ (s ee Lem ma \[ p hief\ ] f or $\ p_f$). Then,by Lemma\[ c ommdia g r\], $\ l b g\ circ \p=\p\ap\circ\lag $ .S ince the funct ion $\ p\ a p$ s ati sfi es conditi on s (DL C 3S) and (L C 3 S ) ( b y Lemma \[phi ef\]) and t h e f unctio ns $\ l ag$ an d $\l bg $ ar e LCA-isomo rphi sms, we g et tha t the fu n ction $ \p$ sa tis fie s co n di ti ons ( D LC3 S )and (LC 3S) as w el l.
The abo v e l emma im plie s the following fa ctment i one d inthe p revi ou s sec tion:
\[dl c3 sax\] Condition (DL C3) in De fin it ion \ [dhc\ ] can b e r epl aced by any of the c o n d it ions (DLC3S), (LC3 )a n d(LC3S) ( i.e.,w eob t ain equi va len t sy s t ems o f ax i om s by the se rep l ac em ents).
\ [compo s1 \]Let $\p: ( A,\r ho,\BB BB)\lra(B,\e t a,\BBBB\ap)$ b e a function b e tw e e nC LCAs an d let $\p$sati s fy c ondi t io ns( DLC1) -(DLC 4) . T h en the function $\p \c uk$ (s ee (\ [cukf\])) sat isfies con d i t ions (DL C1)- ( DL C 5).
Obviously , for every $a\ i n A$, $\ p\cuk (a)\le\p (a)$. Hen c e , $\p\cu k(0 )=0 $,i.e . (D LC1) is fulfi l l ed.Fo r (DLC2 ) a nd (DLC 5)see \[ pf1 \] (d).
Let $a\in\B BB B, b \i n A $ and $a\llx b $. Th en , b y (BC 1 ), the re ex ist$c ,d \ in\ BBBB$ s u ch t hat$a \l lx c \ll xd\llx b$. Thu s $a\ll cr c$ and he n ce $ c^ *\ llcr a^ *$. Now, usin gLemma \[c | satisfying_conditions (DLC1)-(DLC5)_(or conditions (DLC1), (DLC2),_(LC3), (DLC4),_(DLC5))._Then $\p$_satisfies_conditions (DLC3S) and_(LC3S) as well.
Let_$f=f_\p$ (see Lemma \[funcon\]_for $f_\p$) and_$\p\ap=\p_f$_(see Lemma \[phief\] for $\p_f$). Then, by Lemma \[commdiagr\], $\lbg\circ\p=\p\ap\circ\lag$. Since the function $\p\ap$_satisfies_conditions (DLC3S)_and_(LC3S)_(by Lemma \[phief\]) and the_functions $\lag$ and $\lbg$ are_LCA-isomorphisms, we_get that the function $\p$ satisfies conditions (DLC3S)_and_(LC3S) as well.
The_above lemma implies the following fact mentioned in the_previous section:
\[dlc3sax\] Condition (DLC3) in Definition_\[dhc\] can be_replaced_by_any of the conditions_(DLC3S), (LC3) and (LC3S) (i.e., we_obtain equivalent systems of axioms by_these replacements).
\[compos1\] Let $\p:(A,\rho,\BBBB)\lra (B,\eta,\BBBB\ap)$ be a_function between CLCAs and let $\p$_satisfy conditions (DLC1)-(DLC4). Then the_function $\p\cuk$_(see (\[cukf\])) satisfies conditions (DLC1)-(DLC5).
Obviously,_for every $a\in_A$, $\p\cuk(a)\le\p(a)$._Hence, $\p\cuk(0)=0$, i.e._(DLC1) is fulfilled. For (DLC2) and_(DLC5) see \[pf1\](d).
Let_$a\in\BBBB, b\in A$ and $a\llx b$._Then,_by (BC1), there_exist_$c,d\in\BBBB$_such that_$a\llx c\llx d\llx_b$._Thus $a\llcr_c$_and hence $c^*\llcr a^*$. Now, using_Lemma_\[c |
.7 & 7.9 [$\pm$ ]{}0.4 & 11.4 [$\pm$ ]{}0.8 & 5.6 [$\pm$ ]{}0.6 & 61 [$\pm$ ]{}6 & 56 [$\pm$ ]{}6\
100 & 6.79 & 14.2 [$\pm$ ]{}0.7 & 8.2 [$\pm$ ]{}0.4 & 12.5 [$\pm$ ]{}0.8 & 4.6 [$\pm$ ]{}0.7 & 68 [$\pm$ ]{}5 & 52 [$\pm$ ]{}4\
160 & 11.36 & 20.2 [$\pm$ ]{}0.8 & 13.3 [$\pm$ ]{}0.5 & 16.6 [$\pm$ ]{}1.0 & 6.9 [$\pm$ ]{}1.0 & 65 [$\pm$ ]{}5 & 62 [$\pm$ ]{}5\
250 & 18.2 & 26.3 [$\pm$ ]{}4.2 & 19.3 [$\pm$ ]{}3.1 & 19.0 [$\pm$ ]{}5.8 & [b]{} & $>$52[b]{} & 55 [$\pm$ ]{}22\
350 & 24.9 & 26.3 [$\pm$ ]{}6.9 & 22.7 [$\pm$ ]{}6.8 & [c]{} & [c]{} & [c]{} & [c]{}\
[a]{}[The effective beam size (geometric mean) is listed and used to deconvolve the sizes using Equation (\[eqn:convolve\]).]{} [b]{}[The disk is only marginally resolved along one axis at 250 . Therefore $\theta_{\rm{act,min}}$ cannot be derived for the short axis and consequently only the upper limit on the inclination can be calculated.]{} [c]{}[The system is consistent with being unresolved and symmetric at 350 .]{}
The size of the disk can be estimated by fitting 2D Gaussians to the images. First, a model of the photosphere is removed from the image by subtracting a PSF that is scaled to the expected photospheric emission (Section \[sec:photmodel\]). The FWHMs and position angles of the 2D Gaussian models are listed in Table \[tbl:gaussianFits\]. The quoted error on a fitted parameter gives the range of parameter values for which the [$\chi^2$]{} value is within 10% of the minimum [$\ | .7 & 7.9 [ $ \pm$ ] { } 0.4 & 11.4 [ $ \pm$ ] { } 0.8 & 5.6 [ $ \pm$ ] { } 0.6 & 61 [ $ \pm$ ] { } 6 & 56 [ $ \pm$ ] { } 6\
100 & 6.79 & 14.2 [ $ \pm$ ] { } 0.7 & 8.2 [ $ \pm$ ] { } 0.4 & 12.5 [ $ \pm$ ] { } 0.8 & 4.6 [ $ \pm$ ] { } 0.7 & 68 [ $ \pm$ ] { } 5 & 52 [ $ \pm$ ] { } 4\
160 & 11.36 & 20.2 [ $ \pm$ ] { } 0.8 & 13.3 [ $ \pm$ ] { } 0.5 & 16.6 [ $ \pm$ ] { } 1.0 & 6.9 [ $ \pm$ ] { } 1.0 & 65 [ $ \pm$ ] { } 5 & 62 [ $ \pm$ ] { } 5\
250 & 18.2 & 26.3 [ $ \pm$ ] { } 4.2 & 19.3 [ $ \pm$ ] { } 3.1 & 19.0 [ $ \pm$ ] { } 5.8 & [ b ] { } & $ > $ 52[b ] { } & 55 [ $ \pm$ ] { } 22\
350 & 24.9 & 26.3 [ $ \pm$ ] { } 6.9 & 22.7 [ $ \pm$ ] { } 6.8 & [ c ] { } & [ c ] { } & [ c ] { } & [ c]{}\
[ a]{}[The effective beam size (geometric mean) is listed and use to deconvolve the size using Equation (\[eqn: convolve\ ]). ] { } [ b]{}[The disk is entirely marginally resolved along one axis at 250 . consequently $ \theta_{\rm{act, min}}$ cannot be derived for the short axis and consequently merely the upper limit on the inclination can be forecast. ] { } [ c]{}[The system is consistent with being open and symmetrical at 350 . ] { }
The size of the disk can be estimated by fitting 2D Gaussians to the picture. First, a model of the photosphere is remove from the image by subtracting a PSF that is scaled to the expect photospheric emission (Section \[sec: photmodel\ ]). The FWHMs and position angles of the 2D Gaussian model are listed in Table \[tbl: gaussianFits\ ]. The quoted error on a fitted parameter give the range of parameter values for which the [ $ \chi^2 $ ] { } value is within 10% of the minimum [ $ \ | .7 & 7.9 [$\om$ ]{}0.4 & 11.4 [$\pm$ ]{}0.8 & 5.6 [$\pm$ ]{}0.6 & 61 [$\pm$ ]{}6 & 56 [$\pm$ ]{}6\
100 & 6.79 & 14.2 [$\pm$ ]{}0.7 & 8.2 [$\pm$ ]{}0.4 & 12.5 [$\pm$ ]{}0.8 & 4.6 [$\pm$ ]{}0.7 & 68 [$\pm$ ]{}5 & 52 [$\lm$ ]{}4\
160 & 11.36 & 20.2 [$\po$ ]{}0.8 & 13.3 [$\pm$ ]{}0.5 & 16.6 [$\pm$ ]{}1.0 & 6.9 [$\pm$ ]{}1.0 & 65 [$\pm$ ]{}5 & 62 [$\pm$ ]{}5\
250 & 18.2 & 26.3 [$\pm$ ]{}4.2 & 19.3 [$\pm$ ]{}3.1 & 19.0 [$\pm$ ]{}5.8 & [b]{} & $>$52[b]{} & 55 [$\pm$ ]{}22\
350 & 24.9 & 26.3 [$\pm$ ]{}6.9 & 22.7 [$\pm$ ]{}6.8 & [c]{} & [c]{} & [x]{} & [c]{}\
[e]{}[The effective bxzm size (geomefvic mzai) is listed and used to daconvolve the vixed using Equation (\[eqn:convolve\]).]{} [b]{}[The difk is omlj marginally rgsolvtd wlonf one axis at 250 . Therefore $\theta_{\rm{acf,min}}$ cainot be derived for the short axis and cojseqkently only the upoer limit ob thq inclination can be calculated.]{} [c]{}[Thg system is consistent with beine unrzsolved and stmmfjric at 350 .]{}
The wize jf the disk gsn be astimatrd by fitting 2C Geussuans to the images. Ficst, a model of the protosphera ns removed from the inate by subdracgung a KSF tgat is sceled to the expected pyotospheric emissiom (Fvvtion \[sec:photjodel\]). Ehq FWHMs and position angles of the 2D Galssizn models are listed in Table \[tbl:gaussianFits\]. The quotqd error on a fitted parameter gives the range of parajdtex valudw vor which the [$\chi^2$]{} value is within 10% of the miniihm [$\ | .7 & 7.9 [$\pm$ ]{}0.4 & 11.4 & [$\pm$ ]{}0.6 61 [$\pm$ ]{}6 & & 14.2 [$\pm$ & 8.2 [$\pm$ & 12.5 [$\pm$ ]{}0.8 & 4.6 ]{}0.7 & 68 [$\pm$ ]{}5 & 52 [$\pm$ ]{}4\ 160 & 11.36 & [$\pm$ ]{}0.8 & 13.3 [$\pm$ ]{}0.5 & 16.6 [$\pm$ ]{}1.0 & 6.9 [$\pm$ & [$\pm$ & [$\pm$ ]{}5\ 250 & 18.2 & 26.3 [$\pm$ ]{}4.2 & 19.3 [$\pm$ ]{}3.1 & 19.0 [$\pm$ & [b]{} & $>$52[b]{} & 55 [$\pm$ ]{}22\ & 24.9 & 26.3 ]{}6.9 & 22.7 [$\pm$ ]{}6.8 [c]{} [c]{} & & [a]{}[The beam size (geometric is listed and used to deconvolve the sizes using Equation (\[eqn:convolve\]).]{} [b]{}[The disk is only marginally resolved one axis . Therefore cannot derived the short axis only the upper limit on the calculated.]{} [c]{}[The system is consistent with being unresolved symmetric at .]{} The size of the disk be estimated by fitting 2D Gaussians to the First, a model of the photosphere is removed from the image by subtracting a PSF scaled to the expected emission (Section \[sec:photmodel\]). FWHMs position of 2D Gaussian are listed in Table \[tbl:gaussianFits\]. The quoted error on a fitted gives the range of parameter values for which the [$\chi^2$]{} within of the minimum | .7 & 7.9 [$\pm$ ]{}0.4 & 11.4 [$\pm$ ]{}0.8 & 5.6 [$\pm$ ]{}0.6 & 61 [$\pm$ ]{}6 & 56 [$\pm$ ]{}6\
100 & 6.79 & 14.2 [$\pm$ ]{}0.7 & 8.2 [$\pm$ ]{}0.4 & 12.5 [$\pm$ ]{}0.8 & 4.6 [$\pm$ ]{}0.7 & 68 [$\pm$ ]{}5 & 52 [$\pm$ ]{}4\
160 & 11.36 & 20.2 [$\pm$ ]{}0.8 & 13.3 [$\pM$ ]{}0.5 & 16.6 [$\pm$ ]{}1.0 & 6.9 [$\pm$ ]{}1.0 & 65 [$\pm$ ]{}5 & 62 [$\pm$ ]{}5\
250 & 18.2 & 26.3 [$\pM$ ]{}4.2 & 19.3 [$\pm$ ]{}3.1 & 19.0 [$\pm$ ]{}5.8 & [B]{} & $>$52[b]{} & 55 [$\pM$ ]{}22\
350 & 24.9 & 26.3 [$\pm$ ]{}6.9 & 22.7 [$\Pm$ ]{}6.8 & [C]{} & [c]{} & [c]{} & [c]{}\
[A]{}[The Effective beam sIZe (geOmetric mean) is listed and Used tO dEConvOLvE the sIzes usiNG EQUAtiOn (\[EqN:coNvOLvE\]).]{} [b]{}[The DisK is only Marginally ResOlVed along one aXIs At 250 . ThereforE $\thEta_{\rm{act,min}}$ cAnnOt be deRiVed FOr the ShoRt axiS and coNSequenTly only thE uPPer limIT on the iNCLiNatiOn can be calculated.]{} [C]{}[thE System is consisTent wiTh BEiNG UnrEsoLved and symMeTric aT 350 .]{}
the size OF tHE DIsk CAn be estimated By fitting 2D GAUssIans to ThE imAGes. FirSt, a moDeL Of tHe photospheRe is Removed frOm the iMAge by suBTractinG a PSF tHat Is sCaleD To ThE exPeCTed PHoTosPHerIc emissiOn (seCtion \[Sec:pHOTMOdel\]). the fWHMS and pOsition angles Of tHe 2D GAUssIan moDels aRe liStEd in TAble \[tbL:gausSiAnFits\]. The quoted ErroR on a fitteD paRaMetEr Gives THe rangE of ParAmeter vAlues foR WhiCh THE [$\ChI^2$]{} value is within 10% of thE mINImUm [$\ | .7 & 7.9 [$\pm$ ]{}0.4 & 1 1.4 [$\pm$ ]{}0 .8& 5 .6 [$\ pm$]{}0.6 & 61 [$ \ pm$]{}6 & 56 [$\pm$ ]{}6\
100&6 .79& 1 4.2 [ $\pm$ ] { }0 . 7 &8. 2[$\ pm $ ] {}0.4 &12.5 [$ \pm$ ]{}0. 8 & 4 .6 [$\pm$ ]{ } 0. 7 & 68 [$\ pm$ ]{}5 & 52 [ $\p m$ ]{} 4\
16 0 & 11 .36 & 20 .2 [$\ p m$ ]{} 0.8 & 13. 3[ $\pm$] {}0.5 & 1 6. 6 [$ \pm$ ]{}1.0 & 6.9 [$ \ pm$ ]{}1.0 & 6 5 [$\p m$ ]{ } 5 &62[$\pm$ ]{} 5\
250& 18.2 & 26 . 3 [$\ p m$ ]{}4.2 & 1 9.3 [$\pm$] {}3 .1 & 1 9. 0 [ $ \pm$ ] {}5.8 & [b] {} & $>$52[ b]{} & 55 [$\ pm$ ]{ } 22\
350 & 24.9& 26.3 [$ \pm $ ]{ } 6. 9& 2 2. 7 [$ \ pm $ ] { }6. 8 & [c]{ }&[c]{} & [ c ] { } & [ c]{ }\
[ a]{}[ The effective be am s i ze(geom etric mea n) is l istedand u se d to deconvolve the sizes us ing E qua ti on (\ [ eqn:co nvo lve \]).]{} [b]{}[ T hedi s k is only marginally r es o l ve d alongone ax i sat 250 . Th er efo re $ \ t heta_ {\rm { ac t,min}}$ canno t b ederived f or the s hor t a xis a n d co nseque ntly onl y the upper limit on the inclinati o nc a nb e ca lcu lated.]{} [ c]{} [ Thesyst e misc onsis tentwi t hb eing unresolved and s ymmetr ic at 350 .]{}
Th e size oft h e disk ca n be es t imated by fitt ing 2 D Gaussian s to theimage s. First , a model o f the ph oto sph ere is r em oved from the i mage b y subtr act ing a P SFtha t i s s ca led to th e expect ed p ho to sph erice mission(S ect io n \ [sec: p hotmod el\]) . Th eFW H Msand pos i ti o n ang le sof t he2D Gaus sian mod els are listed i n T a ble\[ tb l:gauss ianFits\]. Th equoted err or on a fit t e d parame ter gives the range ofp aramete r v alues for which th e [ $\chi^ 2$] { } valu e is w ithin 1 0%o f them i ni mum [ $\ | .7 &_7.9 [$\pm$ ]{}0.4_& 11.4 [$\pm$ ]{}0.8 &_5.6 [$\pm$ ]{}0.6_&_61 [$\pm$ ]{}6_&_56 [$\pm$ ]{}6\
100 &_6.79 & 14.2_[$\pm$ ]{}0.7 & 8.2 [$\pm$ ]{}0.4_& 12.5 [$\pm$ ]{}0.8_&_4.6 [$\pm$ ]{}0.7 & 68 [$\pm$ ]{}5 & 52 [$\pm$ ]{}4\
160 & 11.36 & 20.2 [$\pm$ ]{}0.8 &_13.3_[$\pm$ ]{}0.5 &_16.6_[$\pm$ ]{}1.0_& 6.9 [$\pm$ ]{}1.0 & 65_[$\pm$ ]{}5 & 62 [$\pm$ ]{}5\
250 &_18.2 &_26.3 [$\pm$ ]{}4.2 & 19.3 [$\pm$ ]{}3.1 & 19.0 [$\pm$ ]{}5.8_&_[b]{} & $>$52[b]{}_& 55 [$\pm$ ]{}22\
350 & 24.9 & 26.3 [$\pm$ ]{}6.9 &_22.7 [$\pm$ ]{}6.8 & [c]{} & [c]{}_& [c]{} &_[c]{}\
[a]{}[The_effective_beam size (geometric mean)_is listed and used to deconvolve_the sizes using Equation (\[eqn:convolve\]).]{} [b]{}[The disk_is only marginally resolved along one axis_at 250 . Therefore $\theta_{\rm{act,min}}$ cannot be_derived for the short axis_and consequently_only the upper limit on_the inclination can_be calculated.]{}_[c]{}[The system is_consistent with being unresolved and symmetric_at 350 .]{}
The size_of the disk can be estimated_by_fitting 2D Gaussians_to_the_images. First,_a model of_the_photosphere is_removed_from the image by subtracting a_PSF_that is scaled to the expected photospheric_emission (Section \[sec:photmodel\]). The FWHMs_and_position angles of the_2D Gaussian models are listed_in Table \[tbl:gaussianFits\]. The quoted error on_a fitted_parameter gives_the range of parameter values for which the [$\chi^2$]{} value is within_10% of the minimum [$\ |
or treated using simple approximations. For instance, Gotoh and Finney [@Gotoh1974] estimated the density of RCP based only on correlations among the contact neighbors. Another example is the ‘granocentric’ model [@Clusel2009], which considers the correlations between the central particle and nearest Voronoi neighbors. Beyond local correlations, the statistical treatment of Song [*et al.*]{} takes a mean-field approximation of the long-range correlations [@Song2008; @Jin2010B]. Other mean-field approaches are developed based on liquid theories [@Berryman1983] and replica theory (RT) of the glass transition [@Parisi2010]. However, in low dimensions the effects of fluctuations are strong and mean-field approximations are insufficient. For example, in 2d, the coarse-grained approximation used in Ref. [@Song2008] works poorly [@Meyer2010], and therefore more sophisticated treatments of correlations become necessary.
In this paper, we aim to establish a framework for random packings that addresses the two problems stated above: [*(i)*]{} we define an ensemble of equiprobable graphs that satisfy the jamming conditions to represent the statistics of all possible contact networks, and [*(ii)*]{} we take into account pair and higher-order particle correlations that are important to describe low-dimensional systems. Inspired by the more advanced liquid theories, our formulation is analogous to the Yvon-Born-Green (YBG) hierarchy [@Hansen1986] augmented to consider the contact network and local and global jamming conditions for packings. We develop a systematic layer-expansion within a Kirkwood-like superposition approximation [@Hansen1986] to provide a phase diagram and predictions of the volume fractions of jammed packings. The theoretical predictions on volume fractions and pair distribution functions agree well with experiments and computer simulations on two-dimensional frictional packings. We also discuss the relation between the present approach and glass theory frameworks in search of unifications of random packings and glasses [@Parisi2010; @Krzakala2007; @Mari2008; @Mezard2011].
The present approach builds up on the Edwards mean-field theory of packings developed by Song et al. [@Song2008], by incorporating correlations between the particle positions. Previous theory [@Song2008] utilizes a mean-field assumption of uniformity of the particle density in the bulk as well as the particles in contact. The present theory is a bottom-up approach to take into account particle-particle correlations | or treated using simple approximations. For case, Gotoh and Finney [ @Gotoh1974 ] estimate the density of RCP based only on correlation coefficient among the contact neighbors. Another exercise is the ‘ granocentric ’ mannequin [ @Clusel2009 ], which considers the correlation between the central particle and cheeseparing Voronoi neighbors. Beyond local correlations, the statistical discussion of Song [ * et al. * ] { } takes a mean - field approximation of the long - range correlations [ @Song2008; @Jin2010B ]. early mean - field approaches are evolve based on liquid theory [ @Berryman1983 ] and replica theory (RT) of the field glass transition [ @Parisi2010 ]. However, in low dimensions the effects of fluctuations are potent and mean - field approximations are insufficient. For example, in 2d, the coarse - grained estimate use in Ref. [ @Song2008 ] works poorly [ @Meyer2010 ], and therefore more sophisticated treatments of correlations become necessary.
In this paper, we aim to establish a model for random packing material that addresses the two trouble stated above: [ * (i) * ] { } we define an ensemble of equiprobable graph that satisfy the jamming condition to represent the statistics of all potential contact networks, and [ * (ii) * ] { } we take into account pair and higher - order particle correlations that are important to describe low - dimensional system. inspire by the more advanced liquid theories, our formulation is analogous to the Yvon - Born - Green (YBG) hierarchy [ @Hansen1986 ] augmented to consider the liaison network and local and ball-shaped jamming conditions for packings. We develop a systematic layer - expansion within a Kirkwood - comparable superposition approximation [ @Hansen1986 ] to supply a phase diagram and predictions of the bulk fractions of jammed packing material. The theoretical predictions on book fractions and pair distribution functions agree well with experiments and calculator simulations on two - dimensional frictional packings. We also hash out the relation between the present approach and glass theory frameworks in search of fusion of random packings and glasses [ @Parisi2010; @Krzakala2007; @Mari2008; @Mezard2011 ].
The present access builds up on the Edwards mean - field theory of packings developed by Song et al. [ @Song2008 ], by incorporate correlations between the particle positions. Previous theory [ @Song2008 ] utilizes a average - sphere premise of uniformity of the particle density in the bulk as well as the particles in contact. The present hypothesis is a bottom - up approach to take into account particle - particle correlations | or treated using simple apkroximations. For nbstancx, Gotoh and Finvey [@Gotoh1974] estimated the densmty if RCK based only on corfelations among tye cibtact neigigors. Another swamplz ms the ‘granocenjric’ model [@Cnusel2009], which cotskdzrs the correlations between the ceneral pattlcle and neareft Vpwonoj neighbors. Beyond local correlatikns, the statistical yreatment of Song [*et al.*]{} tanes w mean-field approxlmation of jge jing-range corfelations [@Song2008; @Jin2010B]. Otger mean-field approaches are dexelopzd based on luqulg theories [@Uerrymwn1983] and repligs theosy (RT) og the glass trsnsmtiob [@Parisi2010]. However, in lmw dimensions the gffects of fmuctuations are srring atd maan-fkwld apkroeimztions arx insufficisnt. For exanple, in 2d, the coarst-grwpmed approximztion tsqd in Ref. [@Song2008] works poorly [@Meyer2010], and thegefode more sophisticated teeatments of correlatlons becoie necessary.
In this paper, we aim to establish a fsamewkfk yir raveol packings that addresses the two problems stwfec sbove: [*(i)*]{} we deflne an ensemble of eeuokrobable graphr that safisfy the jamming fonditijns ti represegt tne statistics of all possiboe contact nvtwoeks, and [*(ii)*]{} we take into accouut pait and nigher-order particle coxrelatjons that age importzvt to describe luw-dpmenvional systems. Inspired by the more advcnced liduid theoryes, our fogmulabhon is analogous tl the Yeon-Born-Grefn (YBG) hierarchy [@Hansen1986] augmentxv to consider tve bontact nztwork and local anq global jammiug condicions wor packinfs. We dxvelop a syseematic layer-adpansion witiin a Kirhwooe-likw superousition approxomation [@Hcusen1986] to peovide a phase dianram xhd predictions if uhw volume fractoonr os uakmqg packings. Tve tfeofrticau predictioks un vplume fractions and [air distribution funcyijns agreg well wieh experimentx and computer simklatimns on teo-dymensional frictional packings. We also fisguss the relaeion between the present approach and glass theory framxworks in search of unidications of random kackings and jlassef [@Parisi2010; @Nrzakala2007; @Mari2008; @Mezard2011].
Rhe present approsch builds up on the Sdwardv meaj-field theory of packings developed by Song et al. [@Song2008], by incorporating cirrelavijns between the partpclz ppsitiogs. Pcetious theory [@Song2008] utilizes a mean-field assumptioi of unifosmnty of the particle density im ghe bulk as wgll as the particles in cohtact. Thr present theory is a bottom-up spproach to take into acgount parricle-larticle currelations | or treated using simple approximations. For instance, Finney estimated the of RCP based contact Another example is ‘granocentric’ model [@Clusel2009], considers the correlations between the central and nearest Voronoi neighbors. Beyond local correlations, the statistical treatment of Song [*et takes a mean-field approximation of the long-range correlations [@Song2008; @Jin2010B]. Other mean-field approaches developed on theories and replica theory (RT) of the glass transition [@Parisi2010]. However, in low dimensions the effects of are strong and mean-field approximations are insufficient. For in 2d, the coarse-grained used in Ref. [@Song2008] works [@Meyer2010], therefore more treatments correlations necessary. In this we aim to establish a framework for random packings that addresses the two problems stated above: [*(i)*]{} define an equiprobable graphs satisfy jamming to represent the all possible contact networks, and [*(ii)*]{} account pair and higher-order particle correlations that are to describe systems. Inspired by the more advanced theories, our formulation is analogous to the Yvon-Born-Green hierarchy [@Hansen1986] augmented to consider the contact network and local and global jamming conditions for develop a systematic layer-expansion a Kirkwood-like superposition [@Hansen1986] provide phase and predictions the volume fractions of jammed packings. The theoretical predictions on volume and pair distribution functions agree well with experiments and computer two-dimensional packings. We also the relation between the approach glass theory frameworks in unifications random [@Parisi2010; @Mari2008; The present approach builds on the Edwards mean-field theory packings developed by Song correlations between the particle positions. Previous theory [@Song2008] a mean-field assumption of uniformity of the density in the bulk as well as the particles in contact. The theory is approach to take into account particle-particle correlations | or treated using simple approXimations. FOr insTanCe, GOtOh anD FinNey [@Gotoh1974] estimaTEd thE density of RCP based only On corReLAtioNS aMong tHe contaCT nEIGhbOrS. ANotHeR ExAmple Is tHe ‘granoCentric’ modEl [@CLuSel2009], which consIDeRs the correLatIons between tHe cEntral PaRtiCLe and NeaRest VOronoi NEighboRs. Beyond lOcAL correLAtions, tHE StAtisTical treatment of SONg [*ET al.*]{} takes a mean-fIeld apPrOXiMATioN of The long-ranGe CorreLAtions [@SONg2008; @jIN2010b]. OtHEr mean-field apProaches are DEveLoped bAsEd oN Liquid TheorIeS [@berRyman1983] and repLica Theory (RT) oF the glASs transITion [@ParIsi2010]. HowEveR, in Low dIMeNsIonS tHE efFEcTs oF FluCtuationS aRe StronG and MEAN-FielD apProxImatiOns are insuffiCieNt. FoR ExaMple, iN 2d, the CoarSe-GrainEd apprOximaTiOn used in Ref. [@Song2008] WorkS poorly [@MeYer2010], AnD thErEfore MOre sopHisTicAted treAtments OF coRrELATiOns become necessary.
in THIs Paper, we aIm to esTAbLiSH a framewOrK foR ranDOM packIngs THaT addressEs the tWO pRoBlems stAtEd abovE: [*(i)*]{} We dEfiNe an eNSembLe of eqUiprobabLe graPHs that satisfy tHE jamming condiTIoNS To REpreSenT the statistIcs oF All pOssiBLe ConTAct neTworkS, aND [*(iI)*]{} We take into account paIr And higHer-orDer particle coRrelations THAT are impoRtanT To DEscribe low-dimeNsionAl systems. INSpired by The moRe advancEd liquid tHEOries, our ForMulAtiOn iS ANaLogous to the YvON-born-grEen (YBG) hIerArchy [@HaNseN1986] auGmeNteD tO consider The contaCt NeTwOrK anD locaL And globaL jAmmInG coNditiONs for pAckinGs. We DeVeLOp a SystemaTIc LAYer-eXpAnSion WitHiN a KirKwooD-LikE superpOsition apProXImatIoN [@HAnsen1986] to Provide a phase DiAgram and prEdIctIons of THE volume fRactions of jammed packingS. the theoRetIcal pRediCtions on vOluMe fracTioNS and paIr distRibutIoN fuNCTions AGReE weLl With experiMENts And coMpUter SimulatIons on two-dimensionAL frIctional packiNgs. we alSO DiScuSS tHE reLaTIon BETween the present Approach anD gLAsS theory fraMEwoRkS in searCh of uniFicatIOns of raNdom packiNgs and glaSsEs [@PaRISi2010; @KRzakala2007; @MarI2008; @Mezard2011].
THe present APproaCH bUilds Up oN the EdWaRds Mean-fIeld thEOry Of pacKings dEvEloped By SonG eT al. [@Song2008], bY incorporating correlatiOns betWeen tHe pArticle poSitIOns. previous tHeorY [@Song2008] utiliZes A meAn-fieLd aSSumptIon oF UnIfoRMity oF the PArticle deNSiTy iN THe Bulk as well aS THE paRticlEs iN ContacT. The Present theory is a bOTtom-up approach To taKE IntO acCOunt PaRticle-particle CorReLATions | or treated using simple a pproximati ons.For in st ance , Go toh and Finney [@Go toh1974] estimated the dens it y ofR CP base d onlyo nc o rre la ti ons a m on g the co ntact n eighbors.Ano th er example i s t he ‘granoc ent ric’ model [ @Cl usel20 09 ],w hichcon sider s thec orrela tions bet we e n thec entralp a rt icle and nearest Voro n oi neighbors. Bey ond lo ca l c o r rel ati ons, the s ta tisti c al trea t me n t ofS ong [*et al.* ]{} takes a mea n-fiel dapp r oximat ion o ft helong-rangecorr elations[@Song 2 008; @J i n2010B] . Othe r m ean -fie l dap pro ac h esa re de v elo ped base don liqu id t h e o r ies[@B erry man19 83] and repli catheo r y ( RT) o f the gla ss tran sition [@Pa ri si2010]. Howeve r, i n low dim ens io nsth e eff e cts of fl uct uations are st r ong a n d me an-field approxima ti o n sare insu fficie n t. F o r exampl e, in 2d, t he co arse - gr ained ap proxim a ti on used i nRef. [ @S ong 200 8] wo r ks p oorly[@Meyer2 010], and thereforem ore sophistic a te d tr e atme nts of correla tion s bec omen ec ess a ry.
In th is pa p er, we aim to estab li sh a f ramew ork for rando m packings t h at addre sses th e two problemsstate d above: [ * (i)*]{}we de fine anensembleo f equipro bab legra phs t ha t satisfy the j ammi ng condit ion s to re pre sen t t hest atisticsof all p os si bl econ tactn etworks, a nd[* (ii )*]{} we tak e int o ac co un t pa ir andh ig h e r-or de rpart icl ecorre lati o nsthat ar e importa ntt o de sc ri be low- dimensional s ys tems. Insp ir edby the m ore adva nced liquid theories, o u r formu lat ion i s an alogous t o t he Yvo n-B o rn-Gre en (YB G) hi er arc h y [@Ha n s en 198 6] augmented t o c onsid er the contac t network and loca l an d global jamm ing con d i ti ons fo r pa ck i ngs . We develop a sy stematic l ay e r- expansionw ith in a Kirk wood-li ke su p erposit ion appro ximation[@ Hans e n 198 6] to prov ide a ph ase diagr a m and pr edict ion s of t he vo lumefracti o nsof ja mmed p ac kings. Theth eoretica l predictions on volume fract ionsand pair dis tri b uti on functi onsagree well wi thexper ime n ts an d co m pu ter simul atio n s on two- d im ens i o na l frictiona l p ack ings. We also d iscu ss the relation b e tween the pres enta p pro ach andgl ass theory fra mew or k s in sear ch of unifica tions of r a ndompackin gs and glasse s [@ P arisi2 010; @K rzakala20 07; @ M ari2008 ;@M e zard20 11].
The pr esenta ppro a c h builds up on t he Ed w a rds m e an- field t heory o f pac kings deve loped by So ng etal.[@Son g2008], b y inco rpo ra ting corre l ations be tween the pa rt icle po sition s. P r e vious the or y [ @Song2008 ] ut i li ze s amean -fiel dassu mption of uniformi tyo f the p ar tic l e densi t yi n the bulkaswella s the parti c lesi nc ontac t. The prese nt theo r y i sa botto m-u p approachto take i n toac coun t partic le -par ti cle c orrelations | or_treated using_simple approximations. For instance,_Gotoh and_Finney_[@Gotoh1974] estimated_the_density of RCP_based only on_correlations among the contact_neighbors. Another example_is_the ‘granocentric’ model [@Clusel2009], which considers the correlations between the central particle and nearest_Voronoi_neighbors. Beyond_local_correlations,_the statistical treatment of Song_[*et al.*]{} takes a mean-field_approximation of_the long-range correlations [@Song2008; @Jin2010B]. Other mean-field approaches_are_developed based on_liquid theories [@Berryman1983] and replica theory (RT) of the_glass transition [@Parisi2010]. However, in low_dimensions the effects_of_fluctuations_are strong and mean-field_approximations are insufficient. For example, in_2d, the coarse-grained approximation used in_Ref. [@Song2008] works poorly [@Meyer2010], and therefore more sophisticated_treatments of correlations become necessary.
In this_paper, we aim to establish_a framework_for random packings that addresses_the two problems_stated above:_[*(i)*]{} we define_an ensemble of equiprobable graphs that_satisfy the jamming_conditions to represent the statistics of_all_possible contact networks,_and_[*(ii)*]{}_we take_into account pair_and_higher-order particle_correlations_that are important to describe low-dimensional_systems._Inspired by the more advanced liquid theories,_our formulation is analogous_to_the Yvon-Born-Green (YBG) hierarchy_[@Hansen1986] augmented to consider the_contact network and local and global_jamming conditions_for packings._We develop a systematic layer-expansion within a Kirkwood-like superposition approximation [@Hansen1986]_to provide a phase diagram and_predictions of the volume_fractions of_jammed_packings. The theoretical_predictions_on volume_fractions and pair distribution functions agree well_with experiments_and computer simulations on two-dimensional frictional_packings. We also discuss_the_relation between the present approach and_glass theory frameworks in search of_unifications of random packings and_glasses_[@Parisi2010;_@Krzakala2007; @Mari2008; @Mezard2011].
The present approach_builds up on the Edwards mean-field_theory of packings_developed by Song et al. [@Song2008], by_incorporating_correlations between the particle positions. Previous_theory_[@Song2008] utilizes a mean-field assumption of_uniformity_of_the particle density in the_bulk as well as the particles_in contact. The present theory is a bottom-up approach_to take into_account particle-particle correlations |
onic isometry group of gets broken down to $SU(1,1) \times SU(2) \times U(1)$ [@Gauntlett:1998fz]. By completing the square term $-(\frac{2\tilde{r}}{\sqrt{\mu}}dt+ \frac{a}{2}\sigma_3)^2$, the near-horizon topology of the BMPV black hole becomes $AdS_2$ fibered over $S^3$ [@Li:2006uq]. This may be a sign of the existence of certain putative dual 2D chiral CFT. From it is easily seen that the tree-level entropy is $$S_{0} = \frac{{\rm Area}}{4 G_5} = \frac{\pi^2}{2 G_5} \mu \sqrt{\mu - a^{2}}. \label{entropy}$$
Entropy from the Iyer-Wald formula
----------------------------------
Let us see how $R^2$-curvature corrections modify. When it comes to the Gauss-Bonnet terms, the result has actually been calculated in [@Guica:2005ig]. Our goal here is to reproduce the known result and meantime set up some notational conventions to be used in the next section.
The 5D ${\mathcal N}=2$ supergravity can arise from compactifying 11D M-theory on Calabi-Yau threefolds $X$ where the number of vector multiplets is $h^{1,1}(X)$. The resultant Einstein-frame low-energy effective action may necessarily contain higher-derivative corrections [@hep-th/9602102; @hep-th/9707013] such as the familiar Gauss-Bonnet term: $$\begin{aligned}
\begin{split}
\delta {\mathcal L}=
\xi (R_{\mu \nu \rho \sigma} R^{\mu \nu \rho \sigma} -4 R_{\mu \nu}R^{ \mu \nu} + R^2)
\label{higher}
\end{split}\end{aligned}$$ where $\xi$ is is proportional to $c_{2 A}Y^{A}$. As mentioned before, $Y^{A}$ denotes the scalar component in vector multiplets evaluated at the horizon. $c_{2 A}$ is given by $$\begin{aligned}
\begin{split}
c_{2 A}=\int_X c_2( | onic isometry group of gets broken down to $ SU(1,1) \times SU(2) \times U(1)$ [ @Gauntlett:1998fz ]. By completing the square condition $ -(\frac{2\tilde{r}}{\sqrt{\mu}}dt+ \frac{a}{2}\sigma_3)^2 $, the approximate - horizon topology of the BMPV black hole become $ AdS_2 $ fibered over $ S^3 $ [ @Li:2006uq ]. This may be a sign of the existence of sealed putative dual 2D chiral CFT. From it is easily seen that the tree - degree entropy is $ $ S_{0 } = \frac{{\rm Area}}{4 G_5 } = \frac{\pi^2}{2 G_5 } \mu \sqrt{\mu - a^{2 } }. \label{entropy}$$
Entropy from the Iyer - Wald formula
----------------------------------
lease us see how $ R^2$-curvature corrections change. When it comes to the Gauss - Bonnet terms, the resultant role has actually been calculated in [ @Guica:2005ig ]. Our goal here is to regurgitate the known result and meantime determine up some notational conventions to be used in the following section.
The 5D $ { \mathcal N}=2 $ supergravity can originate from compactifying 11D M - theory on Calabi - Yau threefolds $ X$ where the number of vector multiplets is $ h^{1,1}(X)$. The resultant Einstein - frame low - department of energy effective action may necessarily contain higher - derivative corrections [ @hep - th/9602102; @hep - th/9707013 ] such as the familiar Gauss - Bonnet condition: $ $ \begin{aligned }
\begin{split }
\delta { \mathcal L}=
\xi (R_{\mu \nu \rho \sigma } R^{\mu \nu \rho \sigma } -4 R_{\mu \nu}R^ { \mu \nu } + R^2)
\label{higher }
\end{split}\end{aligned}$$ where $ \xi$ is is proportional to $ c_{2 A}Y^{A}$. As mentioned before, $ Y^{A}$ denotes the scalar component in vector multiplets evaluated at the horizon. $ c_{2 A}$ is given by $ $ \begin{aligned }
\begin{split }
c_{2 A}=\int_X c_2 ( | onif isometry group of gets broken down to $SU(1,1) \timxs SU(2) \tjmes U(1)$ [@Gxuntlett:1998fz]. By completing the swuare term $-(\frac{2\tilde{r}}{\sqrt{\mu}}at+ \frac{a}{2}\spgma_3)^2$, the bear-iorizon topology of the NIPV gpack iole becomes $AdX_2$ fibered mver $S^3$ [@Li:2006uq]. Thhs mcy be a sign of the existence of cereain puyahive dual 2D chyral SFT. Rgon it is easily seen that the free-levtl entropy is $$S_{0} = \ftac{{\rm Area}}{4 G_5} = \frac{\pi^2}{2 G_5} \mu \dqrt{\lu - a^{2}}. \label{entropy}$$
Fntropy fron thq Iyer-Wald fofmula
----------------------------------
Let us see how $R^2$-corvature corrections modify. When it cpmes to thg Fakvs-Bonnet tecms, thv result has actually been cslculated in [@Gmica:2005ij]. Oue goal here is to repcoduce the known resolt and medncime set up some notariinal wonvantiubs go ge uaed in thx next sectjon.
The 5D ${\marhcal N}=2$ supergravitu sqn arise from compastyfying 11D M-theory on Calabi-Yau threefoldv $X$ where the number of vextor multiplets is $h^{1,1}(X)$. The resujtant Einstein-frame low-energy effective action maf necxsrarnoy covraln higher-derivative corrections [@hep-th/9602102; @hep-th/9707013] stdh ax the familiar Gauss-Bonney heti: $$\begin{alignea}
\begin{spljt}
\delta {\mathcal L}=
\xi (R_{\mu \nu \ryo \sigma} W^{\mu \mu \rho \sigma} -4 R_{\mu \nu}R^{ \mu \nu} + R^2)
\label{highvr}
\ene{split}\end{aligned}$$ wkere $\xi$ is ir prpportoonal to $c_{2 A}Y^{A}$. As mentiuned before, $Y^{A}$ fenotes tgd scalar componevt pn vactor multiplets evaluated at the hirizpn. $c_{2 A}$ ks goven br $$\begin{alihned}
\bccin{split}
c_{2 A}=\int_X c_2( | onic isometry group of gets broken down \times \times U(1)$ By completing the near-horizon of the BMPV hole becomes $AdS_2$ over $S^3$ [@Li:2006uq]. This may be sign of the existence of certain putative dual 2D chiral CFT. From it easily seen that the tree-level entropy is $$S_{0} = \frac{{\rm Area}}{4 G_5} = G_5} \sqrt{\mu a^{2}}. Entropy from the Iyer-Wald formula ---------------------------------- Let us see how $R^2$-curvature corrections modify. When it comes the Gauss-Bonnet terms, the result has actually been in [@Guica:2005ig]. Our goal is to reproduce the known and set up notational to used in the section. The 5D ${\mathcal N}=2$ supergravity can arise from compactifying 11D M-theory on Calabi-Yau threefolds $X$ where number of is $h^{1,1}(X)$. resultant low-energy action may necessarily corrections [@hep-th/9602102; @hep-th/9707013] such as the $$\begin{aligned} \begin{split} \delta {\mathcal L}= \xi (R_{\mu \nu \sigma} R^{\mu \rho \sigma} -4 R_{\mu \nu}R^{ \mu + R^2) \label{higher} \end{split}\end{aligned}$$ where $\xi$ is is to $c_{2 A}Y^{A}$. As mentioned before, $Y^{A}$ denotes the scalar component in vector multiplets evaluated horizon. $c_{2 A}$ is by $$\begin{aligned} \begin{split} A}=\int_X | onic isometry group of gets brOken down to $sU(1,1) \timEs Su(2) \tiMeS U(1)$ [@GaUntlEtt:1998fz]. By completINg thE square term $-(\frac{2\tilde{r}}{\sQrt{\mu}}Dt+ \FRac{a}{2}\SIgMa_3)^2$, the Near-horIZoN TOpoLoGy Of tHe bmPv blacK hoLe becomEs $AdS_2$ fiberEd oVeR $S^3$ [@Li:2006uq]. This maY Be A sign of the ExiStence of certAin PutatiVe DuaL 2d chirAl CfT. FroM it is eASily seEn that the TrEE-level ENtropy iS $$s_{0} = \FrAc{{\rm area}}{4 G_5} = \frac{\pi^2}{2 G_5} \mu \sqrT{\Mu - A^{2}}. \Label{entropy}$$
EnTropy fRoM ThE iYer-walD formula
----------------------------------
LeT uS see hOW $R^2$-curvaTUrE CORreCTions modify. WhEn it comes to THe GAuss-BoNnEt tERms, the ResulT hAS acTually been cAlcuLated in [@GuIca:2005ig]. OUR goal heRE is to reProducE thE knOwn rESuLt And MeANtiME sEt uP SomE notatioNaL cOnvenTionS TO BE useD in The nExt seCtion.
The 5D ${\mathCal n}=2$ supERgrAvity Can arIse fRoM compActifyIng 11D M-ThEory on Calabi-Yau ThreEfolds $X$ whEre ThE nuMbEr of vECtor muLtiPleTs is $h^{1,1}(X)$. THe resulTAnt eiNSTEiN-frame low-energy effEcTIVe Action maY necesSArIlY Contain hIgHer-DeriVATive cOrreCTiOns [@hep-th/9602102; @Hep-th/9707013] sUCh As The famiLiAr GausS-BOnnEt tErm: $$\beGIn{alIgned}
\bEgin{spliT}
\deltA {\Mathcal L}=
\xi (R_{\mu \nU \Rho \sigma} R^{\mu \nu \RHo \SIGmA} -4 r_{\mu \nU}R^{ \mU \nu} + R^2)
\label{hiGher}
\ENd{spLit}\eND{aLigNEd}$$ wheRe $\xi$ iS iS PrOPortional to $c_{2 A}Y^{A}$. As meNtIoned bEfore, $y^{A}$ denotes the sCalar compoNENT in vectoR mulTIpLEts evaluated at The hoRizon. $c_{2 A}$ is gIVen by $$\begIn{aliGned}
\begiN{split}
c_{2 A}=\iNT_x c_2( | onic isometry group of get s broken d own t o $ SU( 1, 1) \ time s SU(2) \times U(1) $ [@Gauntlett:1998fz]. By c om p leti n gthe s quare t e rm $ -(\ fr ac {2\ ti l de {r}}{ \sq rt{\mu} }dt+ \frac {a} {2 }\sigma_3)^2 $ ,the near-h ori zon topology of the B MP V b l ack h ole beco mes $A d S_2$ f ibered ov er $S^3$[ @Li:200 6 u q] . Th is may be a signo ft he existence o f cert ai n p u t ati vedual 2D ch ir al CF T . Fromi ti s eas i ly seen thatthe tree-le v elentrop yis$ $S_{0} = \f ra c {{\ rm Area}}{4 G_5 } = \frac {\pi^2 } {2 G_5} \mu \sq rt{\mu -a^{ 2}}. \l ab el{ en t rop y }$ $
E ntr opy from t he Iyer -Wal d f o rmul a
- ---- ----- ------------- --- ---- - ---
Let us s ee h ow $R^2 $-curv ature c orrections modi fy.When it c ome stoth e Gau s s-Bonn etter ms, the result has a c t u al ly been calculated i n [@ Guica:20 05ig]. Ou rg oal here i s t o re p r oduce the kn own resu lt and me an time se tup som enot ati onalc onve ntions to be u sed i n the next sect i on.
The 5D $ { \m a t hc a l N} =2$ supergravi ty c a n ar isef ro m c o mpact ifyin g1 1D M-theory on Calabi- Ya u thre efold s $X$ where t he numbero f vector m ulti p le t s is $h^{1,1}( X)$.The result a nt Einst ein-f rame low -energy e f f ective a cti onmay ne c e ss arily contain h ighe r- derivat ive correc tio ns[@h ep- th /9602102; @hep-th /9 70 70 13 ] s uch a s the fam il iar G aus s-Bon n et ter m: $$ \beg in {a l ign ed}
\be g in { s plit }\d elta {\ ma thcal L}=
\ xi (R_{ \mu \nu \ rho \sig ma }R^{\mu\nu \rho \sig ma } -4 R_{\m u\nu }R^{ \ m u \nu} +R^2)
\label{higher}
\en d {split} \en d{ali gned }$$ where $\ xi$ is is propor tional to $ c_ {2A } Y^{A} $ . A s m en tioned bef o r e,$Y^{A }$ den otes th e scalar component invector multip let s ev a l ua ted at the h o riz o n . $c_{2 A}$ isgiven by $ $\ b eg in{aligned }
\b eg in{spli t}
c_{2 A}=\ i nt_X c_ 2( | onic isometry_group of_gets broken down to_$SU(1,1) \times_SU(2)_\times U(1)$_[@Gauntlett:1998fz]._By completing the_square term $-(\frac{2\tilde{r}}{\sqrt{\mu}}dt+_\frac{a}{2}\sigma_3)^2$, the near-horizon topology_of the BMPV_black_hole becomes $AdS_2$ fibered over $S^3$ [@Li:2006uq]. This may be a sign of the_existence_of certain_putative_dual_2D chiral CFT. From it_is easily seen that the_tree-level entropy_is $$S_{0} = \frac{{\rm Area}}{4 G_5} = \frac{\pi^2}{2_G_5}_\mu \sqrt{\mu -_a^{2}}. \label{entropy}$$
Entropy from the Iyer-Wald formula
----------------------------------
Let us see how_$R^2$-curvature corrections modify. When it comes_to the Gauss-Bonnet_terms,_the_result has actually been_calculated in [@Guica:2005ig]. Our goal here_is to reproduce the known result_and meantime set up some notational conventions_to be used in the next_section.
The 5D ${\mathcal N}=2$ supergravity_can arise_from compactifying 11D M-theory on_Calabi-Yau threefolds $X$_where the_number of vector_multiplets is $h^{1,1}(X)$. The resultant Einstein-frame_low-energy effective action_may necessarily contain higher-derivative corrections [@hep-th/9602102;_@hep-th/9707013]_such as the_familiar_Gauss-Bonnet_term: $$\begin{aligned}
\begin{split}
\delta_{\mathcal L}=
_\xi_(R_{\mu \nu_\rho_\sigma} R^{\mu \nu \rho \sigma} -4_R_{\mu_\nu}R^{ \mu \nu} + R^2)
\label{higher}
\end{split}\end{aligned}$$ where $\xi$_is is proportional to_$c_{2_A}Y^{A}$. As mentioned before,_$Y^{A}$ denotes the scalar component_in vector multiplets evaluated at the_horizon. $c_{2_A}$ is_given by $$\begin{aligned}
\begin{split}
c_{2 A}=\int_X c_2( |
} (M^{-1})_{ab} h_{bjk},$$ where $y_{ia}$ are the couplings of $q_i (Q^c_1)_a \phi$, $h_{bjk}$ those of $(Q_1)_b q^c_j \sigma_k$, and $M$ the mass matrix of $Q_1 Q^c_1$. Since the family symmetry applies to all three of these quantities, it is well hidden in the resulting effective operator of Eq. (1) and even more so in the resulting mass matrix $$m_{ij} = {f_{ijk} \over \Lambda} \langle \sigma_k \rangle \langle \phi
\rangle.$$ On the positive side, if this particular mechanism is assumed, specific models of family structure may be considered and then compared to the data. Singlet quarks of charge $-1/3$ are contained in the fundamental representation of $E_6$. Hence the $d$ quarks of the SM may owe their family structure wholly or partly [@r00] to such a mechanism.
(400,225)(0,0) (125,200)(200,150) (200,100)(200,150) (200,100)(200,50) (125,0)(200,50) (275,200)(200,150)3 (275,0)(200,50)3 (115,200)\[\][$q$]{} (115,0)\[\][$q^c$]{} (215,125)\[\][$Q^c_2$]{} (201,100)\[\][$\times$]{} (215,75)\[\][$Q_2$]{} (285,200)\[\][$\sigma$]{} (285,0)\[\][$\phi$]{}
Consider next Fig. 2. Here $Q_2 \sim (3,2,1/6)$ and $Q^c_2 \sim
(3^*,2,-1/6)$ are required. Whereas these heavy vector quark doublets are not present in the of $E_6$, the corresponding heavy lepton doublets $L_2 \sim (1,2,-1/2)$ and $L^c_2 \sim (1,2,1/2)$ are, and they have been used for example in a recently proposed model [@hmp05] of late neutrino mass and baryogenesis. Thus the observed lepton family structure may be encoded with this mechanism.
(400 | } (M^{-1})_{ab } h_{bjk},$$ where $ y_{ia}$ are the couplings of $ q_i (Q^c_1)_a \phi$, $ h_{bjk}$ those of $ (Q_1)_b q^c_j \sigma_k$, and $ M$ the mass matrix of $ Q_1 Q^c_1$. Since the family symmetry apply to all three of these quantity, it is well hidden in the resulting effective operator of Eq. (1) and even more then in the resulting mass matrix $ $ m_{ij } = { f_{ijk } \over \Lambda } \langle \sigma_k \rangle \langle \phi
\rangle.$$ On the positivist english, if this particular mechanism is wear, specific models of family social organization may be considered and then compared to the data. Singlet quarks of bang $ -1/3 $ are contained in the cardinal representation of $ E_6$. therefore the $ d$ quark cheese of the SM may owe their family structure wholly or partially [ @r00 ] to such a mechanism.
(400,225)(0,0) (125,200)(200,150) (200,100)(200,150) (200,100)(200,50) (125,0)(200,50) (275,200)(200,150)3 (275,0)(200,50)3 (115,200)\[\][$q$ ] { } (115,0)\[\][$q^c$ ] { } (215,125)\[\][$Q^c_2 $ ] { } (201,100)\[\][$\times$ ] { } (215,75)\[\][$Q_2 $ ] { } (285,200)\[\][$\sigma$ ] { } (285,0)\[\][$\phi$ ] { }
Consider next Fig. 2. Here $ Q_2 \sim (3,2,1/6)$ and $ Q^c_2 \sim
(3^*,2,-1/6)$ are required. Whereas these heavy vector quark doublet are not present in the of $ E_6 $, the corresponding big lepton doublets $ L_2 \sim (1,2,-1/2)$ and $ L^c_2 \sim (1,2,1/2)$ are, and they have been used for example in a recently proposed exemplar [ @hmp05 ] of late neutrino mass and baryogenesis. Thus the observed lepton family structure may be encoded with this mechanism.
(400 | } (M^{-1})_{ah} h_{bjk},$$ where $y_{ia}$ are the couplings of $q_n (Q^c_1)_a \pii$, $h_{bjk}$ those ow $(Q_1)_b q^c_j \sigma_k$, and $M$ the mads matrux of $Q_1 Q^c_1$. Since the fxmily symletry appliew to all thcse of tmzse qhwnticixs, it is well hldden in tha resulting efxeztnve operator of Eq. (1) and even more so yn the tedulting mass mwtriq $$i_{ij} = {f_{ijk} \over \Lambda} \langle \sigma_k \rahgle \laigle \phi
\rangle.$$ On the positive side, if tjis oarticular mechanidm is assumgs, s[wcific modelr of familj structure may be considered and then comoared to the dajc. Sijclet quarks of crarge $-1/3$ are contained in the fundamental rcpresxntarion of $E_6$. Hence the $d$ quarks of the SM iay owe tvenr family structure wyooly ot pardly [@f00] to sudh a mechajisj.
(400,225)(0,0) (125,200)(200,150) (200,100)(200,150) (200,100)(200,50) (125,0)(200,50) (275,200)(200,150)3 (275,0)(200,50)3 (115,200)\[\][$q$]{} (115,0)\[\][$s^c$]{} (215,125)\[\][$Q^c_2$]{} (201,100)\[\][$\times$]{} (215,75)\[\][$Q_2$]{} (285,200)\[\][$\sigma$]{} (285,0)\[\][$\phi$]{}
Considet gvct Fig. 2. Here $S_2 \sim (3,2,1/6)$ agd $Q^c_2 \sim
(3^*,2,-1/6)$ are required. Whereas these htavy bector quark doublets aee not present in the of $E_6$, the corresponding heavy lepton doublets $L_2 \sim (1,2,-1/2)$ and $L^w_2 \sim (1,2,1/2)$ art, qnd tfwy have been used for example in a recently pro[kstd kodel [@hmp05] of lcte neutrino masx wnc baryogenesis. Thus cge observed lepton fwmily sjructuee may be encpded with this mechanism.
(400 | } (M^{-1})_{ab} h_{bjk},$$ where $y_{ia}$ are the $q_i \phi$, $h_{bjk}$ of $(Q_1)_b q^c_j matrix $Q_1 Q^c_1$. Since family symmetry applies all three of these quantities, it well hidden in the resulting effective operator of Eq. (1) and even more in the resulting mass matrix $$m_{ij} = {f_{ijk} \over \Lambda} \langle \sigma_k \rangle \phi On positive if this particular mechanism is assumed, specific models of family structure may be considered and then to the data. Singlet quarks of charge $-1/3$ contained in the fundamental of $E_6$. Hence the $d$ of SM may their structure or partly [@r00] such a mechanism. (400,225)(0,0) (125,200)(200,150) (200,100)(200,150) (200,100)(200,50) (125,0)(200,50) (275,200)(200,150)3 (275,0)(200,50)3 (115,200)\[\][$q$]{} (115,0)\[\][$q^c$]{} (215,125)\[\][$Q^c_2$]{} (201,100)\[\][$\times$]{} (215,75)\[\][$Q_2$]{} (285,200)\[\][$\sigma$]{} (285,0)\[\][$\phi$]{} next Fig. $Q_2 \sim and \sim are required. Whereas vector quark doublets are not present $E_6$, the corresponding heavy lepton doublets $L_2 \sim and $L^c_2 (1,2,1/2)$ are, and they have been for example in a recently proposed model [@hmp05] late neutrino mass and baryogenesis. Thus the observed lepton family structure may be encoded with (400 | } (M^{-1})_{ab} h_{bjk},$$ where $y_{ia}$ are the coupLings of $q_i (Q^C_1)_a \phi$, $H_{bjK}$ thOsE of $(Q_1)_B q^c_j \Sigma_k$, and $M$ the mASs maTrix of $Q_1 Q^c_1$. Since the familY symmEtRY appLIeS to alL three oF ThESE quAnTiTieS, iT Is Well hIddEn in the Resulting eFfeCtIve operator oF eq. (1) And even morE so In the resultiNg mAss matRiX $$m_{iJ} = {F_{ijk} \oVer \lambdA} \langlE \Sigma_k \Rangle \lanGlE \Phi
\ranGLe.$$ On the POSiTive Side, if this particuLAr MEchanism is assuMed, speCiFIc MODelS of Family struCtUre maY Be consiDErED ANd tHEn compared to tHe data. SinglET quArks of ChArgE $-1/3$ Are conTaineD iN The Fundamental ReprEsentatioN of $E_6$. HeNCe the $d$ qUArks of tHe SM maY owE thEir fAMiLy StrUcTUre WHoLly OR paRtly [@r00] to sUcH a MechaNism.
(400,225)(0,0) (125,200)(200,150) (200,100)(200,150) (200,100)(200,50) (125,0)(200,50) (275,200)(200,150)3 (275,0)(200,50)3 (115,200)\[\][$Q$]{} (115,0)\[\][$Q^C$]{} (215,125)\[\][$q^C_2$]{} (201,100)\[\][$\timEs$]{} (215,75)\[\][$Q_2$]{} (285,200)\[\][$\SigmA$]{} (285,0)\[\][$\phi$]{}
COnsider next FiG. 2. HeRe $Q_2 \sIM (3,2,1/6)$ anD $Q^c_2 \siM
(3^*,2,-1/6)$ are rEquiReD. WherEas theSe heaVy Vector quark doubLets Are not preSenT iN thE oF $E_6$, the COrrespOndIng Heavy lePton douBLetS $L_2 \SIM (1,2,-1/2)$ AnD $L^c_2 \sim (1,2,1/2)$ are, and they haVe BEEn Used for eXample IN a ReCEntly proPoSed ModeL [@HMp05] of lAte nEUtRino mass And barYOgEnEsis. ThuS tHe obseRvEd lEptOn famILy stRucturE may be enCoded WIth this mechaniSM.
(400 | } (M^{-1})_{ab} h_{bjk},$$ where $y_ {ia}$ ar e t he cou plin gs of $q_i (Q^ c _1)_ a \phi$, $h_{bjk}$ tho se of $ ( Q_1) _ bq^c_j \sigma _ k$ , and $ M$ th em as s mat rix of $Q_ 1 Q^c_1$.Sin ce the familys ym metry appl ies to all thre e o f thes equa n titie s,it is wellh iddenin the re su l ting e f fective o pe rato r of Eq. (1) ande ve n more so in th e resu lt i ng m ass ma trix $$m_{ ij } = { f _{ijk}\ ov e r \La m bda} \langle\sigma_k \r a ngl e \lan gl e \ p hi
\r angle .$ $ On the positi ve s ide, if t his pa r ticular mechani sm isass ume d, s p ec if icmo d els of fa m ily structu re m ay be con s i d e redand the n com pared to thedat a. S i ngl et qu arksof c ha rge $ -1/3$are c on tained in the f unda mental re pre se nta ti on of $E_6$. He nce the $d $ quark s of t h e SM may owe their fam il y st ructurewholly or p a rtly [@r 00 ] t o su c h a me chan i sm .
(400, 225)(0 , 0) ( 125,200 )( 200,15 0) (2 00, 100)( 2 00,1 50) (2 00,100)( 200,5 0 ) (125,0)(200, 5 0) (275,200)( 2 00 , 1 50 ) 3 (2 75, 0)(200,50)3 (11 5 ,200 )\[\ ] [$ q$] { } (11 5,0)\ [\ ] [$ q ^c$]{} (215,125)\[\ ][ $Q^c_2 $]{}(201,100)\[\] [$\times$] { } (215,75) \[\] [ $Q _ 2$]{} (285,200 )\[\] [$\sigma$] { } (285,0 )\[\] [$\phi$] {}
Consi d e r next F ig. 2. He re$ Q _2 \sim (3,2,1/ 6 ) $ an d$Q^c_2\si m
(3^* ,2, -1/ 6)$ ar erequired. Whereas t he se h eav y vec t or quark d oub le tsare n o t pres ent i n th eof $E_ 6$, the co r r espo nd in g he avy l epton dou b let s $L_2\sim (1,2 ,-1 / 2)$an d$L^c_2\sim (1,2,1/2 )$ are, andth eyhave b e e n used f or example in a recentl y propos edmodel [@h mp05] oflat e neut rin o massand ba ryoge ne sis . Thust h eobs er ved lepton f ami ly st ru ctur e may b e encoded with thi s me chanism.
(40 0 | } (M^{-1})_{ab}_h_{bjk},$$ where_$y_{ia}$ are the couplings_of $q_i_(Q^c_1)_a_\phi$, $h_{bjk}$_those_of $(Q_1)_b q^c_j_\sigma_k$, and $M$_the mass matrix of_$Q_1 Q^c_1$. Since_the_family symmetry applies to all three of these quantities, it is well hidden in_the_resulting effective_operator_of_Eq. (1) and even more so_in the resulting mass matrix_$$m_{ij} =_{f_{ijk} \over \Lambda} \langle \sigma_k \rangle \langle \phi_
\rangle.$$_On the positive_side, if this particular mechanism is assumed, specific models_of family structure may be considered_and then compared_to_the_data. Singlet quarks of_charge $-1/3$ are contained in the_fundamental representation of $E_6$. Hence the_$d$ quarks of the SM may owe_their family structure wholly or partly_[@r00] to such a mechanism.
(400,225)(0,0)_(125,200)(200,150) (200,100)(200,150)_(200,100)(200,50) (125,0)(200,50) (275,200)(200,150)3 (275,0)(200,50)3 (115,200)\[\][$q$]{}_(115,0)\[\][$q^c$]{} (215,125)\[\][$Q^c_2$]{} (201,100)\[\][$\times$]{}_(215,75)\[\][$Q_2$]{} (285,200)\[\][$\sigma$]{}_(285,0)\[\][$\phi$]{}
Consider next Fig. 2._Here $Q_2 \sim (3,2,1/6)$ and $Q^c_2_\sim
(3^*,2,-1/6)$ are_required. Whereas these heavy vector quark_doublets_are not present_in_the_of $E_6$,_the corresponding heavy_lepton_doublets $L_2_\sim_(1,2,-1/2)$ and $L^c_2 \sim (1,2,1/2)$ are,_and_they have been used for example in_a recently proposed model_[@hmp05]_of late neutrino mass_and baryogenesis. Thus the observed_lepton family structure may be encoded_with this_mechanism.
(400 |
}}{dt}} &= -i{\mathbf{\hat{H}}}(t){\left|\psi(t)\right>}, \nonumber \\
& {\left|\psi(0)\right>} = {\left|\psi_0\right>} \label{eq:Schr_full} \\
{\frac{d{\left|\chi(t)\right>}}{dt}} &= -i{\mathbf{\hat{H}}}^\dagger(t){\left|\chi(t)\right>} - \mathcal{C}^{-1}\left[{f_C(\omega)}{\overline{{\left<{\mathbf{\hat{C}}} \right>}}(\omega)}\right]{{\mathbf{\hat{C}}}}{\left|\psi(t)\right>}, \nonumber \\
&{\left|\chi(T)\right>} = \sigma'\left({\left<\psi(T) | \psi(T)\right>}\right){\left|\psi(T)\right>}\label{eq:ihSchr} \\
{\mathbf{\hat{H}}}(t) &= {\mathbf{\hat{H}}}_0 - {\mathbf{\hat{X}}}\epsilon(t) \label{eq:H}\\
\epsilon(t) &= \mathcal{C}^{-1}[{\bar{\epsilon}(\omega)}] \\
{\bar{\epsilon}(\omega)}&= {\bar{\epsilon}_{unc}(\omega)}- {\tilde{f}_{\epsilon}(\omega)}[\lambda_0 + \lambda_T\cos(\omega T)] \label{eq:epsw_full}\\
&{\bar{\epsilon}_{unc}(\omega)}\equiv {f_{\epsilon}(\omega)}\mathcal{C}[\eta(t)], \qquad\qquad \eta(t) \equiv -\frac{\operatorname{Im}{{\left<\chi(t) \left|{\mathbf{\hat{X}}} \right|\psi(t) \right>}}}{\alpha} \label{eq:epswunc_full}\\
&{\epsilon_{unc}}(t) = \mathcal{C}^{-1}[{\bar{\epsilon}_{unc}(\omega)}] \label{eq:epsunc}\\
&\lambda_0 = \frac{{\epsilon_{unc}}(0)d - {\epsilon_{unc}}(T)b}{ad - b^2} \label{eq:lambda0_full} \\
&\lambda_T = \frac{{\epsilon_{unc}}(T)a | } } { dt } } & = -i{\mathbf{\hat{H}}}(t){\left|\psi(t)\right > }, \nonumber \\
& { \left|\psi(0)\right > } = { \left|\psi_0\right > } \label{eq: Schr_full } \\
{ \frac{d{\left|\chi(t)\right>}}{dt } } & = -i{\mathbf{\hat{H}}}^\dagger(t){\left|\chi(t)\right > } - \mathcal{C}^{-1}\left[{f_C(\omega)}{\overline{{\left<{\mathbf{\hat{C } } } \right>}}(\omega)}\right]{{\mathbf{\hat{C}}}}{\left|\psi(t)\right > }, \nonumber \\
& { \left|\chi(T)\right > } = \sigma'\left({\left<\psi(T) | \psi(T)\right>}\right){\left|\psi(T)\right>}\label{eq: ihSchr } \\
{ \mathbf{\hat{H}}}(t) & = { \mathbf{\hat{H}}}_0 - { \mathbf{\hat{X}}}\epsilon(t) \label{eq: H}\\
\epsilon(t) & = \mathcal{C}^{-1}[{\bar{\epsilon}(\omega) } ] \\
{ \bar{\epsilon}(\omega)}&= { \bar{\epsilon}_{unc}(\omega)}- { \tilde{f}_{\epsilon}(\omega)}[\lambda_0 + \lambda_T\cos(\omega T) ] \label{eq: epsw_full}\\
& { \bar{\epsilon}_{unc}(\omega)}\equiv { f_{\epsilon}(\omega)}\mathcal{C}[\eta(t) ], \qquad\qquad \eta(t) \equiv -\frac{\operatorname{Im}{{\left<\chi(t) \left|{\mathbf{\hat{X } } } \right|\psi(t) \right>}}}{\alpha } \label{eq: epswunc_full}\\
& { \epsilon_{unc}}(t) = \mathcal{C}^{-1}[{\bar{\epsilon}_{unc}(\omega) } ] \label{eq: epsunc}\\
& \lambda_0 = \frac{{\epsilon_{unc}}(0)d - { \epsilon_{unc}}(T)b}{ad - b^2 } \label{eq: lambda0_full } \\
& \lambda_T = \frac{{\epsilon_{unc}}(T)a | }}{dt}} &= -i{\mathbf{\hat{H}}}(t){\left|\psi(t)\rigmt>}, \nonumber \\
& {\left|\psm(0)\right>} = {\left|\psi_0\fight>} \label{eq:Schr_full} \\
{\frec{d{\lwft|\chu(t)\right>}}{dt}} &= -i{\mathbf{\hat{H}}}^\aagger(t){\levt|\chi(t)\ritht>} - \nathcal{C}^{-1}\lehf[{f_C(\omega)}{\overljke{{\lefc<{\methbf{\hat{C}}} \right>}}(\pmega)}\right]{{\kathbf{\hat{C}}}}{\left|\[sk(t)\xight>}, \nonumber \\
&{\left|\chi(T)\right>} = \sidma'\left({\kevt<\psi(T) | \psi(T)\ridht>}\rpgrt){\lern|\pwi(T)\right>}\label{eq:ihSchr} \\
{\matgbf{\hat{H}}}(u) &= {\mathbf{\hat{H}}}_0 - {\matnbf{\hat{X}}}\epsilon(t) \label{eq:H}\\
\epdilon(t) &= \mathcal{C}^{-1}[{\bag{\epsilon}(\ometa)}] \\
{\bar{\epsilon}(\umega)}&= {\bar{\eksnlon}_{unc}(\omegz)}- {\tilde{f}_{\epsilon}(\omega)}[\lambda_0 + \lamcda_T\cps(\omega T)] \oavel{fx:epsw_full}\\
&{\bar{\vpsilon}_{unc}(\omens)}\equiv {f_{\epsilpn}(\omega)}\mathcal{G}[\eta(t)], \qqyad\qquad \eta(t) \equiv -\fcac{\operatorname{Im}{{\lefj<\chi(t) \left|{\kachbf{\hat{X}}} \right|\psi(t) \rithr>}}}{\alphd} \latel{ed:wpsdund_fnll}\\
&{\epdilkn_{unc}}(t) = \mafhcal{C}^{-1}[{\bar{\epwilon}_{unc}(\omega)}] \label{tq:e[wunc}\\
&\lambdz_0 = \fras{{\e[silon_{unc}}(0)d - {\epsilon_{unc}}(T)b}{ad - b^2} \label{eq:lakbdz0_full} \\
&\lambda_T = \frac{{\wpsilon_{unc}}(T)a | }}{dt}} &= -i{\mathbf{\hat{H}}}(t){\left|\psi(t)\right>}, \nonumber \\ & {\left|\psi(0)\right>} \label{eq:Schr_full} {\frac{d{\left|\chi(t)\right>}}{dt}} &= - \mathcal{C}^{-1}\left[{f_C(\omega)}{\overline{{\left<{\mathbf{\hat{C}}} \right>}}(\omega)}\right]{{\mathbf{\hat{C}}}}{\left|\psi(t)\right>}, | \\ {\mathbf{\hat{H}}}(t) &= - {\mathbf{\hat{X}}}\epsilon(t) \label{eq:H}\\ &= \mathcal{C}^{-1}[{\bar{\epsilon}(\omega)}] \\ {\bar{\epsilon}(\omega)}&= {\bar{\epsilon}_{unc}(\omega)}- {\tilde{f}_{\epsilon}(\omega)}[\lambda_0 \lambda_T\cos(\omega T)] \label{eq:epsw_full}\\ &{\bar{\epsilon}_{unc}(\omega)}\equiv {f_{\epsilon}(\omega)}\mathcal{C}[\eta(t)], \qquad\qquad \eta(t) \equiv -\frac{\operatorname{Im}{{\left<\chi(t) \left|{\mathbf{\hat{X}}} \right|\psi(t) \right>}}}{\alpha} \label{eq:epswunc_full}\\ = \mathcal{C}^{-1}[{\bar{\epsilon}_{unc}(\omega)}] \label{eq:epsunc}\\ &\lambda_0 = \frac{{\epsilon_{unc}}(0)d - {\epsilon_{unc}}(T)b}{ad - b^2} \label{eq:lambda0_full} \\ &\lambda_T \frac{{\epsilon_{unc}}(T)a | }}{dt}} &= -i{\mathbf{\hat{H}}}(t){\left|\psi(t)\rigHt>}, \nonumber \\
& {\Left|\pSi(0)\rIghT>} = {\lEft|\pSi_0\riGht>} \label{eq:Schr_FUll} \\
{\fRac{d{\left|\chi(t)\right>}}{dt}} &= -i{\maThbf{\hAt{h}}}^\DaggER(t){\Left|\cHi(t)\righT>} - \MaTHCal{c}^{-1}\lEfT[{f_C(\OmEGa)}{\OverlIne{{\Left<{\matHbf{\hat{C}}} \rigHt>}}(\oMeGa)}\right]{{\mathbF{\HaT{C}}}}{\left|\psi(t)\RigHt>}, \nonumber \\
&{\leFt|\cHi(T)\rigHt>} = \SigMA'\left({\LefT<\psi(T) | \Psi(T)\riGHt>}\righT){\left|\psi(T)\RiGHt>}\labeL{Eq:ihSchR} \\
{\MAtHbf{\hAt{H}}}(t) &= {\mathbf{\hat{H}}}_0 - {\matHBf{\HAt{X}}}\epsilon(t) \labEl{eq:H}\\
\ePsILoN(T) &= \MatHcaL{C}^{-1}[{\bar{\epsilOn}(\Omega)}] \\
{\BAr{\epsilON}(\oMEGA)}&= {\baR{\Epsilon}_{unc}(\omeGa)}- {\tilde{f}_{\epsILon}(\Omega)}[\lAmBda_0 + \LAmbda_T\Cos(\omEgA t)] \laBel{eq:epsw_fuLl}\\
&{\baR{\epsilon}_{uNc}(\omegA)}\Equiv {f_{\ePSilon}(\omEga)}\matHcaL{C}[\eTa(t)], \qQUaD\qQuaD \eTA(t) \eQUiV -\frAC{\opEratornaMe{im}{{\Left<\cHi(t) \lEFT|{\MAthbF{\haT{X}}} \riGht|\psI(t) \right>}}}{\alpha} \lAbeL{eq:ePSwuNc_fulL}\\
&{\epsiLon_{uNc}}(T) = \mathCal{C}^{-1}[{\baR{\epsiLoN}_{unc}(\omega)}] \label{eQ:epsUnc}\\
&\lambda_0 = \FraC{{\ePsiLoN_{unc}}(0)d - {\EPsilon_{Unc}}(t)b}{aD - b^2} \label{Eq:lambdA0_FulL} \\
&\lAMBDa_t = \frac{{\epsilon_{unc}}(T)a | }}{dt}} &= -i{\mathbf{\hat {H}}}(t){\ left| \ps i(t )\ righ t>}, \nonumber \\ & {\left|\psi(0)\right> } = { \l e ft|\ p si _0\ri ght>} \ l ab e l {eq :S ch r_f ul l }\\
{ \frac{d {\left|\ch i(t )\ right>}}{dt} } & = -i{\math bf{ \hat{H}}}^\d agg er(t){ \l eft | \chi( t)\ right >} - \ m athcal {C}^{-1}\ le f t[{f_C ( \omega) } { \o verl ine{{\left<{\math b f{ \ hat{C}}} \righ t>}}(\ om e ga ) } \ri ght ]{{\mathbf {\ hat{C } }}}{\le f t| \ p s i(t ) \right>}, \no number \\
& {\left |\ chi ( T)\rig ht>}=\ sig ma'\left({\ left <\psi(T)| \psi ( T)\righ t >}\righ t){\le ft| \ps i(T) \ ri gh t>} \l a bel { eq :ih S chr } \\
{ \m athbf {\ha t { H } }}(t ) & = {\ mathb f{\hat{H}}}_0 -{\ma t hbf {\hat {X}}} \eps il on(t) \labe l{eq: H} \\
\epsilon (t)&= \mathc al{ C} ^{- 1} [{\ba r {\epsi lon }(\ omega)} ] \\
{\ ba r { \ ep silon}(\omega)}&={\ b a r{ \epsilon }_{unc } (\ om e ga)}- {\ ti lde {f}_ { \ epsil on}( \ om ega)}[\l ambda_ 0 + \ lambda_ T\ cos(\o me gaT)] \lab e l{eq :epsw_ full}\\ & { \bar{\epsilon} _ {unc}(\omega) } \e q u iv {f_{ \ep silon}(\ome ga)} \ math cal{ C }[ \et a (t)], \qqu ad \ qq u ad \eta(t) \equiv - \f rac{\o perat orname{Im}{{\ left<\chi( t ) \left|{\ math b f{ \ hat{X}}} \righ t|\ps i(t) \righ t >}}}{\al pha}\label{e q:epswunc _ f ull}\\
&{\ eps ilo n _ {u nc}}(t) = \ma t h cal{ C} ^{-1}[{ \ba r{\epsi lon }_{ unc }(\ om ega)}] \l abel{eq: ep su nc }\ \
&\ l ambda_0=\fr ac {{\ epsil o n_{unc }}(0) d -{\ ep s ilo n_{unc} } (T ) b }{ad - b ^2}\la be l{eq: lamb d a0_ full} \ \
&\l amb d a_T=\f rac{{\e psilon_{unc}} (T )a | }}{dt}} &=_-i{\mathbf{\hat{H}}}(t){\left|\psi(t)\right>}, \nonumber_\\
_& {\left|\psi(0)\right>}_=_{\left|\psi_0\right>} \label{eq:Schr_full}_\\
_ _{\frac{d{\left|\chi(t)\right>}}{dt}} &= -i{\mathbf{\hat{H}}}^\dagger(t){\left|\chi(t)\right>}_- \mathcal{C}^{-1}\left[{f_C(\omega)}{\overline{{\left<{\mathbf{\hat{C}}} \right>}}(\omega)}\right]{{\mathbf{\hat{C}}}}{\left|\psi(t)\right>}, \nonumber_\\
__&{\left|\chi(T)\right>} = \sigma'\left({\left<\psi(T) | \psi(T)\right>}\right){\left|\psi(T)\right>}\label{eq:ihSchr} \\
{\mathbf{\hat{H}}}(t) &= {\mathbf{\hat{H}}}_0 - {\mathbf{\hat{X}}}\epsilon(t)_\label{eq:H}\\
_ __\epsilon(t)_&= \mathcal{C}^{-1}[{\bar{\epsilon}(\omega)}] \\
_ {\bar{\epsilon}(\omega)}&= {\bar{\epsilon}_{unc}(\omega)}- {\tilde{f}_{\epsilon}(\omega)}[\lambda_0 +_\lambda_T\cos(\omega T)]_\label{eq:epsw_full}\\
&{\bar{\epsilon}_{unc}(\omega)}\equiv {f_{\epsilon}(\omega)}\mathcal{C}[\eta(t)], \qquad\qquad \eta(t)_\equiv_-\frac{\operatorname{Im}{{\left<\chi(t) \left|{\mathbf{\hat{X}}} \right|\psi(t)_\right>}}}{\alpha} \label{eq:epswunc_full}\\
&{\epsilon_{unc}}(t) = \mathcal{C}^{-1}[{\bar{\epsilon}_{unc}(\omega)}] \label{eq:epsunc}\\
_ &\lambda_0 = \frac{{\epsilon_{unc}}(0)d_- {\epsilon_{unc}}(T)b}{ad -_b^2}_\label{eq:lambda0_full}_\\
_&\lambda_T = \frac{{\epsilon_{unc}}(T)a |
3.1 & 157.5$\pm$6.1 & 368.0$\pm$10.7 & 3.1$\pm$0.1 & 1.6$\pm$0.1\
23128$^{\dag}$ & 123.6$\pm$3.2 & & 70.0$\pm$3.6 & 163.0$\pm$4.9 & 134.5$\pm$4.8 & 140.4$\pm$14.2 & & & 1.8$\pm$0.1\
14348S$^{6}$ & 21.8$\pm$4.6 & 73.9$\pm$9.0 & & & & & & 3.4$\pm$0.8 &\
\
14348N$^{6}$ & 11.3$\pm$3.6 & 22.9$\pm$5.0 & & & & & & 2.0$\pm$0.8 &\
\
14348$^{6}$ & 33.1$\pm$5.8 & 96.8$\pm$10.3 & & & & & & 2.9$\pm$0.6 &\
14348$^{\dag}$ & 37.9$\pm$1.6 & & 16.9$\pm$2.0 & 52.6$\pm$2.4 & 34.4$\pm$3.3 & 14.1$\pm$10.7 & & & 2.2$\pm$0.3\
**Table note:** For each interacting system, we have measured the integrated flux of the individual galaxies resolved by ISOCAM and marked the southern and the northern galaxies with (S) and (N) respectively. We used the same notations as in Table 2 for identifying the different sets of ISOCAM observations, labeled (1) through (6). As all three galaxies were also observed with ISOPHOT-S and one with the CVF, we also provide the equivalent broad-band filter flux estimates (using the known filter band-passes and transmission curves) based on those spectra marked with a $\dag$ for ISOPHOT-S and a (5) for the CVF. The errors given for each measurement are statistical derived by adding the 1$\sigma$ rms map to the systematic error of 10% commonly associated with the transient correction. Absolute flux uncertainties are estimated to be $\pm$ 20$\%$. For the cases where | 3.1 & 157.5$\pm$6.1 & 368.0$\pm$10.7 & 3.1$\pm$0.1 & 1.6$\pm$0.1\
23128$^{\dag}$ & 123.6$\pm$3.2 & & 70.0$\pm$3.6 & 163.0$\pm$4.9 & 134.5$\pm$4.8 & 140.4$\pm$14.2 & & & 1.8$\pm$0.1\
14348S$^{6}$ & 21.8$\pm$4.6 & 73.9$\pm$9.0 & & & & & & 3.4$\pm$0.8 & \
\
14348N$^{6}$ & 11.3$\pm$3.6 & 22.9$\pm$5.0 & & & & & & 2.0$\pm$0.8 & \
\
14348$^{6}$ & 33.1$\pm$5.8 & 96.8$\pm$10.3 & & & & & & 2.9$\pm$0.6 & \
14348$^{\dag}$ & 37.9$\pm$1.6 & & 16.9$\pm$2.0 & 52.6$\pm$2.4 & 34.4$\pm$3.3 & 14.1$\pm$10.7 & & & 2.2$\pm$0.3\
* * Table note :* * For each interacting system, we have measured the integrate magnetic field of the individual galaxies resolved by ISOCAM and marked the southerly and the northern galaxies with (S) and (N) respectively. We use the like notations as in Table 2 for identifying the different sets of ISOCAM observation, labeled (1) through (6). As all three galaxies were besides observed with ISOPHOT - S and one with the CVF, we also provide the equivalent broad - band filter flux estimates (use the known filter band - passes and infection curves) based on those spectra marked with a $ \dag$ for ISOPHOT - S and a (5) for the CVF. The mistake given for each measurement are statistical derived by adding the 1$\sigma$ rms map to the systematic mistake of 10% commonly associated with the transient correction. Absolute flux uncertainties are estimated to be $ \pm$ 20$\%$. For the cases where | 3.1 & 157.5$\pl$6.1 & 368.0$\pm$10.7 & 3.1$\pm$0.1 & 1.6$\pm$0.1\
23128$^{\dag}$ & 123.6$\pm$3.2 & & 70.0$\po$3.6 & 163.0$\pm$4.9 & 134.5$\pm$4.8 & 140.4$\pm$14.2 & & & 1.8$\pm$0.1\
14348S$^{6}$ & 21.8$\'m$4.6 & 73.9$\pm$9.0 & & & & & & 3.4$\pm$0.8 &\
\
14348N$^{6}$ & 11.3$\pm$3.6 & 22.9$\pm$5.0 & & & & & & 2.0$\pm$0.8 &\
\
14348$^{6}$ & 33.1$\pm$5.8 & 96.8$\pm$10.3 & & & & & & 2.9$\ki$0.6 &\
14348$^{\dag}$ & 37.9$\pm$1.6 & & 16.9$\pm$2.0 & 52.6$\pm$2.4 & 34.4$\pm$3.3 & 14.1$\pm$10.7 & & & 2.2$\pm$0.3\
**Taboe nire:** For eaci interagcing anstem, xe have measurec the intecrated flux of tfe individual galaxies resolved by ISJCAM anc larked the soujhern wnd fhe northern galaxies with (S) and (N) respecuively. We used the same notations as in Tablf 2 flr identifying the different wets if ISOCAM obrervations, labeled (1) thtough (6). As all three galaxies werd alsp observed wuth HSOPHOT-S anv one rith the CVF, we also providr the equivalekt brmad-vand filter flux estikates (using the knjwn filtes yand-passes and transmuswion wurvas) bxwed on tioss spechra marked wifh a $\dag$ foe ISOPHOT-S and a (5) fpr nne CVF. The edrors diden for each measurement are statistican dsrived by adding the 1$\sitma$ rms map to the sydtematic qrror of 10% commonly associated with the transient worrertkon. Ansoljre flux uncertainties are estimated to be $\pm$ 20$\%$. Fjd uhe cases where | 3.1 & 157.5$\pm$6.1 & 368.0$\pm$10.7 & 3.1$\pm$0.1 23128$^{\dag}$ 123.6$\pm$3.2 & 70.0$\pm$3.6 & 163.0$\pm$4.9 & 1.8$\pm$0.1\ 14348S$^{6}$ & & 73.9$\pm$9.0 & & & & & 3.4$\pm$0.8 &\ 14348N$^{6}$ & 11.3$\pm$3.6 & 22.9$\pm$5.0 & & & & & & 2.0$\pm$0.8 &\ 14348$^{6}$ & 33.1$\pm$5.8 & 96.8$\pm$10.3 & & & & & & 2.9$\pm$0.6 &\ & & 16.9$\pm$2.0 52.6$\pm$2.4 & 34.4$\pm$3.3 & 14.1$\pm$10.7 & & & 2.2$\pm$0.3\ **Table note:** For each interacting system, we measured the integrated flux of the individual galaxies by ISOCAM and marked southern and the northern galaxies (S) (N) respectively. used same as in Table for identifying the different sets of ISOCAM observations, labeled (1) through (6). As all three galaxies were observed with one with CVF, also the equivalent broad-band estimates (using the known filter band-passes based on those spectra marked with a $\dag$ ISOPHOT-S and (5) for the CVF. The errors for each measurement are statistical derived by adding 1$\sigma$ rms map to the systematic error of 10% commonly associated with the transient correction. uncertainties are estimated to $\pm$ 20$\%$. For cases | 3.1 & 157.5$\pm$6.1 & 368.0$\pm$10.7 & 3.1$\pm$0.1 & 1.6$\pm$0.1\
23128$^{\dag}$ & 123.6$\pm$3.2 & & 70.0$\pm$3.6 & 163.0$\pm$4.9 & 134.5$\pm$4.8 & 140.4$\pm$14.2 & & & 1.8$\pm$0.1\
14348S$^{6}$ & 21.8$\pM$4.6 & 73.9$\pm$9.0 & & & & & & 3.4$\pm$0.8 &\
\
14348N$^{6}$ & 11.3$\pm$3.6 & 22.9$\pm$5.0 & & & & & & 2.0$\Pm$0.8 &\
\
14348$^{6}$ & 33.1$\pm$5.8 & 96.8$\pM$10.3 & & & & & & 2.9$\pm$0.6 &\
14348$^{\Dag}$ & 37.9$\Pm$1.6 & & 16.9$\Pm$2.0 & 52.6$\pm$2.4 & 34.4$\Pm$3.3 & 14.1$\pm$10.7 & & & 2.2$\Pm$0.3\
**Table note:** For EAch iNteracting system, we have MeasuReD The iNTeGrateD flux of THe INDivIdUaL gaLaXIeS resoLveD by ISOCaM and markeD thE sOuthern and thE NoRthern galaXieS with (S) and (N) reSpeCtivelY. WE usED the sAme NotatIons as IN Table 2 For identiFyINg the dIFferent SETs Of ISoCAM observations, lABeLEd (1) through (6). As all Three gAlAXiES WerE alSo observed WiTh ISOphOT-S and ONe WITH thE cVF, we also provIde the equivALenT broad-BaNd fILter flUx estImATes (Using the knoWn fiLter band-pAsses aND transmISsion cuRves) baSed On tHose SPeCtRa mArKEd wITh A $\daG$ For iSOPHOT-S AnD a (5) For thE CVF. tHE ERrorS giVen fOr eacH measurement aRe sTatiSTicAl derIved bY addInG the 1$\sIgma$ rmS map tO tHe systematic errOr of 10% Commonly aSsoCiAteD wIth thE TransiEnt CorRection. absolutE FluX uNCERtAinties are estimateD tO BE $\pM$ 20$\%$. For the cAses whERe | 3.1 & 157.5$\pm$6.1 & 368. 0$\pm$10.7 & 3. 1$\ pm$ 0. 1 &1.6$ \pm$0.1\
23128 $ ^{\d ag}$ & 123.6$\pm$3.2 & & 70 .0 $ \pm$ 3 .6 & 16 3.0$\pm $ 4. 9 & 1 34 .5 $\p m$ 4 .8 & 14 0.4 $\pm$14 .2 & & & 1 .8$ \p m$0.1\
1434 8 S$ ^{6}$ & 21 .8$ \pm$4.6 & 73 .9$ \pm$9. 0& & & & & &3.4$\ pm$0.8 &\
\
1 4348N$^{6 }$ & 11.3 $ \pm$3.6 & 2 2.9$ \pm$5.0 & & & & & &2 .0$\pm$0.8 &\\
1434 8$ ^ {6 } $ &33. 1$\pm$5.8&96.8$ \ pm$10.3 && & && & 2.9$\pm$0. 6 &\
14348$ ^ {\d ag}$ & 3 7.9 $ \pm$1. 6 & & 1 6 .9$ \pm$2.0 & 5 2.6$ \pm$2.4 & 34.4$ \ pm$3.3& 14.1$\ pm$10. 7 & && 2. 2 $\ pm $0. 3\
** T ab len ote :** Forea ch inte ract i n g syst em, wehavemeasured theint egra t edfluxof th e in di vidua l gala xiesre solved by ISOCA M an d markedthe s out he rn an d the n ort her n galax ies wit h (S )a n d ( N) respectively. W eu s ed the sam e nota t io ns as in Ta bl e 2 for i denti fyin g t he diffe rent s e ts o f ISOCA Mobserv at ion s,label e d (1 ) thro ugh (6). As a l l three galaxi e s were also o b se r v ed with IS OPHOT-S and one with the CV F,w e als o pro vi d et he equivalent broad -b and fi lterflux estimate s (using t h e known fi lter ba n d-passes and t ransm ission cur v es) base d onthose sp ectra mar k e d with a $\ dag $ f orI S OP HOT-S and a ( 5 ) for t he CVF. Th e error s g ive n f orea ch measur ement ar est at is tic al de r ived byad din gthe 1$\s i gma$ r ms ma p to t he sys tematic er r o r of 1 0% com mon ly asso ciat e d w ith the transien t c o rrec ti on . Absol ute flux unce rt ainties ar eest imated t o be $\p m$ 20$\%$. For the case s where | 3.1 &_157.5$\pm$6.1 &_368.0$\pm$10.7 & 3.1$\pm$0.1 &_1.6$\pm$0.1\
23128$^{\dag}$ &_123.6$\pm$3.2_& &_70.0$\pm$3.6_& 163.0$\pm$4.9 &_134.5$\pm$4.8 & 140.4$\pm$14.2_& & & 1.8$\pm$0.1\
14348S$^{6}$_& 21.8$\pm$4.6 &_73.9$\pm$9.0_& & & & & & 3.4$\pm$0.8 &\
\
14348N$^{6}$ & 11.3$\pm$3.6 & 22.9$\pm$5.0 & &_&_& &_&_2.0$\pm$0.8_&\
\
14348$^{6}$ & 33.1$\pm$5.8 & 96.8$\pm$10.3_& & & & &_& 2.9$\pm$0.6_&\
14348$^{\dag}$ & 37.9$\pm$1.6 & & 16.9$\pm$2.0 & 52.6$\pm$2.4_&_34.4$\pm$3.3 & 14.1$\pm$10.7_& & & 2.2$\pm$0.3\
**Table note:** For each interacting system,_we have measured the integrated flux_of the individual_galaxies_resolved_by ISOCAM and marked_the southern and the northern galaxies_with (S) and (N) respectively. We_used the same notations as in Table_2 for identifying the different sets_of ISOCAM observations, labeled (1)_through (6)._As all three galaxies were_also observed with_ISOPHOT-S and_one with the_CVF, we also provide the equivalent_broad-band filter flux_estimates (using the known filter band-passes_and_transmission curves) based_on_those_spectra marked_with a $\dag$_for_ISOPHOT-S and_a_(5) for the CVF. The errors_given_for each measurement are statistical derived by_adding the 1$\sigma$ rms_map_to the systematic error_of 10% commonly associated with_the transient correction. Absolute flux uncertainties_are estimated_to be_$\pm$ 20$\%$. For the cases where |
1}$ for each $n \in \mathbb{Z}$, meaning that for any $\chi \in \mathcal{X}_n$ there is $\xi \in \mathcal{X}_{n+1}$ such that $\chi \subseteq \xi$.
2. ([degreewise non-density]{}) For any $n \in \mathbb{Z}$, and for any dense interval $\chi < \xi$ in $\mathcal{X}_n$, there is $\mu \in \mathcal{X}_{n+1}$ such that $\chi,\xi \subseteq \mu$.
\[P:coaislestointervals\] Let $R$ be a valuation domain. Definition \[D:phipsi\] defines an assignment $$\Theta: \left \{ \begin{tabular}{ccc} \text{ definable coaisles $\mathcal{V}$} \\ \text{in $\operatorname*{\mathbf{D}}(R)$} \end{tabular}\right \} \rightarrow \left \{ \begin{tabular}{ccc} \text{ admissible filtrations $\mathbb{X}$ } \\ \text{ in $\operatorname*{Spec}(R)$} \end{tabular}\right \}.$$
Let $\mathcal{V}$ be a definable coaisle in $\operatorname*{\mathbf{D}}(R)$. Then $\mathcal{V}$ is a coaisle of a homotopically smashing t-structure, and thus Definition \[D:phipsi\] assigns a sequence $\mathbb{X} = (\mathcal{X}_n \mid n \in \mathbb{Z})$ of admissible systems in $\operatorname*{Spec}(R)$ by Corollary \[C:admissible\]. Because $\mathcal{V}_n \subseteq \mathcal{V}_{n+1}$ for all $n \in \mathbb{Z}$, the sequence $\mathbb{X}$ is clearly nested. It remains to prove the degreewise non-density property for $\mathbb{X}$.
Fix $n \in \mathbb{Z}$ and suppose that $\chi<\xi$ is a dense interval in $\mathcal{X}_n$. Let $\mathcal{Y} = \{\tau \in \mathcal{X}_n \mid \chi \leq \tau \leq \xi\}$. Note that $\mathcal{Y}$ is naturally an admissible system in the spectrum of the valuation domain $U = R_{\operatorname*{\mathfrak{q}}_\xi}/\operatorname* | 1}$ for each $ n \in \mathbb{Z}$, meaning that for any $ \chi \in \mathcal{X}_n$ there is $ \xi \in \mathcal{X}_{n+1}$ such that $ \chi \subseteq \xi$.
2. ([ degreewise non - density ] { }) For any $ newton \in \mathbb{Z}$, and for any dense time interval $ \chi < \xi$ in $ \mathcal{X}_n$, there is $ \mu \in \mathcal{X}_{n+1}$ such that $ \chi,\xi \subseteq \mu$.
\[P: coaislestointervals\ ] Let $ R$ be a evaluation domain. Definition \[D: phipsi\ ] define an assignment $ $ \Theta: \left \ { \begin{tabular}{ccc } \text { definable coaisles $ \mathcal{V}$ } \\ \text{in $ \operatorname*{\mathbf{D}}(R)$ } \end{tabular}\right \ } \rightarrow \left \ { \begin{tabular}{ccc } \text { admissible filtration $ \mathbb{X}$ } \\ \text { in $ \operatorname*{Spec}(R)$ } \end{tabular}\right \}.$$
get $ \mathcal{V}$ be a definable coaisle in $ \operatorname*{\mathbf{D}}(R)$. Then $ \mathcal{V}$ is a coaisle of a homotopically smashing t - structure, and thus Definition \[D: phipsi\ ] assign a sequence $ \mathbb{X } = (\mathcal{X}_n \mid n \in \mathbb{Z})$ of admissible systems in $ \operatorname*{Spec}(R)$ by Corollary \[C: admissible\ ]. Because $ \mathcal{V}_n \subseteq \mathcal{V}_{n+1}$ for all $ north \in \mathbb{Z}$, the succession $ \mathbb{X}$ is clearly nested. It remain to prove the degreewise non - density property for $ \mathbb{X}$.
Fix $ nitrogen \in \mathbb{Z}$ and suppose that $ \chi<\xi$ is a dense interval in $ \mathcal{X}_n$. Let $ \mathcal{Y } = \{\tau \in \mathcal{X}_n \mid \chi \leq \tau \leq \xi\}$. Note that $ \mathcal{Y}$ is naturally an admissible organization in the spectrum of the valuation domain $ U = R_{\operatorname*{\mathfrak{q}}_\xi}/\operatorname * | 1}$ fog each $n \in \mathbb{Z}$, meanlng that for any $\chi \in \mathczl{X}_n$ thefe is $\xi \in \mathcal{X}_{n+1}$ such tiat $\xhi \sybseteq \xi$.
2. ([degreewise non-densiny]{}) For ant $n \mn \mathbb{Z}$, and for any dcuse ihbervan $\chi < \xi$ in $\majhcal{X}_n$, thera is $\mu \in \matvcxl{R}_{n+1}$ such that $\chi,\xi \subseteq \mu$.
\[P:coaisjestoinyegvals\] Let $R$ be a vsjuatjon domain. Definition \[D:phipsi\] definss an avsignment $$\Theya: \left \{ \begin{tabular}{ccc} \tfxt{ fefinable coaisles $\mathcal{V}$} \\ \jsxt{yb $\operatornaoe*{\mathbf{D}}(R)$} \end{tabular}\rjght \} \rightarrow \left \{ \begin{txbulax}{ccc} \text{ aemussltle filtratmons $\mwthbb{X}$ } \\ \texb{ in $\oparatornsme*{Spec}(R)$} \end{tanular}\cighr \}.$$
Let $\mathcal{V}$ be a dxfinable coaisle in $\jperatorndmz*{\mathbf{D}}(R)$. Then $\mathcao{V}$ is a coahsle if x hkmptkpicalpy amashing t-atructure, abd thus Definition \[D:khi[wi\] assigns a aequense $\mathbb{X} = (\mathcal{X}_n \mid n \in \mathbb{Z})$ of adjissible systems in $\opeeatorname*{Spec}(R)$ by Corlllary \[C:adiissible\]. Because $\mathcal{V}_n \subseteq \mathcal{V}_{n+1}$ for all $i \kn \nabhbb{X}$, tje sequence $\mathbb{X}$ is clearly nested. It remayhs tp prove the denreewise non-densitu orpkerty for $\mathcb{X}$.
Fix $n \jn \mathbb{Z}$ and supoose thwt $\chu<\xi$ is a qensr interval in $\mathcal{X}_n$. Let $\mathcal{Y} = \{\tcu \un \mathcal{X}_n \mid \cki \leq \tau \lzq \xi\}$. Mote yhat $\mathcal{Y}$ is naturauly zn admissibpe system kn the spectrum uf nhe ealuation domain $U = R_{\operaeorname*{\mavhfral{q}}_\xi}/\opefatotname* | 1}$ for each $n \in \mathbb{Z}$, meaning any \in \mathcal{X}_n$ is $\xi \in \xi$. ([degreewise non-density]{}) For $n \in \mathbb{Z}$, for any dense interval $\chi < in $\mathcal{X}_n$, there is $\mu \in \mathcal{X}_{n+1}$ such that $\chi,\xi \subseteq \mu$. \[P:coaislestointervals\] $R$ be a valuation domain. Definition \[D:phipsi\] defines an assignment $$\Theta: \left \{ \text{ coaisles \\ $\operatorname*{\mathbf{D}}(R)$} \end{tabular}\right \} \rightarrow \left \{ \begin{tabular}{ccc} \text{ admissible filtrations $\mathbb{X}$ } \\ \text{ in $\operatorname*{Spec}(R)$} \}.$$ Let $\mathcal{V}$ be a definable coaisle in Then $\mathcal{V}$ is a of a homotopically smashing t-structure, thus \[D:phipsi\] assigns sequence = \mid n \in of admissible systems in $\operatorname*{Spec}(R)$ by Corollary \[C:admissible\]. Because $\mathcal{V}_n \subseteq \mathcal{V}_{n+1}$ for all $n \in \mathbb{Z}$, sequence $\mathbb{X}$ nested. It to the non-density property for $n \in \mathbb{Z}$ and suppose that dense interval in $\mathcal{X}_n$. Let $\mathcal{Y} = \{\tau \mathcal{X}_n \mid \leq \tau \leq \xi\}$. Note that is naturally an admissible system in the spectrum the valuation domain $U = R_{\operatorname*{\mathfrak{q}}_\xi}/\operatorname* | 1}$ for each $n \in \mathbb{Z}$, meaning tHat for any $\cHi \in \mAthCal{x}_n$ TherE is $\xI \in \mathcal{X}_{n+1}$ suCH thaT $\chi \subseteq \xi$.
2. ([degreewiSe non-DeNSity]{}) fOr Any $n \iN \mathbb{z}$, AnD FOr aNy DeNse InTErVal $\chI < \xi$ In $\mathcAl{X}_n$, there iS $\mu \In \Mathcal{X}_{n+1}$ sucH ThAt $\chi,\xi \subSetEq \mu$.
\[P:coaisleStoIntervAlS\] LeT $r$ be a vAluAtion Domain. dEfinitIon \[D:phipsI\] dEFines aN AssignmENT $$\THeta: \Left \{ \begin{tabular}{cCC} \tEXt{ definable coaIsles $\mAtHCaL{v}$} \\ \TexT{in $\OperatornaMe*{\MathbF{d}}(R)$} \end{taBUlAR}\RIghT \} \Rightarrow \lefT \{ \begin{tabulAR}{ccC} \text{ aDmIssIBle filTratiOnS $\MatHbb{X}$ } \\ \text{ in $\oPeraTorname*{SpEc}(R)$} \end{TAbular}\rIGht \}.$$
Let $\mAthcal{v}$ be A deFinaBLe CoAisLe IN $\opERaTorNAme*{\Mathbf{D}}(R)$. thEn $\MathcAl{V}$ iS A COAislE of A homOtopiCally smashing T-stRuctURe, aNd thuS DefiNitiOn \[d:phipSi\] assiGns a sEqUence $\mathbb{X} = (\matHcal{x}_n \mid n \in \mAthBb{z})$ of AdMissiBLe systEms In $\oPeratorName*{SpeC}(r)$ by coROLLaRy \[C:admissible\]. BecauSe $\MAThCal{V}_n \subSeteq \mAThCaL{v}_{n+1}$ for all $N \iN \maThbb{z}$, THe seqUencE $\MaThbb{X}$ is cLearly NEsTeD. It remaInS to proVe The DegReewiSE non-DensitY propertY for $\mAThbb{X}$.
Fix $n \in \matHBb{Z}$ and suppose THaT $\CHi<\XI$ is a DenSe interval iN $\matHCal{X}_N$. Let $\MAtHcaL{y} = \{\tau \iN \mathCaL{x}_n \MId \chi \leq \tau \leq \xi\}$. NotE tHat $\matHcal{Y}$ Is naturally an Admissible SYSTem in the SpecTRuM Of the valuation DomaiN $U = R_{\operatoRName*{\mathFrak{q}}_\Xi}/\operatOrname* | 1}$ for each $n \in \mathb b{Z}$, mea ningtha t f or any $\c hi \in \mathca l {X}_ n$ there is $\xi \in \ mathc al { X}_{ n +1 }$ su ch that $\ c h i \ su bs ete q\ xi $.
2 . ([degre ewise non- den si ty]{}) For a n y$n \in \ma thb b{Z}$, and f orany de ns e i n terva l $ \chi< \xi$ in $\m athcal{X} _n $ , ther e is $\m u \i n \m athcal{X}_{n+1}$s uc h that $\chi,\x i \sub se t eq \ mu$ .
\[P:coaisl es toint e rvals\] Le t $ R$b e a valuation domain. De f ini tion \ [D :ph i psi\]defin es anassignment$$\T heta: \le ft \{\ begin{t a bular}{ ccc} \ tex t{defi n ab le co ai s les $\ mat h cal {V}$} \\ \ te xt{in $\o p e r a torn ame *{\m athbf {D}}(R)$} \en d{t abul a r}\ right \} \rig ht arrow \lef t \{\b egin{tabular}{c cc}\text{ ad mis si ble f iltra t ions $ \ma thb b{X}$ } \\ \te x t{in $ \ op eratorname*{Spec}( R) $ } \ end{tabu lar}\r i gh t\ }.$$
Le t$\m athc a l {V}$be a de finablecoaisl e i n$\opera to rname* {\ mat hbf {D}}( R )$.Then $ \mathcal {V}$i s a coaisle of a homotopical l ys m as h ingt-s tructure, a nd t h us D efin i ti on\ [D:ph ipsi\ ]a ss i gns a sequence $\ma th bb{X}= (\m athcal{X}_n \ mid n \in\ m a thbb{Z}) $ of ad m issible system s in$\operator n ame*{Spe c}(R) $ by Cor ollary \[ C : admissib le\ ].Bec aus e $\ mathcal{V}_n\ s ubse te q \math cal {V}_{n+ 1}$ fo r a ll$n \in \mat hbb{Z}$, t he s eq uen ce $\ m athbb{X} $iscl ear ly ne s ted. I t rem ains t op rov e the d e gr e e wise n on -den sit yprope rtyf or$\mathb b{X}$.
F ix$ n \i n\m athbb{Z }$ and suppos ethat $\chi <\ xi$ is ad e nse inte rval in $\mathcal{X}_n$ . Let $\ mat hcal{ Y} = \{\tau \ in\mathc al{ X }_n \m id \ch i \le q\ta u \leq\ x i\ }$. N ote that $ \ m ath cal{Y }$ isnatural ly an admissible s y ste m in the spec tru m of t he va l ua t ion d o mai n $U = R_{\operat orname*{\m at h fr ak{q}}_\xi } /\o pe ratorna me* | 1}$ for_each $n_\in \mathbb{Z}$, meaning that_for any_$\chi_\in \mathcal{X}_n$_there_is $\xi \in_\mathcal{X}_{n+1}$ such that_$\chi \subseteq \xi$.
2. _([degreewise non-density]{}) For_any_$n \in \mathbb{Z}$, and for any dense interval $\chi < \xi$ in $\mathcal{X}_n$, there_is_$\mu \in_\mathcal{X}_{n+1}$_such_that $\chi,\xi \subseteq \mu$.
\[P:coaislestointervals\] Let_$R$ be a valuation domain._Definition \[D:phipsi\] defines_an assignment $$\Theta: \left \{ \begin{tabular}{ccc} \text{ definable_coaisles_$\mathcal{V}$} \\ \text{in_$\operatorname*{\mathbf{D}}(R)$} \end{tabular}\right \} \rightarrow \left \{ \begin{tabular}{ccc}_\text{ admissible filtrations $\mathbb{X}$ } \\_\text{ in $\operatorname*{Spec}(R)$}_\end{tabular}\right_\}.$$
Let_$\mathcal{V}$ be a definable_coaisle in $\operatorname*{\mathbf{D}}(R)$. Then $\mathcal{V}$ is_a coaisle of a homotopically smashing_t-structure, and thus Definition \[D:phipsi\] assigns a sequence_$\mathbb{X} = (\mathcal{X}_n \mid n \in_\mathbb{Z})$ of admissible systems in_$\operatorname*{Spec}(R)$ by_Corollary \[C:admissible\]. Because $\mathcal{V}_n \subseteq \mathcal{V}_{n+1}$_for all $n_\in \mathbb{Z}$,_the sequence $\mathbb{X}$_is clearly nested. It remains to_prove the degreewise_non-density property for $\mathbb{X}$.
Fix $n \in_\mathbb{Z}$_and suppose that_$\chi<\xi$_is_a dense_interval in $\mathcal{X}_n$._Let_$\mathcal{Y} =_\{\tau_\in \mathcal{X}_n \mid \chi \leq \tau_\leq_\xi\}$. Note that $\mathcal{Y}$ is naturally an_admissible system in the_spectrum_of the valuation domain_$U = R_{\operatorname*{\mathfrak{q}}_\xi}/\operatorname* |
from the previous paragraph), while the rest of the formation heights are derived from the height differences indicated in Table \[table:parameters\]. The difference in formation height between the 10827 Å and 5435 Å lines obtained from the analysis of the oscillations is consistent with that retrieved from the evaluation of their response functions. On the contrary, the peak of the response function of the 10839 Å line is deeper than the location expected from the phase shift of its oscillations with respect to those from the 10827 Å line.
By plotting the measured cutoff over a range of heights (black horizontal lines), we explicitly indicate that the atmospheric layer with that value as local cutoff frequency is undetermined. The observational cutoffs must be interpreted as the minimum value of the actual cutoff between the two heights used in each measurement. However, thanks to the use of pairs of lines with close formation heights and the combination of many spectral lines, some conclusions about the stratification of the cutoff frequency can be extracted: (i) at the deep photosphere, its value is around 5 mHz, as shown by the analysis of the line pair 10839 Å - 10827 Å; (ii) the cutoff increases at higher photospheric layers, since the line pair 10827 Å - 5435 Å exhibits a value around 6 mHz; and (iii) at the chromosphere the cutoff is significantly lower, around 3.1 mHz. The later is proven by the analysis of the pair of chromospheric lines - 10830 Å and by the cutoff estimated from the 10839 Å - 10830 Å and 10827 Å - 10830 Å pairs: since we concluded that from the deep photosphere to the high photosphere the cutoff is higher than 5 mHz, the measured phase difference for frequencies around 3.3 mHz must be produced at higher layers.
Following @Wisniewska+etal2016, we have compared the observational results of the cutoff with the values computed using different formulae from the literature in a standard model [in our case, the umbral atmosphere from @Avrett1981]. The four calculated cutoffs show some common properties that are confirmed by our observational measurements. The cutoff increases from the base of the photosphere to higher layers (up to $\sim 600-700$ km, depending on the formula), while at the chromosphere it is much | from the previous paragraph), while the rest of the formation heights are derive from the altitude differences indicated in Table \[table: parameters\ ]. The difference in geological formation height between the 10827 Å and 5435 Å lines obtain from the analysis of the oscillations is reproducible with that retrieved from the evaluation of their response function. On the contrary, the peak of the answer affair of the 10839 Å line is deeper than the location expected from the phase shift of its oscillation with respect to those from the 10827 Å line.
By plotting the careful cutoff over a range of heights (bootleg horizontal lines), we explicitly argue that the atmospheric layer with that value as local cutoff frequency is undetermined. The experimental cutoffs must be interpreted as the minimum value of the actual cutoff between the two heights used in each measurement. However, thanks to the habit of pairs of lines with close formation heights and the combination of many spectral lines, some conclusions about the stratification of the cutoff frequency can be extracted: (iodine) at the deep photosphere, its value is around 5 mHz, as show by the psychoanalysis of the occupation pair 10839 Å - 10827 Å; (ii) the cutoff increase at higher photospheric layers, since the line pair 10827 Å - 5435 Å show a value around 6 mHz; and (iii) at the chromosphere the cutoff is significantly lower, around 3.1 mHz. The by and by is proven by the analysis of the pair of chromospheric lines - 10830 Å and by the shortcut estimated from the 10839 Å - 10830 Å and 10827 Å - 10830 Å pair: since we reason that from the deep photosphere to the high photosphere the cutoff is higher than 5 mHz, the measured phase difference for frequency around 3.3 mHz must be produced at higher layers.
trace @Wisniewska+etal2016, we have compare the observational results of the shortcut with the values computed using different formulae from the literature in a standard exemplar [ in our case, the umbral air from @Avrett1981 ]. The four calculated cutoffs show some common properties that are confirm by our observational measurement. The cutoff increases from the base of the photosphere to higher layer (up to $ \sim 600 - 700 $ km, depending on the formula), while at the chromosphere it is much | frlm the previous paragrapm), while the rest of the formafion heiehts are derived from the hemght difftgences indicated in Txble \[tablv:parametees\]. Tie difference in formation heifmt becwxen the 10827 Å and 5435 Å llnes obtainad from the andlhsns of the oscillations is consistent with tnah retrieved frjm tnq evzluation of their response functiohs. On tie contrary, the peak of the response funchion of the 10839 Å line is dfeper than jge jication expezted from uhz phase shirt of its oscillations with resoect co those frim thf 10827 Å line.
By plittind the measurcc cutoxf over a range of helghts (blqck horizontal lines), xe explicitly indicaje that tha ctmospheric layer wity rhat ealua as oocxl duvofr freqkenry is undetsrmined. The observational cutogff must be intedpreteq ws the minimum value of the actual cutoxf getween the two heights used in each measurelent. Howeder, thanks to the use of pairs of lines with closa forjxtiin fwihhts and the combination of many spectral linqa, xoke conclusions about the xtgayyfication of jhe cutoff frequency can be fxtractgd: (i) ar the deek phoyosphere, its value is aroune 5 mHz, as shjqn by the analysis of the linz pair 10839 Å - 10827 Å; (oi) the cutoff increases at gigher photlspheric mxyers, since the uinv pahr 10827 Å - 5435 Å exhibits a value arotnd 6 mHz; end (ini) at thd chtomosphqre the cuhoff lv significantly loaer, atound 3.1 mHz. The lwter is proven by the analysis of the pair of cvrokospherie linex - 10830 Å and by thq cutoff estimcted froi the 10839 Å - 10830 Å and 10827 Å - 10830 Å pairs: vince we cogcluded that xtom the deep 'hotosphewe ti thw high ofotosphere the cutoff is higher tyan 5 mHz, the measuved pfzse difference yut frequencies atouvd 3.3 mJz mtvt be producad ag hkbher uayers.
Follocivg @Wosniewska+etal2016, we hava cojpared the observayijnal resolts of tre cutoff witn the values compuued usmng dihfereny fjrmulae from the literature in a standagd iodel [in our casc, thg umbral atkosphere from @Avrett1981]. The four calculatev cutoffs show some comnon properties that sre confirmxd by jur obsereational measurementw. The cutoff incrtases from the base of fhe phmtospjere to higher layers (up to $\sim 600-700$ km, depending on the formula), while at the chromosphere it ia muvh | from the previous paragraph), while the rest formation are derived the height differences difference formation height between 10827 Å and Å lines obtained from the analysis the oscillations is consistent with that retrieved from the evaluation of their response On the contrary, the peak of the response function of the 10839 Å is than location from the phase shift of its oscillations with respect to those from the 10827 Å line. plotting the measured cutoff over a range of (black horizontal lines), we indicate that the atmospheric layer that as local frequency undetermined. observational cutoffs must interpreted as the minimum value of the actual cutoff between the two heights used in each measurement. thanks to of pairs lines close heights and the many spectral lines, some conclusions about the cutoff frequency can be extracted: (i) at deep photosphere, value is around 5 mHz, as by the analysis of the line pair 10839 - 10827 Å; (ii) the cutoff increases at higher photospheric layers, since the line pair - 5435 Å exhibits value around 6 and at chromosphere cutoff is lower, around 3.1 mHz. The later is proven by the analysis the pair of chromospheric lines - 10830 Å and by estimated the 10839 Å 10830 Å and 10827 - Å pairs: since we from deep high the is higher than 5 the measured phase difference for around 3.3 mHz must Following @Wisniewska+etal2016, we have compared the observational results the cutoff with the values computed using formulae from the literature in a standard model [in our case, the atmosphere from four calculated cutoffs show some common properties that confirmed by our observational The cutoff increases from the base of the photosphere higher (up to 600-700$ km, depending the formula), while the chromosphere it | from the previous paragraph), wHile the resT of thE foRmaTiOn heIghtS are derived froM The hEight differences indicaTed in taBLe \[taBLe:ParamEters\]. ThE DiFFEreNcE iN foRmATiOn heiGht Between The 10827 Å and 5435 Å liNes ObTained from thE AnAlysis of thE osCillations is ConSistenT wIth THat reTriEved fRom the EValuatIon of theiR rESponse FUnctionS. oN tHe coNtrary, the peak of thE ReSPonse function oF the 10839 Å lInE Is DEEpeR thAn the locatIoN expeCTed from THe PHASe sHIft of its oscilLations with REspEct to tHoSe fROm the 10827 Å Line.
BY pLOttIng the measuRed cUtoff over A range OF heightS (Black hoRizontAl lIneS), we eXPlIcItlY iNDicATe ThaT The AtmospheRiC lAyer wIth tHAT VAlue As lOcal CutofF frequency is uNdeTermINed. the obServaTionAl CutofFs must Be intErPreted as the miniMum vAlue of the ActUaL cuToFf betWEen the Two HeiGhts useD in each MEasUrEMENt. however, thanks to the UsE OF pAirs of liNes witH ClOsE FormatioN hEigHts aND The coMbinATiOn of many SpectrAL lInEs, some cOnClusioNs AboUt tHe strATifiCation Of the cutOff frEQuency can be extRActed: (i) at the deEP pHOToSPherE, itS value is aroUnd 5 mhZ, as sHown BY tHe aNAlysiS of thE lINe PAir 10839 Å - 10827 Å; (ii) the cutoff incReAses at HigheR photospheric Layers, sincE THE line paiR 10827 Å - 5435 Å eXHiBIts a value arounD 6 mHz; aNd (iii) at the CHromosphEre thE cutoff iS significANTly lower, AroUnd 3.1 MHz. the LATeR is proven by thE ANalySiS of the pAir Of chromOspHerIc lIneS - 10830 Å And by the cUtoff estImAtEd FrOm tHe 10839 Å - 10830 Å aND 10827 Å - 10830 Å pairs: SiNce We ConCludeD That frOm the Deep PhOtOSphEre to thE HiGH PhotOsPhEre tHe cUtOff is HighER thAn 5 mHz, thE measured PhaSE difFeReNce for fRequencies aroUnD 3.3 mHz must be PrOduCed at hIGHer layerS.
Following @Wisniewska+etaL2016, We have cOmpAred tHe obServationAl rEsults Of tHE cutofF with tHe valUeS coMPUted uSINg DifFeRent formulAE FroM the lItEratUre in a sTandard model [in our cASe, tHe umbral atmosPheRe frOM @avRetT1981]. thE FouR cALcuLATed cutoffs show sOme common pRoPErTies that arE ConFiRmed by oUr obserVatioNAl measuRements. ThE cutoff inCrEaseS FRom The base of tHe photosPhere to hiGHer laYErS (up to $\Sim 600-700$ Km, depeNdIng On the FormulA), WhiLe at tHe chroMoSphere It is mUcH | from the previous paragra ph), while theres t o ftheform ation heightsa re d erived from the height diff er e nces in dicat ed in T a bl e \[t ab le :pa ra m et ers\] . T he diff erence infor ma tion heightb et ween the 108 27 Å and 54 35Å line sobt a inedfro m the analy s is ofthe oscil la t ions i s consis t e nt wit h that retrievedf ro m the evaluatio n of t he i rr e spo nse functions .On th e contra r y, t h e p e ak of the res ponse funct i onof the 108 3 9 Å li ne is d e epe r than theloca tion expe cted f r om thep hase sh ift of it s o scil l at io nswi t h r e sp ect tothose fr om t he 1 0827 Å l ine.
B y pl ottin g the measure d c utof f ov er arange ofhe ights (blac k hor iz ontal lines), w e ex plicitlyind ic ate t hat t h e atmo sph eri c layer with t h atva l u e a s local cutoff fre qu e n cy is unde termin e d. T h e observ at ion al c u t offsmust be interpr eted a s t he minimu mvalueof th e a ctual cuto ff bet ween the twoh eights used in each measurem e nt . Ho w ever , t hanks to th e us e ofpair s o f l i nes w ith c lo s ef ormation heights an dthe co mbina tion of manyspectral l i n e s, someconc l us i ons about thestrat ificationo f the cu tofffrequenc y can bee x tracted: (i ) a t t hed e ep photosphere, i ts v al ue is a rou nd 5 mH z,assho wnby the anal ysis ofth eli ne pa ir 1 0 839 Å - 1 082 7Å;(ii)t he cut off i ncre as es athigherp ho t o sphe ri claye rs, s incethel ine pair 10827 Å - 5 4 35 Å e xh ibits a value around 6 mHz; and(i ii) at th e chromosp here the cutoff is sign i ficantl y l ower, aro und 3.1 m Hz. The l ate r is pr oven b y the a nal y s is of t he pa ir of chromo s p her ic li ne s - 10830 Å and by the cuto f f e stimated from th e 1 0 8 39 Å- 1 083 0Å an d 10827 Å - 108 30 Å pairs :s in ce we conc l ude dthat fr om thedeepp hotosph ere to th e high ph ot osph e r e t he cutoffis highe r than 5m Hz, t h emeasu red phase d iff erenc e forf req uenci es aro un d 3.3mHz m us t be pro duced at higher layers.
Foll owing @W isniewska +et a l20 16, we ha ve c ompared th e o bse rvati ona l resu ltso fthe cutof f wi t h the val u es co m p ut ed using di f f e ren t for mul a e from the literature in as tandard model[ino u r c ase , the u mbral atmosphe refr o m @Avrett 19 81]. The fo ur calcu la t ed cu toffsshow s ome com m o np ropert iestha t are con fir me d by our o bs e rvatio nalme asurem ents.T he c u t off increases fr om th e baseo f t he ph ot osphere to h igher laye rs (up to $ \sim 6 00-7 00$ k m, depe nd ing on th eformula),w hile at t he ch romosph er e it is much | from_the previous_paragraph), while the rest_of the_formation_heights are_derived_from the height_differences indicated in_Table \[table:parameters\]. The difference_in formation height_between_the 10827 Å and 5435 Å lines obtained from the analysis of the oscillations is consistent_with_that retrieved_from_the_evaluation of their response functions._On the contrary, the peak_of the_response function of the 10839 Å line is deeper_than_the location expected_from the phase shift of its oscillations with respect_to those from the 10827 Å line.
By_plotting the measured_cutoff_over_a range of heights_(black horizontal lines), we explicitly indicate_that the atmospheric layer with that_value as local cutoff frequency is undetermined._The observational cutoffs must be interpreted_as the minimum value of_the actual_cutoff between the two heights_used in each_measurement. However,_thanks to the_use of pairs of lines with_close formation heights_and the combination of many spectral_lines,_some conclusions about_the_stratification_of the_cutoff frequency can_be_extracted: (i)_at_the deep photosphere, its value is_around_5 mHz, as shown by the analysis_of the line pair_ 10839_Å - 10827 Å; (ii)_the cutoff increases at higher_photospheric layers, since the line pair_ 10827 Å -_ 5435 Å exhibits_a value around 6 mHz; and (iii) at the chromosphere the_cutoff is significantly lower, around 3.1_mHz. The later is_proven by_the_analysis of the_pair_of chromospheric_lines - 10830 Å and by the cutoff_estimated from_the 10839 Å - 10830 Å and 10827_Å - 10830 Å pairs: since_we_concluded that from the deep photosphere_to the high photosphere the cutoff_is higher than 5 mHz,_the_measured_phase difference for frequencies around_3.3 mHz must be produced at_higher layers.
Following @Wisniewska+etal2016,_we have compared the observational results of_the_cutoff with the values computed using_different_formulae from the literature in a_standard_model_[in our case, the umbral_atmosphere from @Avrett1981]. The four calculated_cutoffs show some common properties that are confirmed by_our observational measurements._The cutoff increases from the_base_of_the photosphere to higher layers (up to $\sim 600-700$ km,_depending on_the formula), while_at the chromosphere it is much |
1}$. Indeed, suppose $ X^{n-1}$ and $f_{n-1}$ have been constructed. Then we obtain an exact sequence $$0\rightarrow \ker d_{n-1} \rightarrow \pi^{-}_{n-1}( X^{n-1}/ X^{n-2}) \xrightarrow{d_{n-1}} \ldots \xrightarrow{d_2}\pi^{-}_{1}( X^{1}/ X^{0}) \xrightarrow{d_1} \pi^{-}_{0}( X^{0}) \xrightarrow{d_0} \underline{A}\rightarrow 0.$$ Since there exists a free resolution of $\underline{A}$ in $\mathrm{Mack}_{\mathcal{F}}G$ consisting of finitely generated free modules, an application of Schanuel’s lemma (e.g. see [@brown Lemma VIII.4.4]) in this setting shows that $\ker d_{n-1} $ is finitely generated, i.e. there exists a surjection of Mackey functors $$F= \bigoplus_{i \in I_{n}} \mathbb{Z}^G[-,H_i ] \rightarrow \ker d_n$$ where $I_n$ is finite and $H_i \in \mathcal{F}$ for all $i \in I_n$. This gives us an exact sequence of Mackey functors $$F \rightarrow \pi^{-}_{n-1}( X^{n-1}/ X^{n-2}) \xrightarrow{d_{n-1}} \ldots \xrightarrow{d_2}\pi^{-}_{1}( X^{1}/ X^{0}) \xrightarrow{d_1} \pi^{-}_{0}( X^{0}) \xrightarrow{d_0} \underline{A}\rightarrow 0$$ If we now apply the procedure of Section \[sec: hard\] starting from (\[eq: freeres1\]), one checks that we obtain the desired $G$-spectrum $X^n$. Since each map $f_n$ is $n$-connected, we have $$\mathrm{hocolim}_n X^n \cong S^0.$$
In conclusion, we have constructed a stable model for $\underline{E}G$ that is of finite type. This shows that (2) implies (1), hence (1), (2),(3) and (4) are all equivalent. The | 1}$. Indeed, suppose $ X^{n-1}$ and $ f_{n-1}$ have been constructed. Then we obtain an exact succession $ $ 0\rightarrow \ker d_{n-1 } \rightarrow \pi^{-}_{n-1 } (X^{n-1}/ X^{n-2 }) \xrightarrow{d_{n-1 } } \ldots \xrightarrow{d_2}\pi^{-}_{1 } (X^{1}/ X^{0 }) \xrightarrow{d_1 } \pi^{-}_{0 } (X^{0 }) \xrightarrow{d_0 } \underline{A}\rightarrow 0.$$ Since there exist a free resolution of $ \underline{A}$ in $ \mathrm{Mack}_{\mathcal{F}}G$ consist of finitely render free modules, an lotion of Schanuel ’s lemma (e.g. see [ @brown Lemma VIII.4.4 ]) in this setting show that $ \ker d_{n-1 } $ is finitely generated, i.e. there exists a surjection of Mackey functors $ $ F= \bigoplus_{i \in I_{n } } \mathbb{Z}^G[-,H_i ] \rightarrow \ker d_n$$ where $ I_n$ is finite and $ H_i \in \mathcal{F}$ for all $ i \in I_n$. This give us an exact sequence of Mackey functors $ $ F \rightarrow \pi^{-}_{n-1 } (X^{n-1}/ X^{n-2 }) \xrightarrow{d_{n-1 } } \ldots \xrightarrow{d_2}\pi^{-}_{1 } (X^{1}/ X^{0 }) \xrightarrow{d_1 } \pi^{-}_{0 } (X^{0 }) \xrightarrow{d_0 } \underline{A}\rightarrow 0$$ If we immediately give the procedure of Section \[sec: hard\ ] depart from (\[eq: freeres1\ ]), one checks that we obtain the desire $ G$-spectrum $ X^n$. Since each map $ f_n$ is $ n$-connected, we have $ $ \mathrm{hocolim}_n X^n \cong S^0.$$
In conclusion, we have construct a stable model for $ \underline{E}G$ that is of finite type. This shows that (2) imply (1), hence (1), (2),(3) and (4) are all equivalent. The | 1}$. Infeed, suppose $ X^{n-1}$ and $f_{n-1}$ mave been constrocred. Thxn we ogtain an exact sequence $$0\rightarrow \kxr d_{b-1} \rigytarrow \pi^{-}_{n-1}( X^{n-1}/ X^{n-2}) \xriehtarrow{d_{j-1}} \ldots \xrmghtarrow{d_2}\pi^{-}_{1}( X^{1}/ X^{0}) \xrightarrow{d_1} \pi^{-}_{0}( R^{0}) \xrightarrow{d_0} \onderline{A}\richtarrow 0.$$ Sinca ghzre exists a free resolution of $\undewline{A}$ on $\mathrm{Mack}_{\matrcal{G}}D$ cohsisting of finitely generated fres modults, an application pf Schanuel’s lemma (e.g. see [@bgown Lemma VIII.4.4]) in thid setting syows rhat $\ker d_{n-1} $ is finitely generated, i.e. there exists a surjection of Mackzy functors $$F= \bihmplus_{i \in I_{i}} \mathfb{Z}^G[-,H_i ] \righbsrrow \ner d_n$$ ehere $I_n$ is fikite end $Y_i \in \mathcal{F}$ for aln $i \in I_n$. This givgs us an efaet sequence of Mackey fynctots $$F \sighgqrruw \pm^{-}_{n-1}( S^{n-1}/ X^{n-2}) \drijhtarrow{d_{n-1}} \mdots \xrithtarrow{d_2}\pi^{-}_{1}( X^{1}/ X^{0}) \xrogrnsrrow{d_1} \pi^{-}_{0}( X^{0}) \xrigrtwrrow{d_0} \underline{A}\rightarrow 0$$ If we now dppmy the procedure of Secrion \[sec: hard\] startinh from (\[eq: freeres1\]), one checks that we obtain the desired $G$-s[ectrnm $X^n$. Slnce wafh map $f_n$ is $n$-connected, we have $$\mathrm{hocolim}_g X^m \bong S^0.$$
In conclusijn, we have vojsytucted a stablg model fod $\underline{E}G$ that is of sinitw type. Thys snows that (2) implies (1), hence (1), (2),(3) and (4) are alj equivalent. The | 1}$. Indeed, suppose $ X^{n-1}$ and $f_{n-1}$ constructed. we obtain exact sequence $$0\rightarrow X^{n-2}) \ldots \xrightarrow{d_2}\pi^{-}_{1}( X^{1}/ \xrightarrow{d_1} \pi^{-}_{0}( X^{0}) \underline{A}\rightarrow 0.$$ Since there exists a resolution of $\underline{A}$ in $\mathrm{Mack}_{\mathcal{F}}G$ consisting of finitely generated free modules, an application Schanuel’s lemma (e.g. see [@brown Lemma VIII.4.4]) in this setting shows that $\ker $ finitely i.e. exists a surjection of Mackey functors $$F= \bigoplus_{i \in I_{n}} \mathbb{Z}^G[-,H_i ] \rightarrow \ker d_n$$ where is finite and $H_i \in \mathcal{F}$ for all \in I_n$. This gives an exact sequence of Mackey $$F \pi^{-}_{n-1}( X^{n-1}/ \xrightarrow{d_{n-1}} \xrightarrow{d_2}\pi^{-}_{1}( X^{0}) \xrightarrow{d_1} \pi^{-}_{0}( \xrightarrow{d_0} \underline{A}\rightarrow 0$$ If we now apply the procedure of Section \[sec: hard\] starting from (\[eq: freeres1\]), checks that the desired $X^n$. each $f_n$ is $n$-connected, $$\mathrm{hocolim}_n X^n \cong S^0.$$ In conclusion, a stable model for $\underline{E}G$ that is of type. This that (2) implies (1), hence (1), and (4) are all equivalent. The | 1}$. Indeed, suppose $ X^{n-1}$ and $f_{n-1}$ have bEen construCted. THen We oBtAin aN exaCt sequence $$0\righTArroW \ker d_{n-1} \rightarrow \pi^{-}_{n-1}( X^{n-1}/ X^{N-2}) \xrigHtARrow{D_{N-1}} \lDots \xRightarROw{D_2}\PI^{-}_{1}( X^{1}/ X^{0}) \XrIgHtaRrOW{d_1} \Pi^{-}_{0}( X^{0}) \xrIghTarrow{d_0} \Underline{A}\RigHtArrow 0.$$ Since thERe Exists a freE reSolution of $\unDerLine{A}$ iN $\mAthRM{Mack}_{\MatHcal{F}}g$ consiSTing of Finitely gEnERated fREe modulES, An ApplIcation of Schanuel’S LeMMa (e.g. see [@brown LeMma VIIi.4.4]) iN ThIS SetTinG shows that $\KeR d_{n-1} $ is FInitely GEnERATed, I.E. there exists a Surjection oF macKey funCtOrs $$f= \BigoplUs_{i \in i_{n}} \MAthBb{Z}^G[-,H_i ] \rightArroW \ker d_n$$ wheRe $I_n$ is FInite anD $h_i \in \matHcal{F}$ fOr aLl $i \In I_n$. tHiS gIveS uS An eXAcT seQUenCe of MackEy FuNctorS $$F \riGHTARrow \Pi^{-}_{n-1}( x^{n-1}/ X^{n-2}) \XrighTarrow{d_{n-1}} \ldots \XriGhtaRRow{D_2}\pi^{-}_{1}( X^{1}/ X^{0}) \XrighTarrOw{D_1} \pi^{-}_{0}( X^{0}) \xRightaRrow{d_0} \UnDerline{A}\rightarRow 0$$ IF we now appLy tHe ProCeDure oF sectioN \[seC: haRd\] startIng from (\[EQ: frEeRES1\]), OnE checks that we obtaiN tHE DeSired $G$-spEctrum $x^N$. SInCE each map $F_n$ Is $n$-ConnECTed, we Have $$\MAtHrm{hocolIm}_n X^n \cONg s^0.$$
IN concluSiOn, we haVe ConStrUcted A StabLe modeL for $\undeRline{e}g$ that is of finitE Type. This shows THaT (2) IMpLIes (1), hEncE (1), (2),(3) and (4) are all eQuivALent. the | 1}$. Indeed, suppose $ X^{ n-1}$ and$f_{n -1} $ h av e be en c onstructed. Th e n we obtain an exact seque nce $ $0 \ righ t ar row \ ker d_{ n -1 } \ri gh ta rro w\p i^{-} _{n -1}( X^ {n-1}/ X^{ n-2 }) \xrightarro w {d _{n-1}} \l dot s \xrighta rro w{d_2} \p i^{ - }_{1} ( X ^{1}/ X^{0} ) \xri ghtarrow{ d_ 1 } \pi ^ {-}_{0} ( X^ {0}) \xrightarrow{d_ 0 }\ underline{A}\r ightar ro w 0 . $ $ S inc e there ex is ts af ree res o lu t i o n o f $\underline{ A}$ in $\ma t hrm {Mack} _{ \ma t hcal{F }}G$co n sis ting of fin itel y generat ed fre e module s , an ap plicat ion of Sch a nu el ’sle m ma( e. g.s ee[@brownLe mm a VII I.4. 4 ] ) in t his set tingshows that $\ ker d_{ n -1} $ is fini tely g enera ted, i .e. t he re exists a sur ject ion of Ma cke yfun ct ors $ $ F= \bi gop lus _{i \in I_{n}} \ma th b b { Z} ^G[-,H_i ] \righta rr o w \ ker d_n$ $ wher e $ I_ n $ is fin it e a nd $ H _ i \in \ma t hc al{F}$ f or all $i \ in I_n$ .This g iv esusan ex a ct s equenc e of Mac key f u nctors $$F \ri g htarrow \pi^ { -} _ { n- 1 }( X ^{n -1}/ X^{n-2 }) \ x righ tarr o w{ d_{ n -1}}\ldot s \ x rightarrow{d_2}\pi^ {- }_{1}( X^{1 }/ X^{0}) \x rightarrow { d _ 1} \pi^ {-}_ { 0} ( X^{0}) \xrig htarr ow{d_0} \u n derline{ A}\ri ghtarrow 0$$ If w e now appl y t hepro ced u r eof Section \[ s e c: h ar d\] sta rti ng from (\ [eq : f ree re s1\]), on e checks t ha twe ob taint he desir ed $G $- spe ctrum $X^n$. Sinc e ea ch m a p $ f_n$ is $n $ - conn ec te d, w e h av e $$\ math r m{h ocolim} _n X^n \c ong S^0. $$
In conc lusion, we ha ve construct ed astable m odel for $\underline{E}G$ thati s of fi nit e typ e. T his shows th at (2) im p lies ( 1), he nce ( 1) , ( 2 ) ,(3)a n d(4) a re all equ i v ale nt. T he | 1}$. Indeed,_suppose $_X^{n-1}$ and $f_{n-1}$ have_been constructed._Then_we obtain_an_exact sequence $$0\rightarrow_\ker d_{n-1} \rightarrow_ \pi^{-}_{n-1}( X^{n-1}/ X^{n-2})_\xrightarrow{d_{n-1}} \ldots __\xrightarrow{d_2}\pi^{-}_{1}( X^{1}/ X^{0}) \xrightarrow{d_1} \pi^{-}_{0}( X^{0}) \xrightarrow{d_0} \underline{A}\rightarrow 0.$$ Since there_exists_a free_resolution_of_$\underline{A}$ in $\mathrm{Mack}_{\mathcal{F}}G$ consisting of_finitely generated free modules, an_application of_Schanuel’s lemma (e.g. see [@brown Lemma VIII.4.4]) in this_setting_shows that $\ker_d_{n-1} $ is finitely generated, i.e. there exists a surjection_of Mackey functors $$F= \bigoplus_{i \in_I_{n}} \mathbb{Z}^G[-,H_i ]_\rightarrow_\ker_d_n$$ where $I_n$ is_finite and $H_i \in \mathcal{F}$ for_all $i \in I_n$. This gives_us an exact sequence of Mackey functors_$$F \rightarrow \pi^{-}_{n-1}( X^{n-1}/ X^{n-2})_\xrightarrow{d_{n-1}} \ldots \xrightarrow{d_2}\pi^{-}_{1}(_X^{1}/ X^{0})_ \xrightarrow{d_1} \pi^{-}_{0}( X^{0})_ \xrightarrow{d_0} \underline{A}\rightarrow_0$$ If_we now apply_the procedure of Section \[sec: hard\]_starting from (\[eq:_freeres1\]), one checks that we obtain_the_desired $G$-spectrum $X^n$._Since_each_map $f_n$_is $n$-connected, we_have_$$\mathrm{hocolim}_n X^n_\cong_S^0.$$
In conclusion, we have constructed a_stable_model for $\underline{E}G$ that is of finite_type. This shows that_(2)_implies (1), hence (1),_(2),(3) and (4) are all_equivalent. The |
(2)| = m_2(c) \le |{\partial}{\Delta}_c|^2/4 = |W|^2/4.$$
Hence, in view of Lemma \[8bb\], see also inequalities –, we may assume that, for every [pseudobracket]{} $b \in \cup_{i=0}^\ell B_i$, it is true that $$\begin{aligned}
\label{ieq1}
0 & \le b(1), b(2) \le |W|, \\ \notag
0 & \le | b(3) | \le (|W|_{a_1} + (|n_1| +|n_2|) |{\Delta}_c(2)|)/2 \\ \label{ieq2}
& \le (|W|_{a_1} + |W|^2/2)/2 \le |W|^2/2 , \\ \label{ieq3}
0 & \le b(4) \le |{\Delta}_c(2)| \le |W|^2/4 .\end{aligned}$$ Thus the space needed to store a [pseudobracket]{} $b \in \cup_{i=0}^\ell B_i$ is $O(\log |W|)$.
For every [pseudobracket system]{} $B_i$, denote $B_i = \{ b_{i, 1}, \ldots, b_{i, k_i} \}$, where, as above, $b_{i, j}(2) \le b_{i, j'}(1)$ whenever $j < j'$.
For every [pseudobracket system]{} $B_i$, we consider a path $c(i) := \Gamma(B_i)$ in ${\mathcal{T}}$ defined by the formula, hence, we have $$\label{pp3}
c(i) := \Gamma(B_i) = c_0(i) \delta(a(b_{i,1})) c_1(i) \ldots \delta(a(b_{i,k_i})) c_{k_i}(i) .$$ As before, for $B_0 = \varnothing$, we set $c(0) := \Gamma(B_0) = c$.
| ( 2)| = m_2(c) \le |{\partial}{\Delta}_c|^2/4 = |W|^2/4.$$
Hence, in view of Lemma \[8bb\ ], see also inequality –, we may wear that, for every [ pseudobracket ] { } $ b \in \cup_{i=0}^\ell B_i$, it is true that $ $ \begin{aligned }
\label{ieq1 }
0 & \le b(1), b(2) \le |W|, \\ \notag
0 & \le | b(3) | \le (|W|_{a_1 } + (|n_1| + |n_2|) |{\Delta}_c(2)|)/2 \\ \label{ieq2 }
& \le (|W|_{a_1 } + |W|^2/2)/2 \le |W|^2/2, \\ \label{ieq3 }
0 & \le b(4) \le |{\Delta}_c(2)| \le |W|^2/4 .\end{aligned}$$ therefore the space want to store a [ pseudobracket ] { } $ b \in \cup_{i=0}^\ell B_i$ is $ O(\log |W|)$.
For every [ pseudobracket arrangement ] { } $ B_i$, denote $ B_i = \ { b_{i, 1 }, \ldots, b_{i, k_i } \}$, where, as above, $ b_{i, j}(2) \le b_{i, j'}(1)$ whenever $ j < j'$.
For every [ pseudobracket arrangement ] { } $ B_i$, we consider a path $ c(i): = \Gamma(B_i)$ in $ { \mathcal{T}}$ define by the formula, hence, we get $ $ \label{pp3 }
c(i): = \Gamma(B_i) = c_0(i) \delta(a(b_{i,1 }) ) c_1(i) \ldots \delta(a(b_{i, k_i }) ) c_{k_i}(i) .$$ As before, for $ B_0 = \varnothing$, we set $ c(0): = \Gamma(B_0) = c$. | (2)| = m_2(f) \le |{\partial}{\Delta}_c|^2/4 = |W|^2/4.$$
Hekce, in view of Lgmna \[8bb\], sxe also inequalkties –, we may assume that, foc evwry [pweudobracket]{} $b \in \cup_{i=0}^\eul B_i$, it ps true tyat $$\uegin{aligned}
\label{ieq1}
0 & \le b(1), b(2) \ls |W|, \\ \uovag
0 & \le | b(3) | \le (|W|_{a_1} + (|n_1| +|n_2|) |{\Dalta}_c(2)|)/2 \\ \label{hed2}
& \le (|W|_{a_1} + |W|^2/2)/2 \le |W|^2/2 , \\ \label{ieq3}
0 & \le b(4) \je |{\Delya}_f(2)| \le |W|^2/4 .\end{alidned}$$ Ehus nht space needed to store a [pseudobdacket]{} $b \in \cup_{i=0}^\ell B_o$ is $O(\log |W|)$.
For every [pseudlbrafket system]{} $B_i$, denohe $B_i = \{ b_{i, 1}, \ldoew, b_{i, k_i} \}$, whefe, as abovt, $y_{i, j}(2) \le b_{i, n'}(1)$ whenever $j < j'$.
For every [pseudocrackzt system]{} $B_i$, ww cltsider a pavh $c(i) := \Gamma(B_i)$ in ${\mathcal{T}}$ definec by the formuka, iencw, we have $$\label{pp3}
c(i) := \Jamma(B_i) = c_0(i) \delta(a(b_{i,1})) c_1(i) \ldods \delta(a(b_{i,k_i})) c_{k_i}(i) .$$ Aw befmre, xor $C_0 = \vxrnktiinf$, we sft $r(0) := \Gamma(B_0) = d$.
| (2)| = m_2(c) \le |{\partial}{\Delta}_c|^2/4 = |W|^2/4.$$ view Lemma \[8bb\], also inequalities –, every $b \in \cup_{i=0}^\ell it is true $$\begin{aligned} \label{ieq1} 0 & \le b(1), \le |W|, \\ \notag 0 & \le | b(3) | \le (|W|_{a_1} + +|n_2|) |{\Delta}_c(2)|)/2 \\ \label{ieq2} & \le (|W|_{a_1} + |W|^2/2)/2 \le |W|^2/2 , \\ 0 \le \le \le |W|^2/4 .\end{aligned}$$ Thus the space needed to store a [pseudobracket]{} $b \in \cup_{i=0}^\ell B_i$ is |W|)$. For every [pseudobracket system]{} $B_i$, denote $B_i \{ b_{i, 1}, \ldots, k_i} \}$, where, as above, j}(2) b_{i, j'}(1)$ $j j'$. every [pseudobracket system]{} we consider a path $c(i) := \Gamma(B_i)$ in ${\mathcal{T}}$ defined by the formula, hence, we have $$\label{pp3} := \Gamma(B_i) \delta(a(b_{i,1})) c_1(i) \delta(a(b_{i,k_i})) .$$ before, for $B_0 we set $c(0) := \Gamma(B_0) = | (2)| = m_2(c) \le |{\partial}{\Delta}_c|^2/4 = |W|^2/4.$$
Hence, in View of LemmA \[8bb\], seE alSo iNeQualItieS –, we may assume thAT, for Every [pseudobracket]{} $b \in \cUp_{i=0}^\elL B_I$, It is TRuE that $$\Begin{alIGnED}
\LabEl{IeQ1}
0 & \le B(1), b(2) \LE |W|, \\ \Notag
0 & \Le | b(3) | \Le (|W|_{a_1} + (|n_1| +|n_2|) |{\DElta}_c(2)|)/2 \\ \label{Ieq2}
& \Le (|w|_{a_1} + |W|^2/2)/2 \le |W|^2/2 , \\ \label{iEQ3}
0 & \lE b(4) \le |{\Delta}_c(2)| \Le |W|^2/4 .\End{aligned}$$ ThUs tHe spacE nEedED to stOre A [pseuDobracKEt]{} $b \in \cUp_{i=0}^\ell B_i$ iS $O(\LOg |W|)$.
For EVery [pseUDObRackEt system]{} $B_i$, denote $B_I = \{ B_{i, 1}, \LDots, b_{i, k_i} \}$, where, aS above, $B_{i, J}(2) \Le B_{I, J'}(1)$ whEneVer $j < j'$.
For evErY [pseuDObrackeT SySTEM]{} $B_i$, WE consider a patH $c(i) := \Gamma(B_i)$ iN ${\MatHcal{T}}$ dEfIneD By the fOrmulA, hENce, We have $$\label{Pp3}
c(i) := \gamma(B_i) = c_0(i) \Delta(a(B_{I,1})) c_1(i) \ldotS \Delta(a(b_{I,k_i})) c_{k_i}(I) .$$ As BefOre, fOR $B_0 = \VaRnoThINg$, wE SeT $c(0) := \GAMma(b_0) = c$.
| (2)| = m_2(c) \le |{\part ial}{\Delt a}_c| ^2/ 4 = | W|^2 /4.$ $
Hence, in v i ew o f Lemma \[8bb\], see a lso i ne q uali t ie s –,we maya ss u m e t ha t, fo re ve ry [p seu dobrack et]{} $b \ in\c up_{i=0}^\el l B _i$, it is tr ue that $$\b egi n{alig ne d}\ label {ie q1}
0 & \le b(1),b(2) \le|W | , \\ \ n otag
0& \l e |b(3) | \le (|W|_ { a_ 1 } + (|n_1| +|n _2|) | {\ D el t a }_c (2) |)/2 \\ \l abel{ i eq2}
& \ l e (|W | _{a_1} + |W|^ 2/2)/2 \le|W| ^2/2 , \\\ label{ ieq3}
0 & \ le b(4) \le |{ \Delta}_c (2)| \ l e |W|^ 2 /4 .\en d{alig ned }$$ Thu s t he sp ac e ne e de d t o st ore a [p se ud obrac ket] { } $ b \i n \ cup_ {i=0} ^\ell B_i$ is $O (\lo g |W |)$.
Forever y[pseu dobrac ket s ys tem]{} $B_i$, d enot e $B_i =\{b_ {i, 1 }, \l d ots, b _{i , k _i} \}$ , where , as a b o v e, $b_{i, j}(2) \leb_ { i ,j'}(1)$whenev e r$j < j'$.
Fo r e very [ pseud obra c ke t system ]{} $B _ i$ ,we cons id er a p at h $ c(i ) :=\ Gamm a(B_i) $ in ${\ mathc a l{T}}$ defined by the formul a ,h e nc e , we ha ve $$\label {pp3 }
c(i ) := \G amm a (B_i) = c_ 0( i )\ delta(a(b_{i,1})) c _1 (i) \ldot s \delta(a(b_ {i,k_i}))c _ { k_i}(i).$$A sb efore, for $B_ 0 = \ varnothing $ , we set $c(0 ) := \Ga mma(B_0)= c$.
| (2)| =_m_2(c) _\le |{\partial}{\Delta}_c|^2/4 = |W|^2/4.$$
Hence,_in view_of_Lemma \[8bb\], see_also_inequalities –, we_may assume that,_for every [pseudobracket]{} $b \in_\cup_{i=0}^\ell B_i$, it_is_true that $$\begin{aligned}
\label{ieq1}
0 & \le b(1), b(2) \le |W|, \\ \notag
0 & \le |_b(3)_| _\le_(|W|_{a_1}_+ (|n_1| +|n_2|) |{\Delta}_c(2)|)/2 \\_ \label{ieq2}
&_\le (|W|_{a_1}_+ |W|^2/2)/2 \le |W|^2/2 , \\_\label{ieq3}
0_& \le b(4)_\le |{\Delta}_c(2)| \le |W|^2/4 .\end{aligned}$$ Thus the_space needed to store a [pseudobracket]{} $b_\in \cup_{i=0}^\ell B_i$_is_$O(\log_|W|)$.
For every [pseudobracket system]{} $B_i$,_denote $B_i = \{ b_{i, 1},_\ldots, b_{i, k_i} \}$, where, as_above, $b_{i, j}(2) \le b_{i, j'}(1)$ whenever_$j < j'$.
For every [pseudobracket system]{} $B_i$,_we consider a path $c(i)_:= \Gamma(B_i)$_in ${\mathcal{T}}$ defined by the_formula, hence, we_have $$\label{pp3}
c(i)_:= \Gamma(B_i) =_c_0(i) \delta(a(b_{i,1})) c_1(i) \ldots_\delta(a(b_{i,k_i})) c_{k_i}(i) .$$_As before, for $B_0 = \varnothing$,_we_set $c(0) :=_\Gamma(B_0)_=_c$.
|
the face identity while our method consistently preserves the identity across different head poses. Our method also consistently preserves the skin color where DR-GAN cannot.
**Quantitative Comparison:** the Fletcher Inception Distance (FID) [@fid2017] is introduced in Table \[tab:tcgan\_res3\] to quantitatively indicate the photo-realisticity of generated images compared to original real images. The closer to real images, the lower FID score. In Table \[tab:tcgan\_res3\], our method clearly achieves significantly lower FID score than other methods.
{width="100.00000%"}
\[fig:attr\_visual\_1\]
{width="100.00000%"}
\[fig:attr\_pose\_1\]
3D Attribute Generation
-----------------------
We manually select $5$ out of $40$ attributes defined in CelebA, which do not indicate face identity and only correlate with the facial area. They are Sunglasses (SG), Wearing lipstick (LS), 5’o clock shadow (SH), Smiling (SM), and Bangs (BA). We strictly follow the CelebA training, validation and testing splitting protocol. [^1]
Traditional attribute generation methods,, FaderNet [@fadernet] and AttGAN [@attgan], are trained on 2D images.
For fair comparison, we apply StarGAN and CycleGAN network structures, trained orthogonally on real, real plus pose augmented images and our UV texture and position maps. For real data in CelebA, we observe a strong head pose bias towards near-frontal poses. We calculate the pose by the reconstructed 3D shape vertices, and use it to split the testing data into $yaw < 45^{\circ}$ and $\ge 45^{\circ}$. As $yaw \ge 45^{\circ}$ testing data is very few, such as, $221$ images for lipstick, where $yaw < 45^{\circ}$ have $9288$ images, we augment the $yaw \ge 45^{\circ}$ data from near-frontal images and achieves $6735$ augmented samples to match the volume of $yaw < 45^{\circ}$.
**Attribute Generation Accuracy:** we apply an off-line attribute classifier, trained on CelebA | the face identity while our method consistently preserves the identity across unlike heading poses. Our method besides consistently preserve the skin coloring material where DR - GAN cannot.
* * Quantitative Comparison :* * the Fletcher Inception Distance (FID) [ @fid2017 ] is inaugurate in Table \[tab: tcgan\_res3\ ] to quantitatively indicate the photo - realisticity of generated images compare to original real images. The cheeseparing to veridical images, the lower FID score. In Table \[tab: tcgan\_res3\ ], our method clearly achieve significantly lower FID score than early methods.
! [ image](fig / sample6_v3.jpg){width="100.00000% " }
\[fig: attr\_visual\_1\ ]
! [ image](fig / attr_pose_7_v2.jpg){width="100.00000% " }
\[fig: attr\_pose\_1\ ]
3D Attribute Generation
-----------------------
We manually select $ 5 $ out of $ 40 $ attributes defined in CelebA, which do not bespeak face identity and only correlate with the facial area. They are Sunglasses (SG), Wearing lipstick (LS), 5’o clock shadow (SH), Smiling (SM), and Bangs (BA). We rigorously follow the CelebA training, validation and testing splitting protocol. [ ^1 ]
Traditional attribute generation method, , FaderNet [ @fadernet ] and AttGAN [ @attgan ], are trained on 2D images.
For fair comparison, we apply StarGAN and CycleGAN network structures, trained orthogonally on real, real plus pose augment images and our UV texture and stead maps. For real data in CelebA, we observe a potent pass pose bias towards near - frontal affectation. We calculate the pose by the reconstructed 3D shape vertices, and use it to split the testing datum into $ yaw < 45^{\circ}$ and $ \ge 45^{\circ}$. As $ yaw \ge 45^{\circ}$ testing data is very few, such as, $ 221 $ images for lipstick, where $ yaw < 45^{\circ}$ have $ 9288 $ images, we augment the $ yaw \ge 45^{\circ}$ data from near - frontal images and achieve $ 6735 $ augment samples to match the volume of $ yaw < 45^{\circ}$.
* * Attribute Generation Accuracy :* * we apply an off - lineage attribute classifier, trained on CelebA | thf face identity while ouv method consistgnrly prxserves the idevtity across different head 'osew. Our method also consistengly presegves the wkin xolor wherx DR-GAN gcnnot.
**Smanticavive Comparison:** the Fletcver Inception Girtcnce (FID) [@fid2017] is introduced in Table \[taf:tcgan\_rrs3\] to quantitatidely yndidate the photo-realisticity of genedated ikages comparec to original real images. Hhe floser to real imahes, the lowgd FYE score. In Txble \[tab:tcgan\_res3\], our mejhod clearly achieves significangly lpwer FID sxoee hvan other mxthods.
{widdh="100.00000%"}
\[fig:atyr\_visual\_1\]
{width="100.00000%"}
\[fig:attr\_pose\_1\]
3V Attribute Generatijn
-----------------------
We manudlmy select $5$ out of $40$ qttritutev dewunea ih RelsbA, whlch do not insicate face identity and only vowgrlate with tge facyaj area. They are Sunglasses (SG), Wearing lppstjck (LS), 5’o clock shadow (SY), Smiling (SM), and Bangd (BA). We serictly follow the CelebA training, validation and testmne skllbtine pgotocol. [^1]
Traditional attribute generation methjss,, FsderNet [@faderneb] and AttGAN [@attgan], age jrained on 2D ioages.
For rair comparison, we apply FtarGQN and CysleGSN network structures, trainwd orthogonajoy on real, real plbs pose augmznted omagex and our UV texture anb posifion maps. Flr real dzga in CelebA, we ubsvrve a strong head pose bias tjwards neer-froutal posds. Wg calcujate the plse bn the reconstructed 3D shcpe vartices, anf use it to split the testing deva into $yaw < 45^{\visc}$ dnd $\ge 45^{\cnrc}$. As $yaw \ge 45^{\circ}$ eesting data ix very yew, suzh as, $221$ imafes for lipstick, wrere $yaw < 45^{\cirw}$ have $9288$ imagev, we augient the $yaw \ge 45^{\zirc}$ data from near-fronnaj imqges and achieves $6735$ auejented samples cu match the voluke uf $raa < 45^{\cysc}$.
**Attribute Cenefatkpn Aczuracy:** we aiplh an off-line attribute cnassjfier, trained on CrlcbA | the face identity while our method consistently identity different head Our method also where cannot. **Quantitative Comparison:** Fletcher Inception Distance [@fid2017] is introduced in Table \[tab:tcgan\_res3\] quantitatively indicate the photo-realisticity of generated images compared to original real images. The to real images, the lower FID score. In Table \[tab:tcgan\_res3\], our method clearly significantly FID than methods. {width="100.00000%"} \[fig:attr\_visual\_1\] {width="100.00000%"} \[fig:attr\_pose\_1\] 3D Attribute Generation ----------------------- We manually select $5$ out of $40$ defined in CelebA, which do not indicate face and only correlate with facial area. They are Sunglasses Wearing (LS), 5’o shadow Smiling and Bangs (BA). strictly follow the CelebA training, validation and testing splitting protocol. [^1] Traditional attribute generation methods,, FaderNet [@fadernet] AttGAN [@attgan], on 2D For comparison, apply StarGAN and structures, trained orthogonally on real, real images and our UV texture and position maps. real data CelebA, we observe a strong head bias towards near-frontal poses. We calculate the pose the reconstructed 3D shape vertices, and use it to split the testing data into $yaw and $\ge 45^{\circ}$. As \ge 45^{\circ}$ testing is few, as, images for where $yaw < 45^{\circ}$ have $9288$ images, we augment the $yaw 45^{\circ}$ data from near-frontal images and achieves $6735$ augmented samples the of $yaw < **Attribute Generation Accuracy:** we an attribute classifier, trained on | the face identity while our meThod consisTentlY prEseRvEs thE ideNtity across difFErenT head poses. Our method alsO consIsTEntlY PrEservEs the skIN cOLOr wHeRe dR-GaN CAnNot.
**QuAntItative comparison:** The flEtcher InceptIOn distance (FId) [@fiD2017] is introduceD in table \[tAb:TcgAN\_res3\] tO quAntitAtivelY IndicaTe the photO-rEAlistiCIty of geNERaTed iMages compared to orIGiNAl real images. ThE closeR tO ReAL ImaGes, The lower FId sCore. IN table \[taB:TcGAN\_Res3\], OUr method clearLy achieves sIGniFicantLy LowER FID scOre thAn OTheR methods.
{widtH="100.00000%"}
\[Fig:attr\_VIsual\_1\]
{wIdTh="100.00000%"}
\[fIg:ATtr\_POsE\_1\]
3D ATTriBute GeneRaTiOn
-----------------------
We mAnuaLLY SElecT $5$ ouT of $40$ aTtribUtes defined in celEbA, wHIch Do not IndicAte fAcE idenTity anD only CoRrelate with the fAciaL area. They Are suNglAsSes (SG), wEaring LipStiCk (LS), 5’o clOck shadOW (SH), smILINg (sM), and Bangs (BA). We striCtLY FoLlow the CElebA tRAiNiNG, validatIoN anD tesTINg splIttiNG pRotocol. [^1]
TRaditiONaL aTtributE gEneratIoN meThoDs,, FadERNet [@FadernEt] and AttgAN [@atTGan], are trained oN 2d images.
For faiR CoMPArISon, wE apPly StarGAN aNd CyCLeGAn netWOrK stRUcturEs, traInED oRThogonally on real, reaL pLus posE augmEnted images anD our UV textURE And positIon mAPs. fOr real data in CeLebA, wE observe a sTRong head Pose bIas towarDs near-froNTAl poses. WE caLcuLatE thE POsE by the reconstRUCted 3d sHape verTicEs, and usE it To sPliT thE tEsting datA into $yaw < 45^{\CiRc}$ AnD $\gE 45^{\ciRc}$. As $yAW \ge 45^{\circ}$ tEsTinG dAta Is verY Few, sucH as, $221$ imAges FoR lIPstIck, wherE $YaW < 45^{\CIrc}$ hAvE $9288$ iMageS, we AuGment The $yAW \ge 45^{\Circ}$ datA from near-FroNTal iMaGeS and achIeves $6735$ augmenteD sAmples to maTcH thE volumE OF $yaw < 45^{\circ}$.
**attribute Generation AccuRAcy:** we apPly An off-Line Attribute ClaSsifieR, trAIned on celebA | the face identity while o ur methodconsi ste ntl ypres erve s the identity acro ss different head pose s. Ou rm etho d a lso c onsiste n tl y pre se rv esth e s kin c olo r where DR-GAN ca nno t.
**Quantita t iv e Comparis on: ** the Fletc her Incep ti onD istan ce(FID) [@fid 2 017] i s introdu ce d in Ta b le \[ta b : tc gan\ _res3\] to quanti t at i vely indicatethe ph ot o -r e a lis tic ity of gen er atedi mages c o mp a r e d t o original rea l images. T h e c loserto re a l imag es, t he low er FID scor e. I n Table \ [tab:t c gan\_re s 3\], ou r meth odcle arly ac hi eve ss ign i fi can t lylower FI Dsc ore t hano t h e r me tho ds.
 {widt h="10 0.00 00 0%"}
\[fig :attr \_ visual\_1\]
{widt h=" 100 .00000% "}
\[f i g:a tt r \ _ po se\_1\]
3D Attrib ut e Ge neration
----- - -- -- - -------- -- --
Wem a nuall y se l ec t $5$ ou t of $ 4 0$ a ttribut es defin ed in Ce lebA, whic h do n ot indic ate f a ce identity an d only correla t ew i th thefac ial area. T heya re S ungl a ss es( SG),Weari ng li p stick (LS), 5’o clo ck shado w (SH ), Smiling (S M), and Ba n g s (BA). W e st r ic t ly follow theCeleb A training , validat ion a nd testi ng splitt i n g protoc ol. [^ 1]
Tr a d it ional attribu t e gen er ation m eth ods,, F ade rNe t [ @fa de rnet] and AttGAN[@ at tg an ],are t r ained on 2 D i ma ges .
Fo r faircompa riso n, w e ap ply Sta r GA N andCy cl eGAN ne tw ork s truc t ure s, trai ned ortho gon a llyon r eal, re al plus poseau gmented im ag esand ou r UV textu re and position maps. F o r realdat a inCele bA, we ob ser ve a s tro n g head posebiasto war d s near - f ro nta lposes. Wec a lcu lateth e po se by t he reconstructed 3 D sh ape vertices, an d us e it to sp l itth e te s t ing data into $ yaw < 45^{ \c i rc }$ and $\g e 45 ^{ \circ}$ . As $y aw \g e 45^{\c irc}$ tes ting data i s ve r y fe w, such as , $221$images fo r lips t ic k, wh ere $yaw<45^ {\cir c}$ ha v e $ 9288$ image s, we au gment t he $yaw\ge 45^{\circ}$ data fr om nea r-fro nta l imagesand ach ieves $67 35$augmentedsam ple s tomat c h the vol u me of $yaw< 45 ^ {\circ}$.
* *At t r ib ute Generat i o n Ac curac y:* * we ap plyan off-line attri b ute classifier , tr a i ned on Cele bA | the_face identity_while our method consistently_preserves the_identity_across different_head_poses. Our method_also consistently preserves_the skin color where_DR-GAN cannot.
**Quantitative Comparison:**_the_Fletcher Inception Distance (FID) [@fid2017] is introduced in Table \[tab:tcgan\_res3\] to quantitatively indicate the photo-realisticity of_generated_images compared_to_original_real images. The closer to_real images, the lower FID_score. In_Table \[tab:tcgan\_res3\], our method clearly achieves significantly lower FID_score_than other methods.
{width="100.00000%"}
\[fig:attr\_visual\_1\]
{width="100.00000%"}
\[fig:attr\_pose\_1\]
3D_Attribute Generation
-----------------------
We manually select $5$ out of $40$ attributes_defined in CelebA, which do not_indicate face identity_and_only_correlate with the facial_area. They are Sunglasses (SG), Wearing_lipstick (LS), 5’o clock shadow (SH),_Smiling (SM), and Bangs (BA). We strictly_follow the CelebA training, validation and_testing splitting protocol. [^1]
Traditional attribute_generation methods,,_FaderNet [@fadernet] and AttGAN [@attgan], are trained_on 2D images.
For_fair comparison,_we apply StarGAN_and CycleGAN network structures, trained orthogonally_on real, real_plus pose augmented images and our_UV_texture and position_maps._For_real data_in CelebA, we_observe_a strong_head_pose bias towards near-frontal poses. We_calculate_the pose by the reconstructed 3D shape_vertices, and use it_to_split the testing data_into $yaw < 45^{\circ}$ and_$\ge 45^{\circ}$. As $yaw \ge 45^{\circ}$_testing data_is very_few, such as, $221$ images for lipstick, where $yaw < 45^{\circ}$_have $9288$ images, we augment the_$yaw \ge 45^{\circ}$ data_from near-frontal_images_and achieves $6735$_augmented_samples to_match the volume of $yaw < 45^{\circ}$.
**Attribute_Generation Accuracy:**_we apply an off-line attribute classifier,_trained on CelebA |
mid r$ by Chinese remainder theorem, and hence $$\label{eq:varrp}
\varrho_r(p)=\begin{cases}
\varrho(p) & \text{ if } p\nmid r;\\
1 &\text{ if } p\mid r.
\end{cases}$$ This implies $\varrho_r(p)=1+O(p^{-\frac{1}{2}})$ uniformly for any $p\gg 1$ and any $r\in{{\mathbb {N}}}_{\neq 0}$, so by Mertens’ first theorem on arithmetic progressions (c.f. [@Iwaniec-Kolwalski Theorem 2.2]), the hypothesis is verified for the arithmetic function $\varrho_r$ and the set ${{\mathcal {P}}}_{Q_2}^\prime$ uniformly for any $r$ (that is, the remainder term $K$ in depends only on the polynomial $Q_1$ and is independent of $r$).
Next we need to evaluate for each $d\in{{\mathbb {N}}}_{\neq 0}$ square-free, the cardinality of the subsequence ${{\mathcal {A}}}(T)_{rd}$, using and the definition of $\omega_r,\varrho_r$. $$\begin{aligned}
\#{{\mathcal {A}}}(T)_{rd}&=\frac{\omega(dr)}{(dr)^M}T^M+O(\omega(dr)T^{M-1})\\ &=\frac{\omega_r(d)}{d^M}\frac{\omega(r)}{r^{M}}T^{M}+O((dr)^{M-1+\varepsilon}T^{M-1})\\ &=\frac{\varrho_r(d)}{d}X_r+O((dr)^{M-1+\varepsilon}T^{M-1}).\end{aligned}$$ Define $$R_r(d):=\#{{\mathcal {A}}}(T)_{rd}-\frac{\varrho_r(d)}{d}X_r.$$ The above computation shows that $$R_r(d)=O((dr)^{M-1+\varepsilon}T^{M-1}).$$ On applying Theorem \[thm:I-F\] with $z=T^\lambda,y=T^\beta$, for $\lambda,\beta>0$ small enough, we get, for each $r$ square-full with all prime divisors in ${{\mathcal {P}}}_{Q_2}$,
$$\begin | mid r$ by Chinese remainder theorem, and hence $ $ \label{eq: varrp }
\varrho_r(p)=\begin{cases }
\varrho(p) & \text { if } p\nmid r;\\
1 & \text { if } p\mid r.
\end{cases}$$ This implies $ \varrho_r(p)=1+O(p^{-\frac{1}{2}})$ uniformly for any $ p\gg 1 $ and any $ r\in{{\mathbb { N}}}_{\neq 0}$, therefore by Mertens ’ inaugural theorem on arithmetic progressions (c.f. [ @Iwaniec - Kolwalski Theorem 2.2 ]), the hypothesis is control for the arithmetic function $ \varrho_r$ and the laid $ { { \mathcal { P}}}_{Q_2}^\prime$ uniformly for any $ r$ (that is, the end term $ K$ in depends only on the polynomial $ Q_1 $ and is independent of $ r$).
Next we need to measure for each $ d\in{{\mathbb { N}}}_{\neq 0}$ square - free, the cardinality of the subsequence $ { { \mathcal { A}}}(T)_{rd}$, using and the definition of $ \omega_r,\varrho_r$. $ $ \begin{aligned }
\#{{\mathcal { A}}}(T)_{rd}&=\frac{\omega(dr)}{(dr)^M}T^M+O(\omega(dr)T^{M-1})\\ & = \frac{\omega_r(d)}{d^M}\frac{\omega(r)}{r^{M}}T^{M}+O((dr)^{M-1+\varepsilon}T^{M-1})\\ & = \frac{\varrho_r(d)}{d}X_r+O((dr)^{M-1+\varepsilon}T^{M-1}).\end{aligned}$$ specify $ $ R_r(d):=\#{{\mathcal { A}}}(T)_{rd}-\frac{\varrho_r(d)}{d}X_r.$$ The above computation shows that $ $ R_r(d)=O((dr)^{M-1+\varepsilon}T^{M-1}).$$ On applying Theorem \[thm: I - F\ ] with $ z = T^\lambda, y = T^\beta$, for $ \lambda,\beta>0 $ small enough, we get, for each $ r$ square - entire with all prime divisors in $ { { \mathcal { P}}}_{Q_2}$,
$ $ \begin | mid r$ by Chinese remainder uheorem, and hence $$\label{ex:varrp}
\barrho_r(p)=\cegin{cases}
\varrho(p) & \text{ if } '\nmie r;\\
1 &\ttqt{ if } p\mid r.
\end{cases}$$ This imppies $\vareho_r(k)=1+O(p^{-\frac{1}{2}})$ uniformly for any $p\gg 1$ zkd anv $c\in{{\mathbb {N}}}_{\neq 0}$, so by Merdens’ first themrdm on arithmetic progressions (c.f. [@Iwanyec-Kolwsldki Theorem 2.2]), tre hjpjtheaps is verified for the arithmetjc funcuion $\varrho_r$ and tne set ${{\mathcal {P}}}_{Q_2}^\prime$ univormpy for any $r$ (that ls, the remaundew term $K$ in ddpends onlj on the pomynomial $Q_1$ and is independent ow $r$).
Nert we need jo egdluate for xach $d\pn{{\mathbb {N}}}_{\neq 0}$ square-xree, thr cardinality pf vhe wubsequence ${{\mathcal {A}}}(V)_{rd}$, using and the desinition mf $\omega_r,\varrho_r$. $$\betib{aligted}
\#{{\mxrhcxl {Z}}}(T)_{cd}&=\fdac{\omeha(dc)}{(dr)^M}T^M+O(\omegz(dr)T^{M-1})\\ &=\frac{\onega_r(d)}{d^M}\frac{\omega(r)}{r^{K}}T^{I}+I((dr)^{M-1+\varepsiloh}T^{M-1})\\ &=\frwc{\darrho_r(d)}{d}X_r+O((dr)^{M-1+\varepsilon}T^{M-1}).\end{aligned}$$ Dtfine $$R_r(d):=\#{{\mathcal {A}}}(T)_{rd}-\frac{\vareho_r(d)}{d}X_r.$$ The above colputation shows that $$R_r(d)=O((dr)^{M-1+\varepsilon}T^{M-1}).$$ On applying Theoram \[thj:K-F\] clbh $z=G^\oalbda,y=T^\beta$, for $\lambda,\beta>0$ small enough, we get, rot vach $r$ square-full with all ptile qivisors in ${{\mxthcal {P}}}_{Q_2}$,
$$\gegin | mid r$ by Chinese remainder theorem, and \varrho_r(p)=\begin{cases} & \text{ } p\nmid r;\\ r. This implies $\varrho_r(p)=1+O(p^{-\frac{1}{2}})$ for any $p\gg and any $r\in{{\mathbb {N}}}_{\neq 0}$, so Mertens’ first theorem on arithmetic progressions (c.f. [@Iwaniec-Kolwalski Theorem 2.2]), the hypothesis is for the arithmetic function $\varrho_r$ and the set ${{\mathcal {P}}}_{Q_2}^\prime$ uniformly for any (that the term in depends only on the polynomial $Q_1$ and is independent of $r$). Next we need to for each $d\in{{\mathbb {N}}}_{\neq 0}$ square-free, the cardinality the subsequence ${{\mathcal {A}}}(T)_{rd}$, and the definition of $\omega_r,\varrho_r$. \#{{\mathcal &=\frac{\omega_r(d)}{d^M}\frac{\omega(r)}{r^{M}}T^{M}+O((dr)^{M-1+\varepsilon}T^{M-1})\\ &=\frac{\varrho_r(d)}{d}X_r+O((dr)^{M-1+\varepsilon}T^{M-1}).\end{aligned}$$ $$R_r(d):=\#{{\mathcal The computation shows that On applying Theorem \[thm:I-F\] with $z=T^\lambda,y=T^\beta$, for $\lambda,\beta>0$ small enough, we get, for each $r$ square-full with prime divisors {P}}}_{Q_2}$, $$\begin | mid r$ by Chinese remainder theOrem, and henCe $$\labEl{eQ:vaRrP}
\varRho_r(P)=\begin{cases}
\varRHo(p) & \tExt{ if } p\nmid r;\\
1 &\text{ if } p\mid r.
\End{caSeS}$$ this IMpLies $\vArrho_r(p)=1+o(P^{-\fRAC{1}{2}})$ unIfOrMly FoR AnY $p\gg 1$ aNd aNy $r\in{{\maThbb {N}}}_{\neq 0}$, so By MErTens’ first theOReM on arithmeTic Progressions (C.f. [@IWaniec-koLwaLSki ThEorEm 2.2]), the HypothESis is vErified foR tHE arithMEtic funCTIoN $\varRho_r$ and the set ${{\mathCAl {p}}}_{q_2}^\prime$ uniformlY for anY $r$ (THaT IS, thE reMainder terM $K$ In depENds only ON tHE POlyNOmial $Q_1$ and is inDependent of $R$).
nexT we neeD tO evALuate fOr eacH $d\IN{{\maThbb {N}}}_{\neq 0}$ squAre-fRee, the carDinaliTY of the sUBsequenCe ${{\mathCal {a}}}(T)_{rD}$, usiNG aNd The DeFIniTIoN of $\OMegA_r,\varrho_R$. $$\bEgIn{aliGned}
\#{{\MATHCal {A}}}(t)_{rd}&=\Frac{\Omega(Dr)}{(dr)^M}T^M+O(\omega(Dr)T^{m-1})\\ &=\fraC{\OmeGa_r(d)}{d^m}\frac{\OmegA(r)}{R^{M}}T^{M}+O((Dr)^{M-1+\varEpsilOn}t^{M-1})\\ &=\frac{\varrho_r(d)}{d}x_r+O((dR)^{M-1+\varepsiLon}t^{M-1}).\End{AlIgned}$$ dEfine $$R_R(d):=\#{{\mAthCal {A}}}(T)_{rd}-\Frac{\varRHo_r(D)}{d}x_R.$$ tHe Above computation shOwS THaT $$R_r(d)=O((dr)^{M-1+\VarepsILoN}T^{m-1}).$$ on applyiNg theOrem \[THM:I-F\] wiTh $z=T^\LAmBda,y=T^\betA$, for $\laMBdA,\bEta>0$ smalL eNough, wE gEt, fOr eAch $r$ sQUare-Full wiTh all priMe divISors in ${{\mathcal {P}}}_{q_2}$,
$$\Begin | mid r$ by Chinese remainde r theorem, andhen ce$$ \lab el{e q:varrp}
\varr h o_r( p)=\begin{cases}
\varr ho(p) & \tex t {if }p\nmidr ;\ \ 1 & \t ex t{if }p\mid r.
\end{c ases}$$ Th isim plies $\varr h o_ r(p)=1+O(p ^{- \frac{1}{2}} )$unifor ml y f o r any $p \gg 1 $ anda ny $r\ in{{\math bb {N}}}_ { \neq 0} $ , s o by Mertens’ first t h eo r em on arithmet ic pro gr e ss i o ns(c. f. [@Iwani ec -Kolw a lski Th e or e m 2.2 ] ), the hypoth esis is ver i fie d forth e a r ithmet ic fu nc t ion $\varrho_r $ an d the set ${{\m a thcal { P }}}_{Q_ 2}^\pr ime $ u nifo r ml yfor a n y $ r $(th a t i s, the r em ai nderterm $ K $ indep ends only on the polyn omi al $ Q _1$ andis in depe nd ent o f $r$) .
Ne xt we need to eva luat e for eac h $ d\ in{ {\ mathb b {N}}} _{\ neq 0}$ sq uare-fr e e,th e c ar dinality of the su bs e q ue nce ${{\ mathca l { A} } }(T)_{rd }$ , u sing a nd th e de f in ition of $\ome g a_ r, \varrho _r $. $$\ be gin {al igned }
\#{{\ mathcal{A}}} ( T)_{rd}&=\frac { \omega(dr)}{( d r) ^ M }T ^ M+O( \om ega(dr)T^{M -1}) \ \ &= \fra c {\ ome g a_r(d )}{d^ M} \ fr a c{\omega(r)}{r^{M}} T^ {M}+O( (dr)^ {M-1+\varepsi lon}T^{M-1 } ) \ \ &=\fra c{\v a rr h o_r(d)}{d}X_r+ O((dr )^{M-1+\va r epsilon} T^{M- 1}).\end {aligned} $ $ Define$$R _r( d): =\# { { \m athcal {A}}}( T ) _{rd }- \frac{\ var rho_r(d )}{ d}X _r. $$Th e above c omputati on s ho ws th at $$ R _r(d)=O( (d r)^ {M -1+ \vare p silon} T^{M- 1}). $$ O n ap plyingT he o r em \ [t hm :I-F \]wi th $z =T^\ l amb da,y=T^ \beta$, f or$ \lam bd a, \beta>0 $ small enoug h, we get, f or ea ch $r$ s quare-fu ll with all prime divis o rs in $ {{\ mathc al { P}}}_{Q_2 }$,
$$\b egi n | mid r$_by Chinese_remainder theorem, and hence_$$\label{eq:varrp}
\varrho_r(p)=\begin{cases}
\varrho(p) &_\text{_if }_p\nmid_r;\\
1 &\text{ if_} p\mid r.
\end{cases}$$_This implies $\varrho_r(p)=1+O(p^{-\frac{1}{2}})$ uniformly_for any $p\gg_1$_and any $r\in{{\mathbb {N}}}_{\neq 0}$, so by Mertens’ first theorem on arithmetic progressions (c.f._[@Iwaniec-Kolwalski_Theorem 2.2]),_the_hypothesis_is verified for the arithmetic_function $\varrho_r$ and the set_${{\mathcal {P}}}_{Q_2}^\prime$_uniformly for any $r$ (that is, the remainder_term_$K$ in depends_only on the polynomial $Q_1$ and is independent of_$r$).
Next we need to evaluate for_each $d\in{{\mathbb {N}}}_{\neq_0}$_square-free,_the cardinality of the_subsequence ${{\mathcal {A}}}(T)_{rd}$, using and the_definition of $\omega_r,\varrho_r$. $$\begin{aligned}
_ \#{{\mathcal {A}}}(T)_{rd}&=\frac{\omega(dr)}{(dr)^M}T^M+O(\omega(dr)T^{M-1})\\ &=\frac{\omega_r(d)}{d^M}\frac{\omega(r)}{r^{M}}T^{M}+O((dr)^{M-1+\varepsilon}T^{M-1})\\ &=\frac{\varrho_r(d)}{d}X_r+O((dr)^{M-1+\varepsilon}T^{M-1}).\end{aligned}$$ Define $$R_r(d):=\#{{\mathcal_{A}}}(T)_{rd}-\frac{\varrho_r(d)}{d}X_r.$$ The above computation shows that_$$R_r(d)=O((dr)^{M-1+\varepsilon}T^{M-1}).$$ On applying Theorem \[thm:I-F\]_with $z=T^\lambda,y=T^\beta$,_for $\lambda,\beta>0$ small enough, we_get, for each_$r$ square-full_with all prime_divisors in ${{\mathcal {P}}}_{Q_2}$,
$$\begin |
ultured 4T1 cancer cells have been seeded and then attached over a specific portion of the surface of a petri dish held by the LT device (see \[fig:LT\]). The bottom of the dish has been filled with a radioactive incubation medium containing both glucose at physiological concentration 1 g/L, i.e. 5.5mM, and an amount of FDG corresponding to about $10^6$ Bq, diluted in a volume of 3 mL. Notice that the amount of FDG can be considered negligible with respect to that of glucose, i.e., FDG has to be regarded as a perturbation of glucose. The dish has been subject to a periodic motion around its axis, inclined from the vertical; at each rotation cycle, lasting one minute, the LT device has collected the radioactivity emitted by the cells. Experiments have been performed for a total time interval of 180 minutes.
In the course of a typical experiment, radioactive FDG molecules, initially added to the incubation medium, have been uptaken and then retained by the cell culture. For each experiment we have considered the time dependent total activity (Bq) of the cell culture $\mathcal{A}_T$, and the corresponding input function $A_i$, describing the activity inside the incubation medium. Both activity curves have been decay corrected. Since the LT is a closed system for radioactive molecules, the two curves satisfy the conservation law $A_i + \mathcal{A}_T= A_{i0}$, where the known constant $A_{i0}$ represents the activity in the medium available at the beginning of the measuring procedure, after absorption by wet surfaces, which occurs in a very short time interval. The total activity $\mathcal{A}_T$ has been reconstructed on the basis of the counts of the detector available with the LT. The input function has been determined by the conservation law as $A_i = A_{i0}-\mathcal{A}_T$.
Following [@Milo], we have chosen the value of the intracellular relative size of the ER with respect to the cytosol as $v = 0.17$, which holds for a rough ER in a liver hepatocyte cell. We have considered six LT experiments, denoted as e$i$, with $i =1,\dots,6$, differing between each other for number of cells $N_c$ and initial amount of FDG in the medium $A_{i0}$. \[tab:exp\_data\_cells\] reports the experimental values of | ultured 4T1 cancer cells have been seeded and then attached over a specific portion of the open of a petri smasher held by the LT device (see \[fig: LT\ ]). The bottom of the cup of tea has been filled with a radioactive incubation metier contain both glucose at physiological assiduity 1 deoxyguanosine monophosphate / L, i.e. 5.5mM, and an amount of FDG represent to about $ 10 ^ 6 $ Bq, diluted in a book of 3 mL. Notice that the sum of FDG can be considered negligible with respect to that of glucose, i.e., FDG has to be regarded as a perturbation of glucose. The smasher has been subject to a periodic motion around its axis, slope from the vertical; at each rotation cycle, lasting one minute, the LT device has roll up the radioactivity emitted by the cells. Experiments have been performed for a entire time interval of 180 minutes.
In the course of a typical experiment, radioactive FDG molecules, initially added to the incubation medium, have been uptaken and then retained by the cell culture. For each experiment we have considered the time dependent total activity (Bq) of the cellular telephone acculturation $ \mathcal{A}_T$, and the corresponding stimulation function $ A_i$, describing the activity inside the incubation culture medium. Both activity curves have been decay corrected. Since the LT is a shut system for radioactive molecules, the two curves satisfy the conservation law $ A_i + \mathcal{A}_T= A_{i0}$, where the known constant $ A_{i0}$ represents the activity in the medium available at the beginning of the measuring procedure, after concentration by wet open, which occurs in a very short time interval. The total natural process $ \mathcal{A}_T$ has been reconstructed on the basis of the counts of the detector available with the LT. The stimulation function has been determined by the conservation law as $ A_i = A_{i0}-\mathcal{A}_T$.
Following [ @Milo ], we have choose the value of the intracellular relative size of the ER with respect to the cytosol as $ volt = 0.17 $, which holds for a harsh ER in a liver hepatocyte cellular telephone. We have considered six LT experiments, announce as e$i$, with $ i = 1,\dots,6 $, differing between each other for issue of cells $ N_c$ and initial sum of FDG in the medium $ A_{i0}$. \[tab: exp\_data\_cells\ ] report the experimental values of | ultkred 4T1 cancer cells have been seeded anb then ettaches over a specific portion of the surhace of a petri dish held by thd LT devibe (see \[fit:LT\]). Uhe bottom of the dish has been nilleb xith a radioactlve incubathon medium condaknnng both glucose at physiological cogcentrayiln 1 g/L, i.e. 5.5mM, agd am amohnt of FDG corresponding to about $10^6$ Bq, dillted in a volume pf 3 mL. Notice that the amoknt lf FDG can be consldered neglugibjw with respezt to that of glucose, j.e., FDG has to be regarded as a oertuxbation of tlycodg. The dish hes beeg subject to a periogic motoon around its axms, ibclined from the vertmcal; at each rotatiog cycle, ldscing one minute, the LR eevicg has coluwctdd uhe rzdioachivmty emitted by the celos. Experiments have bqvm performed ror a eoeal time interval of 180 minutes.
In the cougse kf a typical experiment, radioactive FDG molefules, inieially added to the incubation medium, have been u[takei xnd tmen fwtwined by the cell culture. For each experiment se hsve considered the time drpfncgnt total actixity (Bq) or the cell culture $\mathcaj{A}_T$, abd the cowreslonding input function $A_i$, dwscribing thv acrivity inside the nncubation mzdium. Noth sctivity curves have bezn deczy correctef. Since tgd LT is a closed syxtam for radioactive moleculqs, the twi cuxves satksfy the cjnservatioj law $A_i + \mathcal{A}_T= A_{i0}$, wjere jhe knmwn constajt $A_{i0}$ represents the activity ii the medium svdildble at che beninning of the ieasuring procgdure, aftzr absurption by wet sucfaces, which occurs in a egry short timx intervaj. Thw toral actkxity $\mathcal{A}_T$ has been reconstruxted on the basis pf jhs counts of the deuextor available wigh ehv LV. The hnput functimn hxs cren ddtermintb bn tfe cpnservation law as $A_h = A_{j0}-\mathcal{A}_T$.
Followinb [@Iilo], we yave chofen the value of the intracellupar rxlativx size of the ER with respect to the cyfosol as $g = 0.17$, which holds for a rough ER nn a liver hepatocyte cell. We have consivered six LT experimenjs, denoted as e$i$, with $l =1,\dots,6$, diffecing bqtween eawh other for number if cells $N_c$ and ikitial amount of FDG ih the kediul $A_{i0}$. \[tab:exp\_data\_cells\] reports the experimental values of | ultured 4T1 cancer cells have been seeded attached a specific of the surface by LT device (see The bottom of dish has been filled with a incubation medium containing both glucose at physiological concentration 1 g/L, i.e. 5.5mM, and amount of FDG corresponding to about $10^6$ Bq, diluted in a volume of mL. that amount FDG can be considered negligible with respect to that of glucose, i.e., FDG has to be as a perturbation of glucose. The dish has subject to a periodic around its axis, inclined from vertical; each rotation lasting minute, LT device has the radioactivity emitted by the cells. Experiments have been performed for a total time interval of 180 In the a typical radioactive molecules, added to the have been uptaken and then retained culture. For each experiment we have considered the dependent total (Bq) of the cell culture $\mathcal{A}_T$, the corresponding input function $A_i$, describing the activity the incubation medium. Both activity curves have been decay corrected. Since the LT is a for radioactive molecules, the curves satisfy the law + A_{i0}$, the known $A_{i0}$ represents the activity in the medium available at the beginning the measuring procedure, after absorption by wet surfaces, which occurs very time interval. The activity $\mathcal{A}_T$ has been on basis of the counts detector with input has determined by the conservation as $A_i = A_{i0}-\mathcal{A}_T$. Following we have chosen the size of the ER with respect to the as $v = 0.17$, which holds for rough ER in a liver hepatocyte cell. We have considered six LT denoted as $i =1,\dots,6$, differing between each other for number cells $N_c$ and initial of FDG in the medium $A_{i0}$. \[tab:exp\_data\_cells\] reports the values | ultured 4T1 cancer cells have beEn seeded anD then AttAchEd Over A speCific portion of THe suRface of a petri dish held bY the Lt dEVice (SEe \[Fig:LT\]). the bottOM oF THe dIsH hAs bEeN FiLled wIth A radioaCtive incubAtiOn Medium contaiNInG both glucoSe aT physiologicAl cOncentRaTioN 1 G/L, i.e. 5.5mm, anD an amOunt of fdG corrEsponding To ABout $10^6$ Bq, DIluted iN A VoLume Of 3 mL. Notice that the AMoUNt of FDG can be coNsiderEd NEgLIGibLe wIth respect To That oF Glucose, I.E., Fdg HAs tO Be regarded as a PerturbatioN Of gLucose. thE diSH has beEn subJeCT to A periodic moTion Around its Axis, inCLined frOM the verTical; aT eaCh rOtatIOn CyCle, LaSTinG OnE miNUte, The LT devIcE hAs colLectED THE radIoaCtivIty emItted by the celLs. EXperIMenTs havE been PerfOrMed foR a totaL time InTerval of 180 minutes.
in thE course of A tyPiCal ExPerimENt, radiOacTivE FDG molEcules, iNItiAlLY ADdEd to the incubation mEdIUM, hAve been uPtaken ANd ThEN retaineD bY thE celL CUlturE. For EAcH experimEnt we hAVe CoNsidereD tHe time DePenDenT totaL ActiVity (Bq) Of the celL cultURe $\mathcal{A}_T$, and THe correspondiNG iNPUt FUnctIon $a_i$, describinG the ACtivIty iNSiDe tHE incuBatioN mEDiUM. Both activity curves HaVe been Decay Corrected. SincE the LT is a cLOSEd system For rADiOActive moleculeS, the tWo curves saTIsfy the cOnserVation laW $A_i + \mathcaL{a}_t= A_{i0}$, where The KnoWn cOnsTANt $a_{i0}$ represents tHE ActiViTy in the MedIum avaiLabLe aT thE beGiNning of thE measuriNg PrOcEdUre, After ABsorptioN bY weT sUrfAces, wHIch occUrs in A verY sHoRT tiMe interVAl. tHE totAl AcTiviTy $\mAtHcal{A}_t$ has BEen ReconstRucted on tHe bASis oF tHe Counts oF the detector aVaIlable with ThE LT. the inpUT Function Has been determined by the cONservatIon Law as $a_i = A_{i0}-\Mathcal{A}_T$.
folLowing [@milO], We have Chosen The vaLuE of THE intrACElLulAr Relative siZE Of tHe ER wItH resPect to tHe cytosol as $v = 0.17$, which hOLds For a rough ER in A liVer hEPAtOcyTE cELl. WE hAVe cONSidered six LT expEriments, deNoTEd As e$i$, with $i =1,\dOTs,6$, dIfFering bEtween eAch otHEr for nuMber of celLs $N_c$ and inItIal aMOUnt Of FDG in the Medium $A_{i0}$. \[Tab:exp\_datA\_Cells\] REpOrts tHe eXperimEnTal ValueS of | ultured 4T1 cancer cells h ave been s eeded an d t he n at tach ed over a spec i ficportion of the surface of a p e trid is h hel d by th e L T dev ic e(se e\[ fig:L T\] ). Thebottom ofthe d ish has been fi lled witha r adioactive i ncu bation m edi u m con tai ningboth g l ucoseat physio lo g ical c o ncentra t i on 1 g /L, i.e. 5.5mM, a n da n amount of FD G corr es p on d i ngtoabout $10^ 6$ Bq,d ilutedi na v olu m e of 3 mL. No tice that t h e a mountof FD G can b e con si d ere d negligibl e wi th respec t to t h at of g l ucose,i.e.,FDG ha s to be r ega rd e d a s a pe r tur bation o fgl ucose . Th e d i sh h asbeen subj ect to a peri odi c mo t ion arou nd it s ax is , inc linedfromth e vertical; ateach rotation cy cl e,la sting one mi nut e,the LTdeviceh asco l l e ct ed the radioactivi ty e mi tted bythe ce l ls .E xperimen ts ha ve b e e n per form e dfor a to tal ti m ein tervalof 180 m in ute s.
In t h e co urse o f a typi cal e x periment, radi o active FDG mo l ec u l es , ini tia lly added t o th e inc ubat i on me d ium,havebe e nu ptaken and then ret ai ned by thecell culture. For eache x p erimentwe h a ve considered the time dependent total ac tivit y (Bq) o f the cel l culture$\m ath cal {A} _ T $, and the corr e s pond in g input fu nction$A_ i$, de scr ib ing the a ctivityin si de t heincub a tion med iu m.Bo thactiv i ty cur ves h avebe en dec ay corr e ct e d . Si nc etheLTis a cl osed sys tem for radioact ive mole cu le s, thetwo curves sa ti sfy the co ns erv ationl a w $A_i + \mathcal{A}_T= A_{i0}$ , wherethe know n co nstant $A _{i 0}$ re pre s ents t he act ivity i n t h e medi u m a vai la ble at the b egi nning o f th e measu ring procedure, af t erabsorption by we t su r f ac es, wh i choc c urs i n a very shorttime inter va l .The totala cti vi ty $\ma thcal{A }_T$h as been reconstr ucted onth e ba s i s o f the coun ts of th e detecto r avai l ab le wi ththe LT .The inpu t func t ion hasbeen d et ermine d byth e conser vation law as $A_i = A_ {i0}-\ mathc al{ A}_T$.
F oll o win g [@Milo] , we have chos enthe valu e o f theintr a ce llu l ar re lati v e size of th e E R wi th respectt o the cyto sol as $v= 0. 17$, which holdsf or a rough ERin a l ive r h e pato cy te cell. We ha veco n s idered s ix LT experim ents, de no t ed as e$i$, with$i =1,\ d o ts , 6$, di ffer ing betweeneac ho ther fo rnu m ber of cel ls $N_c$ and i n itia l amount of FDG in them e dium$ A_{ i0}$. \[tab:e x p\_d ata\_cells \] reportsthe ex peri menta l value sof | ultured 4T1_cancer cells_have been seeded and_then attached_over_a specific_portion_of the surface_of a petri_dish held by the_LT device (see_ \[fig:LT\])._The bottom of the dish has been filled with a radioactive incubation medium containing_both_glucose at_physiological_concentration_1 g/L, i.e. 5.5mM, and_an amount of FDG corresponding_to about_$10^6$ Bq, diluted in a volume of 3_mL._Notice that the_amount of FDG can be considered negligible with respect_to that of glucose, i.e., FDG_has to be_regarded_as_a perturbation of glucose._The dish has been subject to_a periodic motion around its axis,_inclined from the vertical; at each rotation_cycle, lasting one minute, the LT_device has collected the radioactivity_emitted by_the cells. Experiments have been_performed for a_total time_interval of 180_minutes.
In the course of a typical_experiment, radioactive FDG_molecules, initially added to the incubation_medium,_have been uptaken_and_then_retained by_the cell culture._For_each experiment_we_have considered the time dependent total_activity_(Bq) of the cell culture $\mathcal{A}_T$, and_the corresponding input function_$A_i$,_describing the activity inside_the incubation medium. Both activity_curves have been decay corrected. Since_the LT_is a_closed system for radioactive molecules, the two curves satisfy the conservation_law $A_i + \mathcal{A}_T= A_{i0}$, where_the known constant $A_{i0}$_represents the_activity_in the medium_available_at the_beginning of the measuring procedure, after absorption_by wet_surfaces, which occurs in a very_short time interval. The_total_activity $\mathcal{A}_T$ has been reconstructed on_the basis of the counts of_the detector available with the_LT._The_input function has been determined_by the conservation law as $A_i_= A_{i0}-\mathcal{A}_T$.
Following [@Milo],_we have chosen the value of the_intracellular_relative size of the ER with_respect_to the cytosol as $v =_0.17$,_which_holds for a rough ER_in a liver hepatocyte cell. We_have considered six LT experiments, denoted as e$i$, with_$i =1,\dots,6$, differing_between each other for number_of_cells_$N_c$ and initial amount of FDG in the medium $A_{i0}$._ \[tab:exp\_data\_cells\] reports_the experimental values_of |
\equiv0(mod\,7)$.
Let $A=B\cdot L_{1}(7^{k}n)=2^{7^{k}3n}-1$. Let’s prove by induction on $k\geq0$, that $$\textcolor[rgb]{0.00,0.00,0.50}{2^{7^{k}3n}-1\equiv0(mod\,7^{k+1}).}\label{T3:1}$$ The case $k=0$ is obvious. Let’s make the inductive assumption, that for $k\leq m-1$ the comparison is fair. Let’s consider expression $$\begin{gathered}
\notag \textcolor[rgb]{0.00,0.00,0.50}{2^{7^{m}3n}-1=(2^{7^{m-1}3n})^{7}-1=}\\
\textcolor[rgb]{0.00,0.00,0.50}{=(2^{7^{m-1}3n}-1)\cdot\sum_{i=0}^6(2^{7^{m-1}3n})^{i}\equiv0(mod\,7^{m}\cdot7)\equiv0(mod\,7^{m+1}).}\end{gathered}$$
Numbers $L_{2}(n) = 2^{2n} + 2^{n} - 1$
=======================================
First five prime numbers $L_{2}(1)=5$, $L_{2}(2)=19$, $L_{2}(3)=71$,\
$L_{2}(4)=271$, $L_{2}(6)=4159$. From **the statement 1** validity of the\
comparisons follows $$L_{2}(4k+1)\equiv0(mod\,5),\ L_{2}(10k+7)\equiv L_{2}(10k+8)\equiv0(mod\,11),$$ where $k\geq0$ is integer.
The numbers $L_{2}(n)$ are not free from squares $L_{2}(68)\equiv 0(mod\,11^{3})$,\
$L_{2}(97)\equiv 0(mod\,11^{2})$.
The author has checked up the following worthy to attention facts for numbers $L_{2}(n)$.
1\) For prime numbers $p\leq5003$ the | \equiv0(mod\,7)$.
Let $ A = B\cdot L_{1}(7^{k}n)=2^{7^{k}3n}-1$. Let ’s prove by evocation on $ k\geq0 $, that $ $ \textcolor[rgb]{0.00,0.00,0.50}{2^{7^{k}3n}-1\equiv0(mod\,7^{k+1}).}\label{T3:1}$$ The subject $ k=0 $ is obvious. Let ’s make the inductive assumption, that for $ k\leq m-1 $ the comparison is average. Let ’s consider construction $ $ \begin{gathered }
\notag \textcolor[rgb]{0.00,0.00,0.50}{2^{7^{m}3n}-1=(2^{7^{m-1}3n})^{7}-1=}\\
\textcolor[rgb]{0.00,0.00,0.50}{=(2^{7^{m-1}3n}-1)\cdot\sum_{i=0}^6(2^{7^{m-1}3n})^{i}\equiv0(mod\,7^{m}\cdot7)\equiv0(mod\,7^{m+1}).}\end{gathered}$$
number $ L_{2}(n) = 2^{2n } + 2^{n } - 1 $
= = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = =
First five premier numbers $ L_{2}(1)=5 $, $ L_{2}(2)=19 $, $ L_{2}(3)=71$,\
$ L_{2}(4)=271 $, $ L_{2}(6)=4159$. From * * the statement 1 * * robustness of the\
comparisons follows $ $ L_{2}(4k+1)\equiv0(mod\,5),\ L_{2}(10k+7)\equiv L_{2}(10k+8)\equiv0(mod\,11),$$ where $ k\geq0 $ is integer.
The number $ L_{2}(n)$ are not free from squares $ L_{2}(68)\equiv 0(mod\,11^{3})$,\
$ L_{2}(97)\equiv 0(mod\,11^{2})$.
The author has checked up the following worthy to care facts for numbers $ L_{2}(n)$.
1\) For prime number $ p\leq5003 $ the | \equlv0(mod\,7)$.
Let $A=B\cdot L_{1}(7^{k}n)=2^{7^{k}3n}-1$. Leu’s prove by inducjiin on $n\geq0$, tgat $$\textzolor[rgb]{0.00,0.00,0.50}{2^{7^{k}3n}-1\equiv0(mod\,7^{k+1}).}\label{T3:1}$$ Thx cawe $k=0$ us obvious. Let’s make tfe inductpve assumptioi, that for $k\leq m-1$ the compariskk is yamr. Let’s considet expression $$\begin{gathered}
\togay \textcolor[rgb]{0.00,0.00,0.50}{2^{7^{m}3n}-1=(2^{7^{m-1}3n})^{7}-1=}\\
\textcolor[rgb]{0.00,0.00,0.50}{=(2^{7^{m-1}3n}-1)\cdot\fum_{i=0}^6(2^{7^{m-1}3n})^{i}\rqkiv0(mod\,7^{m}\cdot7)\equyv0(moc\,7^{i+1}).}\end{fathered}$$
Numbers $L_{2}(n) = 2^{2n} + 2^{n} - 1$
=======================================
First fibe primt numbers $L_{2}(1)=5$, $L_{2}(2)=19$, $L_{2}(3)=71$,\
$L_{2}(4)=271$, $L_{2}(6)=4159$. From **the statement 1** validlty lf the\
comparisons vollows $$L_{2}(4k+1)\ewuiv0(iid\,5),\ L_{2}(10k+7)\equiv U_{2}(10k+8)\equiv0(mod\,11),$$ where $k\geq0$ js integer.
The numbers $L_{2}(n)$ are nog frez from squatzw $L_{2}(68)\fxuiv 0(mod\,11^{3})$,\
$L_{2}(97)\eqniv 0(moq\,11^{2})$.
The author mss chewked up the following wocthy to attention facts fmr numbers $L_{2}(n)$.
1\) For krime numbara $p\leq5003$ the | \equiv0(mod\,7)$. Let $A=B\cdot L_{1}(7^{k}n)=2^{7^{k}3n}-1$. Let’s prove by $k\geq0$, $$\textcolor[rgb]{0.00,0.00,0.50}{2^{7^{k}3n}-1\equiv0(mod\,7^{k+1}).}\label{T3:1}$$ The $k=0$ is obvious. that $k\leq m-1$ the is fair. Let’s expression $$\begin{gathered} \notag \textcolor[rgb]{0.00,0.00,0.50}{2^{7^{m}3n}-1=(2^{7^{m-1}3n})^{7}-1=}\\ \textcolor[rgb]{0.00,0.00,0.50}{=(2^{7^{m-1}3n}-1)\cdot\sum_{i=0}^6(2^{7^{m-1}3n})^{i}\equiv0(mod\,7^{m}\cdot7)\equiv0(mod\,7^{m+1}).}\end{gathered}$$ Numbers = 2^{2n} + 2^{n} - 1$ ======================================= First five prime numbers $L_{2}(1)=5$, $L_{2}(2)=19$, $L_{2}(4)=271$, $L_{2}(6)=4159$. From **the statement 1** validity of the\ comparisons follows $$L_{2}(4k+1)\equiv0(mod\,5),\ L_{2}(10k+7)\equiv where is The $L_{2}(n)$ are not free from squares $L_{2}(68)\equiv 0(mod\,11^{3})$,\ $L_{2}(97)\equiv 0(mod\,11^{2})$. The author has checked up the worthy to attention facts for numbers $L_{2}(n)$. 1\) prime numbers $p\leq5003$ the | \equiv0(mod\,7)$.
Let $A=B\cdot L_{1}(7^{k}n)=2^{7^{k}3n}-1$. Let’S prove by inDuctiOn oN $k\gEq0$, That $$\TextColor[rgb]{0.00,0.00,0.50}{2^{7^{k}3n}-1\equiV0(Mod\,7^{k+1}).}\Label{T3:1}$$ The case $k=0$ is obviouS. Let’s MaKE the INdUctivE assumpTIoN, THat FoR $k\Leq M-1$ tHE cOmparIsoN is fair. let’s considEr eXpRession $$\begin{GAtHered}
\notag \TexTcolor[rgb]{0.00,0.00,0.50}{2^{7^{m}3n}-1=(2^{7^{m-1}3N})^{7}-1=}\\
\teXtcoloR[rGb]{0.00,0.00,0.50}{=(2^{7^{m-1}3N}-1)\Cdot\sUm_{i=0}^6(2^{7^{M-1}3n})^{i}\eqUiv0(mod\,7^{M}\Cdot7)\eqUiv0(mod\,7^{m+1}).}\enD{gAThered}$$
nUmbers $L_{2}(N) = 2^{2N} + 2^{N} - 1$
=======================================
FIrst Five prime numbers $L_{2}(1)=5$, $l_{2}(2)=19$, $l_{2}(3)=71$,\
$L_{2}(4)=271$, $l_{2}(6)=4159$. from **the statemeNt 1** valiDiTY oF THe\
cOmpArisons folLoWs $$L_{2}(4k+1)\eQUiv0(mod\,5),\ L_{2}(10K+7)\EqUIV l_{2}(10k+8)\eQUiv0(mod\,11),$$ where $k\gEq0$ is integer.
tHe nUmbers $l_{2}(n)$ Are NOt free From sQuARes $l_{2}(68)\equiv 0(mod\,11^{3})$,\
$L_{2}(97)\eQuiv 0(Mod\,11^{2})$.
The autHor has CHecked uP The follOwing wOrtHy tO attENtIoN faCtS For NUmBerS $l_{2}(n)$.
1\) FOr prime nUmBeRs $p\leQ5003$ the | \equiv0(mod\,7)$.
Let $A= B\cdot L_{ 1}(7^ {k} n)= 2^ {7^{ k}3n }-1$. Let’s pr o ve b y induction on $k\geq0 $, th at $$\t e xt color [rgb]{0 . 00 , 0 .00 ,0 .5 0}{ 2^ { 7^ {k}3n }-1 \equiv0 (mod\,7^{k +1} ). }\label{T3:1 } $$ The case$k= 0$ is obviou s.Let’sma ket he in duc tiveassump t ion, t hat for $ k\ l eq m-1 $ the co m p ar ison is fair. Let’s c o ns i der expression $$\be gi n {g a t her ed}
\notag \t ex tcolo r [rgb]{0 . 00 , 0 . 00, 0 .50}{2^{7^{m} 3n}-1=(2^{7 ^ {m- 1}3n}) ^{ 7}- 1 =}\\
\ textc ol o r[r gb]{0.00,0. 00,0 .50}{=(2^ {7^{m- 1 }3n}-1) \ cdot\su m_{i=0 }^6 (2^ {7^{ m -1 }3 n}) ^{ i }\e q ui v0( m od\ ,7^{m}\c do t7 )\equ iv0( m o d \ ,7^{ m+1 }).} \end{ gathered}$$
Num bers $L_ {2}(n ) = 2 ^{2n }+ 2^{ n} - 1 $
=== == =============== ==== ========= === == =
Fi rst f i ve pri menum bers $L _{2}(1) = 5$, $ L _ { 2} (2)=19$, $L_{2}(3) =7 1 $ ,\
$L_{2}( 4)=271 $ ,$L _ {2}(6)=4 15 9$. Fro m **the sta t em ent 1**validi t yof the\
c om pariso ns fo llo ws $$ L _{2} (4k+1) \equiv0( mod\, 5 ),\ L_{2}(10k + 7)\equiv L_{2 } (1 0 k +8 ) \equ iv0 (mod\,11),$ $ wh e re $ k\ge q 0$ is integ er.
Th e n u mbers $L_{2}(n)$ ar enot fr ee fr om squares $L _{2}(68)\e q u i v 0(mod\ ,11^ { 3} ) $,\
$L_{2}(97) \equi v 0(mod\,1 1 ^{2})$.
Theauthor h as checke d up the f oll owi ngwor t h yto attentionf a ctsfo r numbe rs$L_{2}( n)$ .
1\) Fo rprime num bers $p\ le q5 00 3$ th e | \equiv0(mod\,7)$.
Let $A=B\cdot_L_{1}(7^{k}n)=2^{7^{k}3n}-1$. Let’s_prove by induction on_$k\geq0$, that_$$\textcolor[rgb]{0.00,0.00,0.50}{2^{7^{k}3n}-1\equiv0(mod\,7^{k+1}).}\label{T3:1}$$_The case_$k=0$_is obvious. Let’s_make the inductive_assumption, that for $k\leq_m-1$ the comparison_is_fair. Let’s consider expression $$\begin{gathered}
\notag \textcolor[rgb]{0.00,0.00,0.50}{2^{7^{m}3n}-1=(2^{7^{m-1}3n})^{7}-1=}\\
\textcolor[rgb]{0.00,0.00,0.50}{=(2^{7^{m-1}3n}-1)\cdot\sum_{i=0}^6(2^{7^{m-1}3n})^{i}\equiv0(mod\,7^{m}\cdot7)\equiv0(mod\,7^{m+1}).}\end{gathered}$$
Numbers $L_{2}(n) = 2^{2n} + 2^{n} - 1$
=======================================
First five_prime_numbers $L_{2}(1)=5$,_$L_{2}(2)=19$,_$L_{2}(3)=71$,\
$L_{2}(4)=271$,_$L_{2}(6)=4159$. From **the statement 1**_validity of the\
comparisons follows $$L_{2}(4k+1)\equiv0(mod\,5),\_ L_{2}(10k+7)\equiv_L_{2}(10k+8)\equiv0(mod\,11),$$ where $k\geq0$ is integer.
The numbers $L_{2}(n)$ are_not_free from squares_$L_{2}(68)\equiv 0(mod\,11^{3})$,\
$L_{2}(97)\equiv 0(mod\,11^{2})$.
The author has checked up the following_worthy to attention facts for numbers_$L_{2}(n)$.
1\) For prime_numbers_$p\leq5003$_the |
sections [@ellis2]; although more accurate determinations of the scalar matrix elements discussed further above constitute the most urgent issue in reducing hadronic uncertainties in dark matter searches, $\Delta s$ also plays a significant role in that context.
A number of lattice QCD investigations of strange quark degrees of freedom in the nucleon have recently been undertaken[@bali1; @bali2; @bali3; @jlqcd1; @jlqcd2; @jlqcd3; @freeman1; @freeman2; @kfliu1; @kfliu2; @kfliu3; @duerr; @drach; @qcdsf; @babich; @young1; @young2], the majority of which have focused specifically on the scalar content. Studies of the latter have proceeded via two avenues: On the one hand, one can directly determine the matrix element $\langle N|\bar{s} s|N\rangle $ via the appropriate disconnected three-point function; this methodology was adopted in[^1] [@bali1; @bali3; @jlqcd2; @jlqcd3; @freeman1; @freeman2; @kfliu2; @drach; @babich] and also in the present work, as described in detail further below. A study of techniques suited to improve the efficiency of this approach has been presented in [@cyprus]. On the other hand, a somewhat less direct inference of the scalar strange quark content of the nucleon is possible via the study of the baryon spectrum, which is related via the Feynman-Hellmann theorem $$\langle B | \bar{q} q | B\rangle = \frac{\partial m_B }{\partial m_q }$$ to the corresponding sigma terms for the baryon state $|B\rangle $ and quark flavor $q$. This avenue has been pursued in [@jlqcd1; @jlqcd3; @duerr; @qcdsf], and a related methodology, combining lattice hadron spectrum data with chiral perturbation theory, was pursued in [@young1; @young2].
The characteristics of these various investigations of the scalar strange quark content of the nucleon are diverse. They include $N_f =2$ calculations, in which the strange quark degrees of freedom are quenched [@bali3; @jlqcd1; @jlqcd2; @babich], but also $N_ | sections [ @ellis2 ]; although more accurate determinations of the scalar matrix elements discussed far above establish the most urgent issue in reduce hadronic uncertainties in dark topic searches, $ \Delta s$ also plays a meaning function in that context.
A number of wicket QCD investigations of strange quark degrees of freedom in the nucleon have recently been undertaken[@bali1; @bali2; @bali3; @jlqcd1; @jlqcd2; @jlqcd3; @freeman1; @freeman2; @kfliu1; @kfliu2; @kfliu3; @duerr; @drach; @qcdsf; @babich; @young1; @young2 ], the majority of which have focus specifically on the scalar content. report of the latter have proceeded via two avenues: On the one hired hand, one can directly determine the matrix element $ \langle N|\bar{s } s|N\rangle $ via the appropriate confused three - point function; this methodology was adopted in[^1 ] [ @bali1; @bali3; @jlqcd2; @jlqcd3; @freeman1; @freeman2; @kfliu2; @drach; @babich ] and besides in the present work, as identify in detail further below. A report of techniques suited to improve the efficiency of this approach has been presented in [ @cyprus ]. On the early hired hand, a somewhat less direct inference of the scalar strange quark content of the nucleon is possible via the study of the baryon spectrum, which is related via the Feynman - Hellmann theorem $ $ \langle B | \bar{q } q | B\rangle = \frac{\partial m_B } { \partial m_q } $ $ to the corresponding sigma terms for the baryon state of matter $ |B\rangle $ and quark flavor $ q$. This avenue has been pursued in [ @jlqcd1; @jlqcd3; @duerr; @qcdsf ], and a related methodology, blend lattice hadron spectrum data with chiral perturbation theory, was quest for in [ @young1; @young2 ].
The characteristics of these various investigations of the scalar strange quark contentedness of the nucleon are diverse. They include $ N_f = 2 $ calculations, in which the strange quark degrees of freedom are quench [ @bali3; @jlqcd1; @jlqcd2; @babich ], but also $ N _ | seftions [@ellis2]; although move accurate detetmunatiois of tge scalaf matrix elements discussed hurtyer avove constitute the mort urgent issue ib revucing hadronic nhcertaikcies jk darn matter searchgs, $\Delta s$ anso plays a sicnkfncant role in that context.
A number os lattive QCD investigajions jf sfgakge quark degrees of freedom in fhe nucneon have recrntly been undertaken[@bali1; @hali2; @bali3; @jlqcd1; @jlqcd2; @ulqcd3; @freemqn1; @fwweman2; @kfliu1; @yfliu2; @kflil3; @duerr; @dradh; @qcdsf; @babich; @young1; @young2], the majoxity of whixh hagg focused spxcificwlly on the scalar cmntent. Xtudies of the lavter have proceeded via txo avenues: On the ong hand, one ccn directly determine tye majrix alemdbt $\uanflx N|\gar{s} s|J\raigle $ via tge appropriqte disconnected thteq-ipint functioh; this mqthodology was adopted in[^1] [@bali1; @bali3; @jlqbd2; @jmqcd3; @freeman1; @freeman2; @kfoiu2; @drach; @babich] and wlso in tre present work, as described in detail further benow. A rtubn ud hechniques suited to improve the efficiency os tnix approach has been presemtfd yn [@cyprus]. On jhe othzd gand, a somewhat leds direst inderence os thr scalar strange quark contwnt of the nlclein is possible via the study uf tne batyon spectrum, which is xelates via the Ffynman-Helmoann theorem $$\lanele B | \bar{q} q | B\rangle = \frac{\pareial m_B }{\pertiak m_q }$$ tu thg correfponding slgma barms for the baryoj staje $|B\ratgle $ and euark flavor $q$. This avenue has uxen pursued im [@blqbd1; @jlqcd3; @buerr; @acdsf], and a rqlated methodokogy, coibinivg lattice hadron spectrum dwta with chirdp perturbatimn theorr, waw puesued iv [@young1; @young2].
Thr charactvrnstics of these various invcstigxfions of the sccuqr strange quarl cuntqnn oh the tucleon are givefse. Yhey knclude $N_f =2$ zalcilations, in which tha stdange quark degreex jf freedim are qtenched [@bali3; @klqcd1; @jlqcd2; @babich], but elso $N_ | sections [@ellis2]; although more accurate determinations of matrix discussed further constitute the most uncertainties dark matter searches, s$ also plays significant role in that context. A of lattice QCD investigations of strange quark degrees of freedom in the nucleon recently been undertaken[@bali1; @bali2; @bali3; @jlqcd1; @jlqcd2; @jlqcd3; @freeman1; @freeman2; @kfliu1; @kfliu2; @kfliu3; @drach; @babich; @young2], majority of which have focused specifically on the scalar content. Studies of the latter have proceeded two avenues: On the one hand, one can determine the matrix element N|\bar{s} s|N\rangle $ via the disconnected function; this was in[^1] @bali3; @jlqcd2; @jlqcd3; @freeman2; @kfliu2; @drach; @babich] and also in the present work, as described in detail further below. A of techniques improve the of approach been presented in the other hand, a somewhat less the scalar strange quark content of the nucleon possible via study of the baryon spectrum, which related via the Feynman-Hellmann theorem $$\langle B | q | B\rangle = \frac{\partial m_B }{\partial m_q }$$ to the corresponding sigma terms for state $|B\rangle $ and flavor $q$. This has pursued [@jlqcd1; @duerr; @qcdsf], a related methodology, combining lattice hadron spectrum data with chiral perturbation was pursued in [@young1; @young2]. The characteristics of these various the strange quark content the nucleon are diverse. include =2$ calculations, in which quark of [@bali3; @jlqcd2; but also $N_ | sections [@ellis2]; although more Accurate deTermiNatIonS oF the ScalAr matrix elemenTS disCussed further above consTitutE tHE mosT UrGent iSsue in rEDuCINg hAdRoNic UnCErTaintIes In dark mAtter searcHes, $\deLta s$ also playS A sIgnificant RolE in that conteXt.
A Number Of LatTIce QCd inVestiGationS Of straNge quark dEgREes of fREedom in THE nUcleOn have recently beeN UnDErtaken[@bali1; @balI2; @bali3; @jLqCD1; @jLQCd2; @jLqcD3; @freeman1; @frEeMan2; @kfLIu1; @kfliu2; @KFlIU3; @DUerR; @Drach; @qcdsf; @babIch; @young1; @youNG2], thE majorItY of WHich haVe focUsED spEcifically oN the Scalar conTent. StUDies of tHE latter Have prOceEdeD via TWo AvEnuEs: oN thE OnE haND, onE can direCtLy DeterMine THE MAtriX elEmenT $\langLe N|\bar{s} s|N\rangLe $ vIa thE AppRopriAte diSconNeCted tHree-poInt fuNcTion; this methodoLogy Was adopteD in[^1] [@BaLi1; @bAlI3; @jlqcD2; @Jlqcd3; @fReeMan1; @Freeman2; @Kfliu2; @drACh; @bAbICH] AnD also in the present wOrK, AS dEscribed In detaIL fUrTHer below. a sTudY of tECHniquEs suITeD to improVe the eFFiCiEncy of tHiS approAcH haS beEn preSEnteD in [@cypRus]. On the Other HAnd, a somewhat leSS direct infereNCe OF ThE ScalAr sTrange quark ContENt of The nUClEon IS possIble vIa THe STudy of the baryon specTrUm, whicH is reLated via the FeYnman-HellmANN Theorem $$\lAnglE b | \bAR{q} q | B\rangle = \frac{\PartiAl m_B }{\partiaL M_q }$$ to the cOrresPonding sIgma terms FOR the baryOn sTatE $|B\rAngLE $ AnD quark flavor $q$. tHIs avEnUe has beEn pUrsued iN [@jlQcd1; @JlqCd3; @dUeRr; @qcdsf], anD a relateD mEtHoDoLogY, combINing lattIcE haDrOn sPectrUM data wIth chIral PeRtURbaTion theORy, WAS purSuEd In [@yoUng1; @YoUng2].
ThE chaRActEristicS of these vAriOUs inVeStIgationS of the scalar sTrAnge quark cOnTenT of the NUCleon are Diverse. They include $N_f =2$ calCUlationS, in Which The sTrange quaRk dEgrees Of fREedom aRe quenChed [@bAlI3; @jlQCD1; @jlqcD2; @BAbIch], BuT also $N_ | sections [@ellis2]; altho ugh more a ccura tedet er mina tion s of the scala r mat rix elements discussed furt he r abo v econst itute t h em o stur ge ntis s ue in r edu cing ha dronic unc ert ai nties in dar k m atter sear che s, $\Delta s $ a lso pl ay s a signi fic ant r ole in that c ontext.
An umbero f latti c e Q CD i nvestigations ofs tr a nge quark degr ees of f r ee d o m i n t he nucleon h ave r e centlyb ee n u nde r taken[@bali1; @bali2; @b a li3 ; @jlq cd 1;@ jlqcd2 ; @jl qc d 3;@freeman1;@fre eman2; @k fliu1; @kfliu2 ; @kfliu 3; @du err ; @ drac h ;@q cds f; @ba b ic h;@ you ng1; @yo un g2 ], th e ma j o r i ty o f w hich have focused spec ifi call y on thescala r co nt ent.Studie s ofth e latter have p roce eded viatwo a ven ue s: On the on e h and , one c an dire c tly d e t e rm ine the matrix ele me n t $ \langleN|\bar { s} s | N\rangle $ vi a th e appro pria t edisconne cted t h re e- point f un ction; t his me thodo l ogywas ad opted in [^1][ @bali1; @bali3 ; @jlqcd2; @jl q cd 3 ; @ f reem an1 ; @freeman2 ; @k f liu2 ; @d r ac h;@ babic h] an da ls o in the present wor k, as de scrib ed in detailfurther be l o w . A stud y of te c hniques suited to i mprove the efficien cy of this ap proach ha s been pre sen ted in [@ c y pr us]. On the o t h er h an d, a so mew hat les s d ire ctinf er ence of t he scala rst ra ng e q uarkc ontent o fthe n ucl eon i s possi ble v ia t he s t udy of the ba r y on s pe ct rum, wh ic h isrela t edvia the Feynman- Hel l mann t he orem $$ \langle B | \ ba r{q} q | B \r ang le = \ f r ac{\part ial m_B }{\partial m_q} $$ to t hecorre spon ding sigm a t erms f ort he bar yon st ate $ |B \ra n g le $a n dqua rk flavor $q $ . Th is av en ue h as been pursued in [@jlqc d 1;@jlqcd3; @due rr; @qc d s f] , a n da re la t edm e thodology, comb ining latt ic e h adron spec t rum d ata wit h chira l per t urbatio n theory, was purs ue d in [ @yo ung1; @you ng2].
T he charac t erist i cs of t hes e vari ou s i nvest igatio n s o f the scala rstrang e qua rk content of the nucleon are div erse.Theyinc lude $N_f =2 $ ca lculation s, i n which th e s tra nge q uar k degr eeso ffre e dom a re q u enched [@ b al i3; @ jl qcd1; @jlqc d 2 ; @b abich ],b ut als o $N _ | sections_[@ellis2]; although_more accurate determinations of_the scalar_matrix_elements discussed_further_above constitute the_most urgent issue_in reducing hadronic uncertainties_in dark matter_searches,_$\Delta s$ also plays a significant role in that context.
A number of lattice QCD_investigations_of strange_quark_degrees_of freedom in the nucleon_have recently been undertaken[@bali1; @bali2;_@bali3; @jlqcd1;_@jlqcd2; @jlqcd3; @freeman1; @freeman2; @kfliu1; @kfliu2; @kfliu3; @duerr;_@drach;_@qcdsf; @babich; @young1;_@young2], the majority of which have focused specifically on_the scalar content. Studies of the_latter have proceeded_via_two_avenues: On the one_hand, one can directly determine the_matrix element $\langle N|\bar{s} s|N\rangle $_via the appropriate disconnected three-point function; this_methodology was adopted in[^1] [@bali1; @bali3;_@jlqcd2; @jlqcd3; @freeman1; @freeman2; @kfliu2;_@drach; @babich]_and also in the present_work, as described_in detail_further below. A_study of techniques suited to improve_the efficiency of_this approach has been presented in_[@cyprus]._On the other_hand,_a_somewhat less_direct inference of_the_scalar strange_quark_content of the nucleon is possible_via_the study of the baryon spectrum, which_is related via the_Feynman-Hellmann_theorem $$\langle B |_\bar{q} q | B\rangle =_\frac{\partial m_B }{\partial m_q }$$ to_the corresponding_sigma terms_for the baryon state $|B\rangle $ and quark flavor $q$. This_avenue has been pursued in [@jlqcd1;_@jlqcd3; @duerr; @qcdsf], and_a related_methodology,_combining lattice hadron_spectrum_data with_chiral perturbation theory, was pursued in [@young1;_@young2].
The characteristics_of these various investigations of the_scalar strange quark content_of_the nucleon are diverse. They include_$N_f =2$ calculations, in which the_strange quark degrees of freedom_are_quenched_[@bali3; @jlqcd1; @jlqcd2; @babich], but_also $N_ |
.
The proposal that local interstellar gas may be deficient in $r$-process products relative to Solar System material is curiously supported by interstellar abundances of other lightly depleted or undepleted elements with a $r$-process contribution. We emphasize directions probing material toward Sco OB2, which includes $\rho$ Oph A. For Solar System material, about 50% of Kr, 50% of Cd, and 30% of Sn are attributable to the $r$-process (Beer et al. 1992; Raiteri et al. 1993; Arlandini et al. 1999). The interstellar Kr abundance within a few hundred parsecs of the Sun is constant at 60% of the solar-wind value (Cartledge et al. 2001, 2003). Though there are fewer determinations of the interstellar Cd abundance, it is constant at 80% of the Solar System value (Sofia, Meyer, & Cardelli 1999). Sofia et al. find a Sn abundance that is essentially solar, if not slightly greater than solar, for sight lines showing little depletion in other elements. Arsenic, which is mostly $r$-process (Beer et al. 1992; Raiteri et al. 1993) and has only 1 isotope, is seen currently only in three reddened directions (Cardelli et al. 1993; Federman et al. 2003). It has a higher condensation temperature (see Cardelli et al. 1993) and thus it has more depletion, especially in reddened directions. Measurements of As [II]{} in lightly reddened sight lines, where little depletion is expected, appear necessary to confirm our hypothesis.
Variations in space and time of the mix of $s$- and $r$-process products in the Galaxy’s interstellar medium should be reflected in the abundances of the heavy elements in unevolved stars. Relative abundances of elements such as Sr, Y, and Zr and heavier elements such as Ba, Ce, and Eu for dwarfs of the thin disk do not vary from star to star by more than the measurement errors (Edvardsson et al. 1993; Chen et al. 2000; Reddy et al. 2003). These measurements would seem to exclude the larger variations expected for stars formed from gas with a large amplitude in the ratio of $s$- to $r$-process products. Our interstellar results suggest that differences in $n$-capture nucleosynthesis may be local to the solar neighborhood.
Concluding Remarks
==================
Our estimate of the Rb | .
The proposal that local interstellar gas may be deficient in $ r$-process products proportional to Solar System fabric is curiously defend by interstellar abundances of early lightly consume or undepleted elements with a $ r$-process contribution. We emphasize management probing material toward Sco OB2, which include $ \rho$ Oph A. For Solar System material, about 50% of Kr, 50% of Cd, and 30% of Sn are attributable to the $ r$-process (Beer et al. 1992; Raiteri et al. 1993; Arlandini et al. 1999). The interstellar Kr abundance within a few hundred parsec of the Sun is constant at 60% of the solar - wind value (Cartledge et al. 2001, 2003). Though there are fewer determination of the interstellar Cd abundance, it is constant at 80% of the Solar System value (Sofia, Meyer, & Cardelli 1999). Sofia et al. witness a Sn abundance that is essentially solar, if not slightly great than solar, for sight lines showing short depletion in other elements. Arsenic, which is mostly $ r$-process (Beer et al. 1992; Raiteri et al. 1993) and have only 1 isotope, is seen currently only in three reddened direction (Cardelli et al. 1993; Federman et al. 2003). It has a higher condensation temperature (see Cardelli et al. 1993) and thus it hold more depletion, especially in reddened directions. Measurements of As [ II ] { } in lightly reddened sight credit line, where little depletion is expected, look necessary to confirm our guess.
Variations in space and time of the mix of $ s$- and $ r$-process products in the Galaxy ’s interstellar culture medium should be reflected in the abundances of the heavy elements in unevolved asterisk. Relative abundances of elements such as Sr, Y, and Zr and heavier elements such as Ba, Ce, and Eu for dwarfs of the sparse disk do not vary from star to star by more than the measurement errors (Edvardsson et al. 1993; Chen et al. 2000; Reddy et al. 2003). These measurement would seem to eject the large variations expected for stars formed from accelerator with a large amplitude in the ratio of $ s$- to $ r$-process products. Our interstellar result suggest that differences in $ n$-capture nucleosynthesis may be local to the solar neighborhood.
Concluding Remarks
= = = = = = = = = = = = = = = = = =
Our estimate of the Rb | .
The proposal that local inttrstellar gas may be defmcient jn $r$-procdss products relative to Soler Ststem material is curiously supportef by intwrsttllar abundances of other lightmn depnxted or undeplejed elements with a $r$-procevs clntribution. We emphasize directions probinb laterial towarq Scp OB2, shich includes $\rho$ Oph A. For Solar System material, aboit 50% of Kr, 50% of Cd, and 30% of Sj arf attributable to hhe $r$-procesw (Beqe et al. 1992; Raigeri et al. 1993; Arlandini gt al. 1999). The interstellar Kr abundxnce cithin a feq yunfted parsecs if thv Sun is consbsnt at 60% of thr solar-wind vakue (Caetledge et al. 2001, 2003). Thougi there are fewer dejerminatiots of the intersteloae Cd dbungancd, it is cpnatant wt 80% of the Somar System calue (Sofia, Meyer, & Vaweelli 1999). Sofia st al. sigd a Sn abundance that is essentially smlad, if not slightly greatwr than solar, for sigjt lines fhowing little depletion in other elements. Arseniw, whirh is mjrrlj $r$-process (Beer et al. 1992; Raiteri et al. 1993) and has knky 1 isotope, is scen currently only ij yrree reddened direccjohs (Cardelli et al. 1993; Federmwn et al. 2003). It hws a higher condensation temperqture (see Cagdeloi et al. 1993) and thus it has morz deplgtion, rspecially in reddened birectjons. Measurfments of Xs [II]{} in lightly fedceted sighu lines, where littje depletmon ix expecged, sppear necessary to confirm our hypothesls.
Varnatiots in spacf and time of the mix of $s$- and $c$-'rocess produvtv it the Gajaxy’s interstellar medium should be refjectea in the agundancxs of the hewvy elements hj unevolved vtars. Rejatice avundancdr of elements xuch as Sg, V, and Zr qnd heavier elemenbs suzg as Ba, Ce, and Zb dor dwarfs of tne ghig fixk go not vary xrom stxt to rtar by morc tfan yhe measurement erross (Esvardsson et al. 1993; Cnek et al. 2000; Teddy et wl. 2003). These meaxurements would setm to xxcludx the karder variations expected for stzrs formef fvom gas with w lavge wmplitude nn the ratio of $s$- to $r$-process products. Onr interstellar results suggest that diffetenges in $n$-captnre nusleosynthasis may be local to the solar neighbprhood.
Concluding Remadks
==================
Our estilate of the Rb | . The proposal that local interstellar gas deficient $r$-process products to Solar System interstellar of other lightly or undepleted elements a $r$-process contribution. We emphasize directions material toward Sco OB2, which includes $\rho$ Oph A. For Solar System material, 50% of Kr, 50% of Cd, and 30% of Sn are attributable to $r$-process et 1992; et al. 1993; Arlandini et al. 1999). The interstellar Kr abundance within a few hundred parsecs the Sun is constant at 60% of the value (Cartledge et al. 2003). Though there are fewer of interstellar Cd it constant 80% of the System value (Sofia, Meyer, & Cardelli 1999). Sofia et al. find a Sn abundance that is essentially if not than solar, sight showing depletion in other which is mostly $r$-process (Beer et et al. 1993) and has only 1 isotope, seen currently in three reddened directions (Cardelli et 1993; Federman et al. 2003). It has a condensation temperature (see Cardelli et al. 1993) and thus it has more depletion, especially in Measurements of As [II]{} lightly reddened sight where depletion expected, necessary to our hypothesis. Variations in space and time of the mix of and $r$-process products in the Galaxy’s interstellar medium should be the of the heavy in unevolved stars. Relative of such as Sr, Y, and elements Ce, Eu dwarfs of the thin do not vary from star star by more than al. 1993; Chen et al. 2000; Reddy et 2003). These measurements would seem to exclude larger variations expected for stars formed from gas with a large amplitude the ratio to $r$-process products. Our interstellar results suggest that in $n$-capture nucleosynthesis may local to the solar neighborhood. Concluding Remarks ================== Our of Rb | .
The proposal that local interStellar gas May be DefIciEnT in $r$-ProcEss products relATive To Solar System material iS curiOuSLy suPPoRted bY intersTElLAR abUnDaNceS oF OtHer liGhtLy depleTed or undepLetEd Elements with A $R$-pRocess contRibUtion. We emphaSizE direcTiOns PRobinG maTeriaL towarD sco OB2, wHich incluDeS $\Rho$ Oph a. for SolaR sYsTem mAterial, about 50% of Kr, 50% oF cd, ANd 30% of Sn are attriButablE tO ThE $R$-ProCesS (Beer et al. 1992; RAiTeri eT Al. 1993; ArlanDInI ET Al. 1999). THE interstellar kr abundance WIthIn a few HuNdrED parseCs of tHe sUn iS constant at 60% Of thE solar-winD value (cArtledgE Et al. 2001, 2003). ThoUgh theRe aRe fEwer DEtErMinAtIOns OF tHe iNTerStellar CD aBuNdancE, it iS CONStanT at 80% Of thE SolaR System value (SOfiA, MeyER, & CaRdellI 1999). SofiA et aL. fInd a SN abundAnce tHaT is essentially sOlar, If not sligHtlY gReaTeR than SOlar, foR siGht Lines shOwing liTTle DePLETiOn in other elements. ARsENIc, Which is mOstly $r$-PRoCeSS (Beer et aL. 1992; RAitEri eT AL. 1993) and hAs onLY 1 iSotope, is Seen cuRReNtLy only iN tHree reDdEneD diRectiONs (CaRdelli Et al. 1993; FedeRman eT Al. 2003). It has a higher COndensation teMPeRATuRE (see carDelli et al. 1993) anD thuS It haS morE DePleTIon, esPeciaLlY In REddened directions. MeAsUremenTs of AS [II]{} in lightly rEddened sigHT LInes, wherE litTLe DEpletion is expeCted, aPpear necesSAry to conFirm oUr hypothEsis.
VariaTIOns in spaCe aNd tIme Of tHE MiX of $s$- and $r$-proceSS ProdUcTs in the galAxy’s intErsTelLar MedIuM should be ReflecteD iN tHe AbUndAnces OF the heavY eLemEnTs iN unevOLved stArs. ReLatiVe AbUNdaNces of eLEmENTs suCh As sr, Y, aNd ZR aNd heaVier ELemEnts sucH as Ba, Ce, anD Eu FOr dwArFs Of the thIn disk do not vaRy From star to StAr bY more tHAN the measUrement errors (Edvardsson ET al. 1993; Chen Et aL. 2000; ReddY et aL. 2003). These meaSurEments WouLD seem tO excluDe the LaRgeR VAriatIONs ExpEcTed for starS FOrmEd froM gAs wiTh a largE amplitude in the ratIO of $S$- to $r$-process prOduCts. OUR InTerSTeLLar ReSUltS SUggest that diffeRences in $n$-cApTUrE nucleosynTHesIs May be loCal to thE solaR NeighboRhood.
ConcLuding RemArKs
==================
OuR EStiMate of the RB | .
The proposal that local interstel lar g asmay b e de fici ent in $r$-pro c essproducts relative to S olarSy s temm at erial is cur i ou s l y s up po rte db yinter ste llar ab undances o f o th er lightly d e pl eted or un dep leted elemen tswith a $ r$- p roces s c ontri bution . We em phasize d ir e ctions probing m at eria l toward Sco OB2, wh i ch includes $\ rho$ O ph A. F orSol ar Systemma teria l , about 50 % o f K r , 50% of Cd,and 30% ofS n a re att ri but a ble to the$r $ -pr ocess (Beer etal. 1992; Raite r i et al . 1993;Arland ini et al. 19 99 ).Th e in t er ste l lar Kr abun da nc e wit hina f e w hu ndr ed p arsec s of the Suniscons t ant at 6 0% of the s olar- wind v alue(C artledge et al. 200 1, 2003). Th ou ghth ere a r e fewe r d ete rminati ons oft hein t e r st ellar Cd abundance ,i t i s consta nt at8 0% o f the Sol ar Sy stem v alue(Sof i a, Meyer,& Card e ll i1999).So fia et a l.fin d a S n abu ndance that is esse n tially solar,i f not slightl y g r e at e r th ansolar, forsigh t lin es s h ow ing littl e dep le t io n in other elements. A rsenic , whi ch is mostly$r$-proces s ( Beer etal.1 99 2 ; Raiteri et a l. 19 93) and ha s only 1isoto pe, is s een curre n t ly onlyinthr eered d e ne d directions( C arde ll i et al . 1 993; Fe der man et al .2003). It has a h ig he rco nde nsati o n temper at ure ( see Card e lli et al.1993 )an d th us it h a sm o re d ep le tion , e sp ecial ly i n re ddeneddirection s.M easu re me nts ofAs [II]{} inli ghtly redd en edsightl i nes, whe re little depletion ise xpected , a ppear nec essary to co nfirmour hypoth esis.
Vari at ion s in sp a c eand t ime of the m ixof $s $- and $r$-pr ocess products int heGalaxy’s inte rst ella r me diu m s h oul db e r e f lected in the a bundancesof th e heavy el e men ts in une volvedstars . Relati ve abunda nces of e le ment s suc h as Sr, Y , and Zr and heav i er el e me nts s uch as Ba ,Ce, andEu for dwa rfs o f theth in dis k dono t vary f rom star to star by mor e than themea surementerr o rs(Edvardss on e t al. 1993 ; C hen et a l.2 000;Redd y e t a l . 200 3).T hese meas u re men t s w ould seem t o e xcl ude t hel argervari ations expected f o r stars formed fro m gas wi t h ala rge amplitudeinth e ratio of $ s$- to $r$- processpr o ducts . Ourinters tellarr e su l ts sug gest th at differ enc es in $n$- ca pt u re nuc leos yn thesis may b e loc a l to the solar ne ighbo r h ood.Con cludi ng Remark s
=== ========== =====
Ourestima te o f the Rb | .
The proposal_that local_interstellar gas may be_deficient in_$r$-process_products relative_to_Solar System material_is curiously supported_by interstellar abundances of_other lightly depleted_or_undepleted elements with a $r$-process contribution. We emphasize directions probing material toward Sco OB2,_which_includes $\rho$_Oph_A._For Solar System material, about_50% of Kr, 50% of_Cd, and_30% of Sn are attributable to the $r$-process_(Beer_et al. 1992;_Raiteri et al. 1993; Arlandini et al. 1999). The_interstellar Kr abundance within a few_hundred parsecs of_the_Sun_is constant at 60%_of the solar-wind value (Cartledge et_al. 2001, 2003). Though there are_fewer determinations of the interstellar Cd abundance,_it is constant at 80% of_the Solar System value (Sofia,_Meyer, &_Cardelli 1999). Sofia et al._find a Sn_abundance that_is essentially solar,_if not slightly greater than solar,_for sight lines_showing little depletion in other elements._Arsenic,_which is mostly_$r$-process_(Beer_et al._1992; Raiteri et_al._1993) and_has_only 1 isotope, is seen currently_only_in three reddened directions (Cardelli et al._1993; Federman et al._2003)._It has a higher_condensation temperature (see Cardelli et_al. 1993) and thus it has_more depletion,_especially in_reddened directions. Measurements of As [II]{} in lightly reddened sight lines, where_little depletion is expected, appear necessary_to confirm our hypothesis.
Variations_in space_and_time of the_mix_of $s$-_and $r$-process products in the Galaxy’s interstellar_medium should_be reflected in the abundances of_the heavy elements in_unevolved_stars. Relative abundances of elements such_as Sr, Y, and Zr and_heavier elements such as Ba,_Ce,_and_Eu for dwarfs of the_thin disk do not vary from_star to star_by more than the measurement errors (Edvardsson_et_al. 1993; Chen et al. 2000;_Reddy_et al. 2003). These measurements would_seem_to_exclude the larger variations expected_for stars formed from gas with_a large amplitude in the ratio of $s$- to_$r$-process products. Our_interstellar results suggest that differences_in_$n$-capture_nucleosynthesis may be local to the solar neighborhood.
Concluding Remarks
==================
Our estimate_of the_Rb |
{:=}}y\in Y_0$ for $y\in Y_0\setminus Y_\cap$ and $\phi_*(y){\mathrel{:=}}y\in Y_1$ for $y\in Y_1\setminus Y_\cap$. Notice that this is a continuous, $G$[-]{}equivariant map. Thus $Z\in\mathcal{S}$ implies $Y\in\mathcal{S}$.
A cycle for ${\mathrm{RKK}}^{G{\mathbin{\ltimes}}X}(Z;A,A)$ is a triple $(f_0,f_1,f_\cap)$, where $f_j$ is a cycle for ${\mathrm{RKK}}^{G{\mathbin{\ltimes}}X} (Y_j;A,A)$ for $j=0,1$ and $f_\cap$ is a homotopy in ${\mathrm{RKK}}^{G{\mathbin{\ltimes}}X}(Y_\cap;A,A)$ between $f_0|_{Y_\cap}$ and $f_1|_{Y_\cap}$. Fix any such cycle. The cycles $f_0$ and $f_1$ are homotopic to $0$ because $Y_0,Y_1\in\mathcal{S}$ and $A\in{{\mathcal{CC}}}$. This yields a homotopy between $(f_0,f_1,f_\cap)$ and $(0,0,f_\cap')$, where $f_\cap'$ is some cycle for ${\mathrm{RKK}}^{G{\mathbin{\ltimes}}X}(Y_\cap\times [0,1];A,A)$ whose restrictions to $0$ and $1$ vanish. Thus $f_\cap'$ is equivalent to a cycle for $${\mathrm{RKK}}^{G{\mathbin{\ltimes}}X\times [0,1]}(Y_\cap;C([0,1],A),\Sigma A)
\cong {\mathrm{RKK}}^{G{\mathbin{\ltimes}}X}(Y_\cap;A,\Sigma A).$$ Apply to the coordinate projection $X\times Y_\cap \times [0,1]\to X\times Y_\cap$ to get this isomorphism. We have ${\mathrm{RKK}}^{G{\mathbin{\ltimes}}X}(Y_\cap;A,\Sigma A)=0$ | {: = } } y\in Y_0 $ for $ y\in Y_0\setminus Y_\cap$ and $ \phi_*(y){\mathrel{:=}}y\in Y_1 $ for $ y\in Y_1\setminus Y_\cap$. Notice that this is a continuous, $ G$[-]{}equivariant map. Thus $ Z\in\mathcal{S}$ implies $ Y\in\mathcal{S}$.
A cycle for $ { \mathrm{RKK}}^{G{\mathbin{\ltimes}}X}(Z;A, A)$ is a ternary $ (f_0,f_1,f_\cap)$, where $ f_j$ is a hertz for $ { \mathrm{RKK}}^{G{\mathbin{\ltimes}}X } (Y_j;A, A)$ for $ j=0,1 $ and $ f_\cap$ is a homotopy in $ { \mathrm{RKK}}^{G{\mathbin{\ltimes}}X}(Y_\cap;A, A)$ between $ f_0|_{Y_\cap}$ and $ f_1|_{Y_\cap}$. Fix any such cycle. The cycles $ f_0 $ and $ f_1 $ are homotopic to $ 0 $ because $ Y_0,Y_1\in\mathcal{S}$ and $ A\in{{\mathcal{CC}}}$. This give way a homotopy between $ (f_0,f_1,f_\cap)$ and $ (0,0,f_\cap')$, where $ f_\cap'$ is some cycle for $ { \mathrm{RKK}}^{G{\mathbin{\ltimes}}X}(Y_\cap\times [ 0,1];A, A)$ whose restrictions to $ 0 $ and $ 1 $ vanish. therefore $ f_\cap'$ is equivalent to a cycle for $ $ { \mathrm{RKK}}^{G{\mathbin{\ltimes}}X\times [ 0,1]}(Y_\cap;C([0,1],A),\Sigma A)
\cong { \mathrm{RKK}}^{G{\mathbin{\ltimes}}X}(Y_\cap;A,\Sigma A).$$ Apply to the coordinate project $ X\times Y_\cap \times [ 0,1]\to X\times Y_\cap$ to get this isomorphism. We have $ { \mathrm{RKK}}^{G{\mathbin{\ltimes}}X}(Y_\cap;A,\Sigma A)=0 $ | {:=}}y\in Y_0$ for $y\in Y_0\setminus Y_\cak$ and $\phi_*(y){\mathrel{:=}}i\ib Y_1$ foc $y\in Y_1\aetminus Y_\cap$. Notice that this is a rontunuouw, $G$[-]{}equivariant map. Thur $Z\in\mathbal{S}$ implues $B\in\mathcal{S}$.
A cycle for ${\mathrm{RIN}}^{G{\machuin{\ltimes}}X}(Z;A,A)$ ix a triple $(f_0,f_1,f_\cap)$, where $f_b$ ks a cycle for ${\mathrm{RKK}}^{G{\mathbin{\ltimef}}X} (Y_j;A,A)$ flr $j=0,1$ and $f_\cap$ if a njmotkiy in ${\mathrm{RKK}}^{G{\mathbin{\ltimes}}X}(Y_\dap;A,A)$ bttween $f_0|_{Y_\cap}$ and $f_1|_{U_\cap}$. Fix any such cycle. Thf cyfles $f_0$ and $f_1$ are holotopic to $0$ vecatwe $Y_0,Y_1\in\mathcxl{S}$ and $A\in{{\mathcal{CC}}}$. Tgis yields a homotopy between $(f_0,w_1,f_\cap)$ and $(0,0,f_\cap')$, qhwre $x_\cap'$ is somx cyclv for ${\mathrm{RKK}}^{G{\mathbhn{\ltimex}}X}(Y_\cap\times [0,1];A,A)$ whmse restrictions to $0$ and $1$ vanish. Thus $f_\cap'$ if equivalanc to a cycle for $${\mathem{EKK}}^{G{\mdthbhn{\ltknes}}B\tijex [0,1]}(G_\cap;C([0,1],A),\Digja A)
\cong {\mathrm{RKK}}^{G{\nathbin{\ltimes}}X}(Y_\cap;A,\Xidna A).$$ Apply to the cjowdinate projection $X\times Y_\cap \times [0,1]\to X\tjmes Y_\cap$ to get this iwomorphism. We have ${\majhrm{RKK}}^{G{\maehbin{\ltimes}}X}(Y_\cap;A,\Sigma A)=0$ | {:=}}y\in Y_0$ for $y\in Y_0\setminus Y_\cap$ and for Y_1\setminus Y_\cap$. that this is $Z\in\mathcal{S}$ $Y\in\mathcal{S}$. A cycle ${\mathrm{RKK}}^{G{\mathbin{\ltimes}}X}(Z;A,A)$ is a $(f_0,f_1,f_\cap)$, where $f_j$ is a cycle ${\mathrm{RKK}}^{G{\mathbin{\ltimes}}X} (Y_j;A,A)$ for $j=0,1$ and $f_\cap$ is a homotopy in ${\mathrm{RKK}}^{G{\mathbin{\ltimes}}X}(Y_\cap;A,A)$ between $f_0|_{Y_\cap}$ $f_1|_{Y_\cap}$. Fix any such cycle. The cycles $f_0$ and $f_1$ are homotopic to because and This a homotopy between $(f_0,f_1,f_\cap)$ and $(0,0,f_\cap')$, where $f_\cap'$ is some cycle for ${\mathrm{RKK}}^{G{\mathbin{\ltimes}}X}(Y_\cap\times [0,1];A,A)$ whose restrictions $0$ and $1$ vanish. Thus $f_\cap'$ is equivalent a cycle for $${\mathrm{RKK}}^{G{\mathbin{\ltimes}}X\times A) \cong {\mathrm{RKK}}^{G{\mathbin{\ltimes}}X}(Y_\cap;A,\Sigma A).$$ Apply the projection $X\times \times X\times to get this We have ${\mathrm{RKK}}^{G{\mathbin{\ltimes}}X}(Y_\cap;A,\Sigma A)=0$ | {:=}}y\in Y_0$ for $y\in Y_0\setminus Y_\cap$ anD $\phi_*(y){\mathrEl{:=}}y\in y_1$ foR $y\iN Y_1\SetmInus y_\cap$. Notice that THis iS a continuous, $G$[-]{}equivariaNt map. thUS $Z\in\MAtHcal{S}$ Implies $y\In\MAThcAl{s}$.
A CycLe FOr ${\MathrM{RKk}}^{G{\mathbIn{\ltimes}}X}(Z;a,A)$ iS a Triple $(f_0,f_1,f_\cap)$, WHeRe $f_j$ is a cycLe fOr ${\mathrm{RKK}}^{G{\MatHbin{\ltImEs}}X} (y_J;A,A)$ foR $j=0,1$ aNd $f_\caP$ is a hoMOtopy iN ${\mathrm{RKk}}^{G{\MAthbin{\LTimes}}X}(Y_\CAP;A,a)$ betWeen $f_0|_{Y_\cap}$ and $f_1|_{Y_\cap}$. fIx ANy such cycle. The Cycles $F_0$ aND $f_1$ ARE hoMotOpic to $0$ becaUsE $Y_0,Y_1\in\MAthcal{S}$ ANd $a\IN{{\MatHCal{CC}}}$. This yielDs a homotopy BEtwEen $(f_0,f_1,f_\CaP)$ anD $(0,0,F_\cap')$, whEre $f_\cAp'$ IS soMe cycle for ${\mAthrM{RKK}}^{G{\mathBin{\ltiMEs}}X}(Y_\cap\TImes [0,1];A,A)$ wHose reStrIctIons TO $0$ aNd $1$ VanIsH. thuS $F_\cAp'$ iS EquIvalent tO a CyCle foR $${\matHRM{rkK}}^{G{\mAthBin{\lTimes}}x\times [0,1]}(Y_\cap;C([0,1],A),\SIgmA A)
\coNG {\maThrm{RkK}}^{G{\maThbiN{\lTimes}}x}(Y_\cap;A,\sigma a).$$ APply to the coordiNate ProjectioN $X\tImEs Y_\CaP \timeS [0,1]\To X\timEs Y_\Cap$ To get thIs isomoRPhiSm. wE HAvE ${\mathrm{RKK}}^{G{\mathbin{\LtIMEs}}x}(Y_\cap;A,\SiGma A)=0$ | {:=}}y\in Y_0$ for $y\in Y _0\setminu s Y_\ cap $ a nd $\p hi_* (y){\mathrel{: = }}y\ in Y_1$ for $y\in Y_1\ setmi nu s Y_\ c ap $. No tice th a tt h isis a co nt i nu ous,$G$ [-]{}eq uivariantmap .Thus $Z\in\m a th cal{S}$ im pli es $Y\in\mat hca l{S}$.
A c y cle f or${\ma thrm{R K K}}^{G {\mathbin {\ l times} } X}(Z;A, A ) $is a triple $(f_0,f_1 , f_ \ cap)$, where $ f_j$ i sa c y c lefor ${\mathrm {R KK}}^ { G{\math b in { \ l tim e s}}X} (Y_j;A, A)$ for $j= 0 ,1$ and $ f_ \ca p $ is a homo to p y i n ${\mathrm {RKK }}^{G{\ma thbin{ \ ltimes} } X}(Y_\c ap;A,A )$bet ween $f _0 |_{ Y_ \ cap } $and $f_ 1|_{Y_\c ap }$ . Fix any s u c h cy cle . Th e cyc les $f_0$ and $f _1$a rehomot opicto $ 0$ beca use $Y _0,Y_ 1\ in\mathcal{S}$and$A\in{{\m ath ca l{C C} }}$.T his yi eld s a homoto py betw e en$( f _ 0 ,f _1,f_\cap)$ and $( 0, 0 , f_ \cap')$, where $f _\ c ap'$ isso mecycl e for $ {\ma t hr m{RKK}}^ {G{\ma t hb in {\ltime s} }X}(Y_ \c ap\ tim es [0 , 1];A ,A)$ w hose res trict i ons to $0$ and $1$ vanish. T h us $ f_ \ cap' $ i s equivalen t to a cy clef or $$ { \math rm{RK K} } ^{ G {\mathbin{\ltimes}} X\ times[0,1] }(Y_\cap;C([0 ,1],A),\Si g m a A)
\c ong{ \m a thrm{RKK}}^{G{ \math bin{\ltime s }}X}(Y_\ cap;A ,\SigmaA).$$ App l y to the co ord ina tep r oj ection $X\tim e s Y_\ ca p \time s [ 0,1]\to X\ tim esY_\ ca p$ to get this is om or ph is m.We ha v e ${\mat hr m{R KK }}^ {G{\m a thbin{ \ltim es}} X} (Y _ \ca p;A,\Si g ma A )=0$ | {:=}}y\in Y_0$_for $y\in_Y_0\setminus Y_\cap$ and $\phi_*(y){\mathrel{:=}}y\in_Y_1$ for_$y\in_Y_1\setminus Y_\cap$._Notice_that this is_a continuous, $G$[-]{}equivariant_map. Thus $Z\in\mathcal{S}$ implies_$Y\in\mathcal{S}$.
A cycle for_${\mathrm{RKK}}^{G{\mathbin{\ltimes}}X}(Z;A,A)$_is a triple $(f_0,f_1,f_\cap)$, where $f_j$ is a cycle for ${\mathrm{RKK}}^{G{\mathbin{\ltimes}}X} (Y_j;A,A)$ for $j=0,1$ and $f_\cap$_is_a homotopy_in_${\mathrm{RKK}}^{G{\mathbin{\ltimes}}X}(Y_\cap;A,A)$_between $f_0|_{Y_\cap}$ and $f_1|_{Y_\cap}$. Fix_any such cycle. The cycles_$f_0$ and $f_1$_are homotopic to $0$ because $Y_0,Y_1\in\mathcal{S}$ and $A\in{{\mathcal{CC}}}$. This_yields_a homotopy between_$(f_0,f_1,f_\cap)$ and $(0,0,f_\cap')$, where $f_\cap'$ is some cycle for_${\mathrm{RKK}}^{G{\mathbin{\ltimes}}X}(Y_\cap\times [0,1];A,A)$ whose restrictions to $0$_and $1$ vanish. Thus $f_\cap'$_is_equivalent_to a cycle for_$${\mathrm{RKK}}^{G{\mathbin{\ltimes}}X\times [0,1]}(Y_\cap;C([0,1],A),\Sigma A)
\cong {\mathrm{RKK}}^{G{\mathbin{\ltimes}}X}(Y_\cap;A,\Sigma_A).$$ Apply to the coordinate projection_$X\times Y_\cap \times [0,1]\to X\times Y_\cap$ to_get this isomorphism. We have ${\mathrm{RKK}}^{G{\mathbin{\ltimes}}X}(Y_\cap;A,\Sigma_A)=0$ |
$q_n$ are concerned, once again, by considering the event that the first and last rows of $M_n$ are equal (up to a sign), we see that $q_n \geq (2+o(1))^{-n}$. It is commonly believed that this lower bound is tight.
\[conjecture:prob-singularity\] We have $$q_n = (2+o(1))^{-n}.$$
In this paper, we obtain a much stronger upper bound on $q_n$, thereby making progress towards \[conjecture:prob-singularity\].
\[thm:main-thm\] There exists a natural number $n_0 \in \N$ such that for all $n\geq n_0$, $$q_n \leq 2^{-n^{1/4}\sqrt{\log{n}}/1000}.$$
Apart from providing a stronger conclusion, our proof of the above theorem is considerably shorter than previous works, and introduces and extends several novel combinatorial tools and ideas in discrete random matrix theory (some of which are based on joint work of the authors with Luh and Samotij [@FJLS2018]). We believe that these ideas allow for a unified approach to the singularity problem for many different discrete random matrix models, which have previously been handled in an ad-hoc manner. For completeness and for the convenience of the reader, we have included full proofs of all the simple background lemmas that we use from other papers, making this paper completely self contained.
Outline of the proof and comparison with previous work
------------------------------------------------------
In this subsection, we provide a very brief, and rather imprecise, outline of our proof, and compare it to previous works of Nguyen [@nguyen2012inverse] and Vershynin [@vershynin2014invertibility]; for further comparison with the work of Costello, Tao, and Vu, see [@nguyen2012inverse].
Let $x:=(x_1,\ldots,x_n)$ be the first row of $M_n$, let $M^{1}_{n-1}$ denote the bottom-right $(n-1)\times (n-1)$ submatrix of $M_n$, and for $2\leq i,j \leq n$, let $c_{ij}$ denote the cofactor of $M^{1}_{n-1}$ obtained by removing its $(i-1)^{st}$ row | $ q_n$ are concerned, once again, by considering the event that the first and final row of $ M_n$ are equal (up to a sign), we see that $ q_n \geq (2+o(1))^{-n}$. It is normally believed that this lower bound is tight.
\[conjecture: prob - singularity\ ] We hold $ $ q_n = (2+o(1))^{-n}.$$
In this paper, we obtain a a lot firm upper bind on $ q_n$, thereby making progress towards \[conjecture: prob - singularity\ ].
\[thm: main - thm\ ] There exist a natural number $ n_0 \in \N$ such that for all $ n\geq n_0 $, $ $ q_n \leq 2^{-n^{1/4}\sqrt{\log{n}}/1000}.$$
Apart from providing a stronger decision, our proof of the above theorem is considerably shorter than former works, and introduces and extends respective novel combinatorial instrument and ideas in discrete random matrix hypothesis (some of which are based on joint work of the authors with Luh and Samotij [ @FJLS2018 ]). We believe that these mind allow for a unified approach to the singularity problem for many different discrete random matrix model, which have previously been handled in an ad - hoc manner. For completeness and for the convenience of the reader, we have included full proofs of all the simple background lemmas that we use from early document, making this newspaper wholly self control.
Outline of the proof and comparison with previous work
------------------------------------------------------
In this subsection, we provide a very brief, and rather imprecise, outline of our proof, and compare it to former works of Nguyen [ @nguyen2012inverse ] and Vershynin [ @vershynin2014invertibility ]; for further comparison with the work of Costello, Tao, and Vu, see [ @nguyen2012inverse ].
Let $ x:=(x_1,\ldots, x_n)$ be the first row of $ M_n$, let $ M^{1}_{n-1}$ denote the bottom - right $ (n-1)\times (n-1)$ submatrix of $ M_n$, and for $ 2\leq i, j \leq n$, let $ c_{ij}$ denote the cofactor of $ M^{1}_{n-1}$ prevail by removing its $ (i-1)^{st}$ quarrel | $q_n$ are concerned, once agaik, by considering the evxnt thaf the fifst and last rows of $M_n$ are xquao (up uj a sign), we see thxt $q_n \geq (2+o(1))^{-n}$. It iw connonly belixbed thab thia lowzr bound is tighj.
\[conjecture:psob-singularity\] Wd kave $$q_n = (2+o(1))^{-n}.$$
In this paper, we obtain a much syrlnger upper boond om $q_n$, fhereby making progress towards \[cohjecturt:prob-singularity\].
\[thk:main-thm\] There exists a nahurap number $n_0 \in \N$ sufh that for all $b\geq n_0$, $$q_n \led 2^{-n^{1/4}\sqrt{\log{n}}/1000}.$$
Apart from kroviding a stronger conclusion, uur pxoof of the avovf theorem is conspderably shorbvr than previois works, and iktrodnces and extends several iovel combinatorial jools and hdzas in discrete randon natrif thaory (womd or xhidh are baaed on joiht work of rhe authors with Lun wbd Samotij [@FJMS2018]). We fejieve that these ideas allow for a unifped zpproach to the singulaeity problem for many differene discrete random matrix models, which have previogsly uedn kqndlea ij an ad-hoc manner. For completeness and for thq cpnnenience of the rcader, we have inclidfd sull proofs ow all cge simple background lemmas that we use fwom pther papers, making this paper completejt self contained.
Oucline of the propf anc comparison with previuus sork
------------------------------------------------------
In this subsectikv, we provide a vdry bsief, and rather imprecise, jutline oh our proof, xnd vomparq it to prfvious works of Nguyen [@nhuyen2012nnverve] and Verdhynin [@vershynin2014invertibility]; for further comkarhsot with tke worl of Costello, Tao, and Vu, seg [@nguyen2012iuverse].
Uet $x:=(x_1,\ldota,x_n)$ be vhe first ror of $M_n$, let $M^{1}_{t-1}$ denote the uottom-rigrt $(n-1)\rimew (n-1)$ suboxtrix of $M_n$, anc for $2\leq i,j \leq n$, oet $c_{ij}$ denote the cowzctor of $M^{1}_{n-1}$ obtcnntd by removing iys $(k-1)^{st}$ rlw | $q_n$ are concerned, once again, by considering that first and rows of $M_n$ sign), see that $q_n (2+o(1))^{-n}$. It is believed that this lower bound is \[conjecture:prob-singularity\] We have $$q_n = (2+o(1))^{-n}.$$ In this paper, we obtain a much upper bound on $q_n$, thereby making progress towards \[conjecture:prob-singularity\]. \[thm:main-thm\] There exists a number \in such for all $n\geq n_0$, $$q_n \leq 2^{-n^{1/4}\sqrt{\log{n}}/1000}.$$ Apart from providing a stronger conclusion, our proof of above theorem is considerably shorter than previous works, introduces and extends several combinatorial tools and ideas in random theory (some which based joint work of authors with Luh and Samotij [@FJLS2018]). We believe that these ideas allow for a unified approach to singularity problem different discrete matrix which previously been handled ad-hoc manner. For completeness and for the reader, we have included full proofs of the simple lemmas that we use from other making this paper completely self contained. Outline of proof and comparison with previous work ------------------------------------------------------ In this subsection, we provide a very brief, imprecise, outline of our and compare it previous of [@nguyen2012inverse] Vershynin [@vershynin2014invertibility]; further comparison with the work of Costello, Tao, and Vu, see Let $x:=(x_1,\ldots,x_n)$ be the first row of $M_n$, let $M^{1}_{n-1}$ bottom-right (n-1)$ submatrix of and for $2\leq i,j n$, $c_{ij}$ denote the cofactor obtained removing | $q_n$ are concerned, once again, by ConsiderinG the eVenT thAt The fIrst And last rows of $M_N$ Are eQual (up to a sign), we see that $Q_n \geq (2+O(1))^{-n}$. iT is cOMmOnly bElieved THaT THis LoWeR boUnD Is Tight.
\[ConJecture:Prob-singulAriTy\] we have $$q_n = (2+o(1))^{-n}.$$
In THiS paper, we obTaiN a much strongEr uPper boUnD on $Q_N$, therEby MakinG progrESs towaRds \[conjecTuRE:prob-sINgulariTY\].
\[ThM:maiN-thm\] There exists a nATuRAl number $n_0 \in \N$ suCh that FoR AlL $N\Geq N_0$, $$q_n \Leq 2^{-n^{1/4}\sqrt{\loG{n}}/1000}.$$
apart FRom provIDiNG A StrONger conclusioN, our proof of THe aBove thEoRem IS consiDerabLy SHorTer than prevIous Works, and iNtroduCEs and exTEnds sevEral noVel ComBinaTOrIaL toOlS And IDeAs iN DisCrete ranDoM mAtrix TheoRY (SOMe of WhiCh arE baseD on joint work oF thE autHOrs With LUh and samoTiJ [@FJLS2018]). we beliEve thAt These ideas allow For a Unified apProAcH to ThE singULarity ProBleM for manY differENt dIsCRETe Random matrix models, WhICH hAve previOusly bEEn HaNDled in an Ad-Hoc MannER. for coMpleTEnEss and foR the coNVeNiEnce of tHe Reader, We HavE inCludeD Full Proofs Of all the SimplE Background lemmAS that we use froM OtHER pAPers, MakIng this papeR comPLeteLy seLF cOntAIned.
OUtlinE oF ThE Proof and comparison wItH previOus woRk
------------------------------------------------------
In this subseCtion, we proVIDE a very brIef, aND rATher imprecise, oUtlinE of our prooF, And compaRe it tO previouS works of NGUYen [@nguyeN2012inVerSe] aNd VERShYnin [@vershynin2014INVertIbIlity]; foR fuRther coMpaRisOn wIth ThE work of CoStello, TaO, aNd vu, SeE [@ngUyen2012iNVerse].
Let $X:=(x_1,\LdoTs,X_n)$ bE the fIRst row Of $M_n$, lEt $M^{1}_{n-1}$ DeNoTE thE bottom-RIgHT $(N-1)\timEs (N-1)$ sUbmaTriX oF $M_n$, anD for $2\LEq i,J \leq n$, leT $c_{ij}$ denotE thE CofaCtOr Of $M^{1}_{n-1}$ obtAined by removiNg Its $(i-1)^{st}$ row | $q_n$ are concerned, once again, by cons ide rin gtheeven t that the fir s t an d last rows of $M_n$ a re eq ua l (up to a si gn), we se e tha t$q _n\g e q(2+o( 1)) ^{-n}$. It is com mon ly believed th a tthis lower bo und is tight .
\[conj ec tur e :prob -si ngula rity\] We hav e $$q_n = ( 2 +o(1)) ^ {-n}.$$ In thi s paper, we obtai n a much strongerupperbo u nd o n $ q_n $, thereby m aking progres s t o w a rds \[conjecture: prob-singul a rit y\].
\[ thm : main-t hm\]Th e reexists a na tura l number$n_0 \ i n \N$ s u ch that for a ll$n\ geqn _0 $, $$ q_ n \l e q2^{ - n^{ 1/4}\sqr t{ \l og{n} }/10 0 0 } . $$
Apa rt f rom p roviding a st ron gerc onc lusio n, ou r pr oo f ofthe ab ove t he orem is conside rabl y shorter th an pr ev iousw orks,and in troduce s and e x ten ds s e ve ral novel combinat or i a ltools an d idea s i nd iscretera ndo m ma t r ix th eory (s ome of w hich a r eba sed onjo int wo rk of th e aut h orswith L uh and S amoti j [@FJLS2018]). We believe th a tt h es e ide asallow for a uni f iedappr o ac h t o thesingu la r it y problem for many d if ferent disc rete random m atrix mode l s , which h avep re v iously been ha ndled in an ad- h oc manne r. Fo r comple teness an d for thecon ven ien ceo f t he reader, we h avein cludedful l proof s o f a llthe s imple bac kgroundle mm as t hat we u s e from o th erpa per s, ma k ing th is pa perco mp l ete ly self co n t aine d.
Outl ine o f the pro o f a nd comp arison wi thp revi ou swork
-- ------------- -- ---------- -- --- ------ - - -------- ------
In this subsect i on, wepro videa ve ry brief, an d rath eri mpreci se, ou tline o f o u r proo f , a ndco mpare it t o pre vious w orks of Ngu yen [@nguyen2012in v ers e] and Vershy nin [@v e r sh yni n 20 1 4in ve r tib i l ity]; for furth er compari so n w ith the wo r k o fCostell o, Tao, andV u, see[@nguyen2 012invers e] .
L e t $x :=(x_1,\ld ots,x_n) $ be thef irstr ow of $ M_n $, let $ M^{ 1}_{n -1}$ d e not e the botto m- right$(n-1 )\ times (n -1)$ submatrix of $M_n$ , andfor $ 2\l eq i,j \l eqn $,let $c_{i j}$denote the co fac tor o f $ M ^{1}_ {n-1 } $obt a inedby r e moving it s $ (i- 1 ) ^{ st}$ row | $q_n$_are concerned,_once again, by considering_the event_that_the first_and_last rows of_$M_n$ are equal_(up to a sign),_we see that_$q_n_\geq (2+o(1))^{-n}$. It is commonly believed that this lower bound is tight.
\[conjecture:prob-singularity\] We have_$$q_n_= (2+o(1))^{-n}.$$
In_this_paper,_we obtain a much stronger_upper bound on $q_n$, thereby_making progress_towards \[conjecture:prob-singularity\].
\[thm:main-thm\] There exists a natural number $n_0_\in_\N$ such that_for all $n\geq n_0$, $$q_n \leq 2^{-n^{1/4}\sqrt{\log{n}}/1000}.$$
Apart from providing_a stronger conclusion, our proof of_the above theorem_is_considerably_shorter than previous works,_and introduces and extends several novel_combinatorial tools and ideas in discrete_random matrix theory (some of which are_based on joint work of the_authors with Luh and Samotij_[@FJLS2018]). We_believe that these ideas allow_for a unified_approach to_the singularity problem_for many different discrete random matrix_models, which have_previously been handled in an ad-hoc_manner._For completeness and_for_the_convenience of_the reader, we_have_included full_proofs_of all the simple background lemmas_that_we use from other papers, making this_paper completely self contained.
Outline_of_the proof and comparison_with previous work
------------------------------------------------------
In this subsection,_we provide a very brief, and_rather imprecise,_outline of_our proof, and compare it to previous works of Nguyen [@nguyen2012inverse]_and Vershynin [@vershynin2014invertibility]; for further comparison_with the work of_Costello, Tao,_and_Vu, see [@nguyen2012inverse].
Let_$x:=(x_1,\ldots,x_n)$_be the_first row of $M_n$, let $M^{1}_{n-1}$ denote_the bottom-right_$(n-1)\times (n-1)$ submatrix of $M_n$, and_for $2\leq i,j \leq_n$,_let $c_{ij}$ denote the cofactor of_$M^{1}_{n-1}$ obtained by removing its $(i-1)^{st}$_row |
1.02 1.08
: Mean Values of Dependent Variables[]{data-label="tbl:average"}
We obtained 10 (drawings) $\times$ 4 (conditions) $\times$ 35 (subjects) = 1440 experimental data entries for each dependent variable. The mean values were computed and the results are shown in Table \[tbl:average\]. First, on performance measures, it can be seen that the subjects generally spent less time, exerted less effort and were more accurate when iteration number increased from condition c3, c6, c9 to c12. The performance efficiency also showed a clear increase across the conditions. In other words, the means of the performance measures were in good agreement with the pre-known overall quality.
Variable *F*(3,57) *p* Condition Pairs Detected Different (out of 6)
----------------- ----------- ------- -----------------------------------------------
Time 2.804 0.048 (c3, c9), (c3, c12)
Effort 7.491 0.000 (c3, c6), (c3, c9), (c3, c12)
Accuracy 2.118 0.108 none
Efficiency 7.442 0.000 (c3, c6), (c3, c9), (c3, c12)
Overall quality 28.596 0.000 all pairs but (c9, c12)
To test whether these trends of changes were significant at the level of 0.05, we ran repeated ANOVA tests with post-hoc comparisons of the Least Square Difference (LSD) method on each of the dependent variables. The results are shown in Table \[tbl:anova\]. There was a significant main effect on time, effort and efficiency ($p < 0.05$), but not on accuracy ($p > 0.05$). Post-hoc comparisons revealed that these dependent variables had different levels of capacity in detecting condition differences. More specifically, out of the six condition pairs in total, time data only found that two pairs of the conditions were different; effort found three; and efficiency found three, while no difference was shown in accuracy between any pair of the conditions.
Second, on overall quality, the mean overall quality value was -2.14 for c3, 0.04 for c6, 1.02 for c | 1.02 1.08
: Mean Values of Dependent Variables[]{data - label="tbl: average " }
We obtained 10 (drawings) $ \times$ 4 (condition) $ \times$ 35 (discipline) = 1440 experimental data entries for each dependent variable star. The mean values were computed and the consequence are shown in Table \[tbl: average\ ]. First, on operation bill, it can be seen that the subjects generally spent less time, exerted less attempt and were more accurate when iteration number increased from condition c3, c6, c9 to c12. The operation efficiency also showed a clean increase across the weather. In other words, the means of the performance measures were in effective agreement with the pre - known overall quality.
Variable * F*(3,57) * p * Condition Pairs Detected Different (out of 6)
----------------- ----------- ------- -----------------------------------------------
Time 2.804 0.048 (c3, c9), (c3, c12)
Effort 7.491 0.000 (c3, c6), (c3, c9), (c3, c12)
Accuracy 2.118 0.108 none
Efficiency 7.442 0.000 (c3, c6), (c3, c9), (c3, c12)
Overall quality 28.596 0.000 all pairs but (c9, c12)
To test whether these trends of changes were significant at the level of 0.05, we ran repeated ANOVA tests with post - hoc comparison of the Least Square Difference (LSD) method on each of the dependent variable. The result are shown in Table \[tbl: anova\ ]. There was a significant chief effect on time, effort and efficiency ($ p < 0.05 $), but not on accuracy ($ phosphorus > 0.05 $). Post - hoc comparisons revealed that these dependent variables had different levels of capacity in detecting condition differences. More specifically, out of the six condition pair in total, clock time data only found that two pairs of the conditions were unlike; effort found three; and efficiency found three, while no difference was shown in accuracy between any pair of the conditions.
Second, on overall quality, the mean overall timbre value was -2.14 for c3, 0.04 for c6, 1.02 for c | 1.02 1.08
: Mean Values of Dependtnt Variables[]{data-label="tbn:averafe"}
We obtxined 10 (drawings) $\times$ 4 (condivionw) $\timtf$ 35 (subjects) = 1440 expefimental fata enteies dor each dxlendent variagpe. Tke mean values wgre computed and the resulds axe shown in Table \[tbl:average\]. First, on perforkajce measures, ij can fe ssvn that the subjects generally apent ltss time, exerted lrss effort and were more afcurwte when iteration number inctsasqe from condigion c3, c6, c9 to c12. The petformance efficiency also showed a clzar increasg zcgmss the convitionf. In other words, the means pf the performsncx meqsures were in good ajreement with the prg-known ovesaml quality.
Variavlw *F*(3,57) *p* Cpnsition Pamrs Detectes Different (out of 6)
----------------- ----------- ------- -----------------------------------------------
Time 2.804 0.048 (s3, c9), (c3, c12)
Effort 7.491 0.000 (c3, b6), (c3, d9), (c3, c12)
Accuracy 2.118 0.108 none
Efflciency 7.442 0.000 (c3, c6), (c3, c9), (c3, c12)
Overall quality 28.596 0.000 alo iairr bkt (c9, c12)
To test whether these trends of changes sete significant ab the level of 0.05, we rwn tepeated ANOVA tests wifh post-hoc comparidons of the Oeast Squwre Cifference (LSD) method on eaxh of the deiendwnt variables. The xesults are rhowm in Yable \[tbl:anova\]. There was a sjgnificant lain effedg on time, effort anc afficiency ($p < 0.05$), but not on wccuracy ($' > 0.05$). Ppst-hoc zompsrisonf revealed that these dependent vagiablgs had different levels of capacity in detectinj condition dofxervnces. Morz speclfically, out of the six condijion pairf in gotal, time data oily found thwt two pairs mv the conditmons were difderebt; effofg found three; snd efficpeucy found three, while no dinferevde was shown in acxuracy between sny payr oh the wonditions.
Sewond, on pveraul quality, bhe meam overall quality vanue sas -2.14 for c3, 0.04 for c6, 1.02 fjr c | 1.02 1.08 : Mean Values of Dependent obtained (drawings) $\times$ (conditions) $\times$ 35 entries each dependent variable. mean values were and the results are shown in \[tbl:average\]. First, on performance measures, it can be seen that the subjects generally less time, exerted less effort and were more accurate when iteration number increased condition c6, to The performance efficiency also showed a clear increase across the conditions. In other words, the means the performance measures were in good agreement with pre-known overall quality. Variable *p* Condition Pairs Detected Different of ----------------- ----------- ----------------------------------------------- 2.804 (c3, c9), (c3, Effort 7.491 0.000 (c3, c6), (c3, c9), (c3, c12) Accuracy 2.118 0.108 none Efficiency 7.442 0.000 (c3, (c3, c9), Overall quality 0.000 pairs (c9, c12) To these trends of changes were significant of 0.05, we ran repeated ANOVA tests with comparisons of Least Square Difference (LSD) method on of the dependent variables. The results are shown Table \[tbl:anova\]. There was a significant main effect on time, effort and efficiency ($p < not on accuracy ($p 0.05$). Post-hoc comparisons that dependent had levels of in detecting condition differences. More specifically, out of the six condition in total, time data only found that two pairs of were effort found three; efficiency found three, while difference shown in accuracy between of conditions. quality, mean quality value was -2.14 c3, 0.04 for c6, 1.02 c | 1.02 1.08
: Mean Values of Dependent VariAbles[]{data-lAbel="tBl:aVerAgE"}
We oBtaiNed 10 (drawings) $\timES$ 4 (conDitions) $\times$ 35 (subjects) = 1440 exPerimEnTAl daTA eNtrieS for eacH DePENdeNt VaRiaBlE. thE mean ValUes were Computed anD thE rEsults are shoWN iN Table \[tbl:aVerAge\]. First, on peRfoRmance MeAsuREs, it cAn bE seen That thE SubjecTs generalLy SPent leSS time, exERTeD lesS effort and were morE AcCUrate when iteraTion nuMbER iNCReaSed From conditIoN c3, c6, c9 tO C12. The perFOrMANCe eFFiciency also sHowed a clear INcrEase acRoSs tHE condiTions. in OTheR words, the meAns oF the perfoRmance MEasures WEre in goOd agreEmeNt wIth tHE pRe-KnoWn OVerALl QuaLIty.
variable *f*(3,57) *p* coNditiOn PaIRS dEtecTed diffErent (Out of 6)
----------------- ----------- ------- -----------------------------------------------
Time 2.804 0.048 (c3, c9), (c3, c12)
effOrt 7.491 0.000 (c3, C6), (C3, c9), (c3, C12)
AccuRacy 2.118 0.108 nOne
EFfIcienCy 7.442 0.000 (c3, c6), (c3, c9), (C3, c12)
OveRaLl quality 28.596 0.000 all paiRs buT (c9, c12)
To test WheThEr tHeSe treNDs of chAngEs wEre signIficant AT thE lEVEL oF 0.05, we ran repeated ANOVa tESTs With post-Hoc comPArIsONs of the LEaSt SQuarE dIfferEnce (lsD) Method on Each of THe DePendent VaRiableS. THe rEsuLts arE ShowN in TabLe \[tbl:anoVa\]. TheRE was a significaNT main effect on TImE, EFfORt anD efFiciency ($p < 0.05$), buT not ON accUracY ($P > 0.05$). POst-HOc comParisOnS ReVEaled that these depenDeNt variAbles Had different lEvels of capACITy in deteCtinG CoNDition differenCes. MoRe specificALly, out of The siX conditiOn pairs in TOTal, time dAta OnlY foUnd THAt Two pairs of the CONditIoNs were dIffErent; efForT foUnd ThrEe; And efficiEncy founD tHrEe, WhIle No difFErence waS sHowN iN acCuracY BetweeN any pAir oF tHe COndItions.
SECoND, On ovErAlL quaLitY, tHe meaN oveRAll Quality Value was -2.14 fOr c3, 0.04 FOr c6, 1.02 fOr C | 1.02 1.08
: Mean Valu es of Depe ndent Va ria bl es[] {dat a-label="tbl:a v erag e"}
We obtained 10 (d rawin gs ) $\t i me s$ 4(condit i on s ) $\ ti me s$35 (s ubjec ts) = 1440 experimen tal d ata entriesf or each depe nde nt variable. Th e mean v alu e s wer e c omput ed and the re sults are s h own in Table \ [ t bl :ave rage\]. First, on pe r formance measu res, i tc an b e s een that thesu bject s genera l ly s p ent less time, ex erted lesse ffo rt and w ere more a ccura te whe n iteration num ber incre ased f r om cond i tion c3 , c6,c9toc12. Th eper fo r man c eeff i cie ncy also s ho wed a cle a r i ncre ase acr oss t he conditions . I n ot h erwords , the mea ns of t he per forma nc e measures were ingood agre eme nt wi th thep re-kno wnove rall qu ality. V ar i a b le *F*(3,57 )*p * Co nditio n P ai r s Detect ed Di ffer e n t (ou t of 6)
----- ------ - -- -- - ----- -- ---- - -- --- - - ----- - ---- ------ -------- ----- - -------------- - -
Time 2.8 04 0.048 ( c 3, c 9),( c3 , c 1 2)
Effor t 7.491 0. 000 (c3,c6), (c3, c9) , (c3, c12 ) Accurac y 2.118 0.10 8 none
Efficien cy 7.4 42 0 . 000 (c 3,c6) , ( c3, c 9) , (c3, c12)
O vera ll qualit y 28.596 0.0 00 all pairs but (c9 ,c1 2)
Totestw hether t he setr end s ofc hanges were sig ni fi c ant at the le v e l of 0 .0 5, w e r an repe ated ANO VA test s with po st- h oc c om pa risonsof the LeastSq uare Diffe re nce (LSD) m ethod on each of the dependentv ariable s.The r esul ts are sh own in Ta ble \[tbl: anova\ ]. Th er e w a s a si g n if ica nt main effe c t on time ,effo rt andefficiency ($p < 0 . 05$ ), but not on ac cura c y ( $p> 0 . 05$ ). Pos t - hoc comparisons revealedth a tthese depe n den tvariabl es haddiffe r ent lev els of ca pacity in d etec t i ngconditiondifferen ces. More speci f ic ally, ou t of t he si x con dition pai rs in total ,time d ata o nl y foundthat two pairs of the c onditi ons w ere differen t;e ffo rt foundthre e; and eff ici enc y fou ndt hree, whi l enod iffer ence was shown in ac c u ra cy betweena n y pa ir of th e condi tion s.
Second, on ov e rall quality,them e anove r allqu ality value wa s - 2. 1 4 for c3, 0 .04 for c6, 1.02 fo rc | 1.02 _ 1.08
_ : Mean Values_of Dependent_Variables[]{data-label="tbl:average"}
We_obtained 10_(drawings)_$\times$ 4 (conditions)_$\times$ 35 (subjects)_= 1440 experimental data_entries for each_dependent_variable. The mean values were computed and the results are shown in Table \[tbl:average\]. First,_on_performance measures,_it_can_be seen that the subjects_generally spent less time, exerted_less effort_and were more accurate when iteration number increased_from_condition c3, c6,_c9 to c12. The performance efficiency also showed a_clear increase across the conditions. In_other words, the_means_of_the performance measures were_in good agreement with the pre-known_overall quality.
Variable _ _*F*(3,57) *p* _ Condition Pairs Detected_Different (out_of 6)
----------------- -----------_------- -----------------------------------------------
_Time _ _ _ _2.804 __0.048 _(c3,_c9),_(c3, c12)
_ Effort __ __ _7.491_ 0.000_ (c3, c6),_(c3,_c9), (c3, c12)
_Accuracy _ 2.118_ _ _ 0.108 none
Efficiency _ 7.442 _ _ 0.000__ (c3, c6),_(c3,_c9), (c3,_c12)
Overall quality 28.596_ _ 0.000 _all pairs but (c9,_c12)
To_test whether these trends of changes_were significant at the level of_0.05, we ran repeated ANOVA_tests_with_post-hoc comparisons of the Least_Square Difference (LSD) method on each_of the dependent_variables. The results are shown in Table \[tbl:anova\]._There_was a significant main effect on_time,_effort and efficiency ($p < 0.05$),_but_not_on accuracy ($p > 0.05$)._Post-hoc comparisons revealed that these dependent_variables had different levels of capacity in detecting condition_differences. More specifically,_out of the six condition_pairs_in_total, time data only found that two pairs of the_conditions were_different; effort found_three; and efficiency found three, while no difference was shown_in accuracy between any pair of the_conditions.
Second, on overall quality, the mean overall quality value was -2.14_for c3, 0.04 for c6, 1.02 for c |
:7}$$ [*b) The first order transition line takes over the AT line:*]{} The paradox in [*a)*]{} is resolved by noting that for a given $k$, multiple RS solutions of $q$ may coexist for a given $\alpha$. Indeed, $\alpha^{(0)}(k,q)$ is monotonic in $q$ for $k\!\le\!k^{(0)}_c\!=\!2.89$, and not so otherwise. Among the coexisting stable RS solutions for $k > k^{(0)}_c$, the one with the lowest entropy is relevant. Hence, there is a first order transition in $q$ at a value $\alpha^{(0)}_1(k)$, determined by the vanishing of the entropy difference between the coexisting stable solutions. The jump in $q$ across the transition widens from 0 at $k=k^{(0)}_c$, and when $k$ increases above $k^{(0)}_0=3.11$, $q$ jumps directly from zero to nonzero at the transition point. The line of $\alpha^{(0)}_1(k)$ starts from $k=k^{(0)}_c$; it crosses the line of $\alpha_{\rm AT}(k)$ at $k^{(0)}_0$, replacing it to become the physically observable phase transition line. In the limit of large $k$, the first order transition line is given by $\alpha^{(0)}_1(k)\!=\!0.94 \alpha^{(0)}_c(k)$, where $q$ jumps from 0 to a high value of $1\!-\!0.27k^{-2}$.
The phase diagram illustrates the nature of the phase transitions. For low tolerance $k$, each constraint results in a significant reduction in the version space, and there is a continuous transition to a phase of nonzero value of $q$. For high tolerance $k$, each constraint introduces a volume reduction which is less significant, resembling a percolation process, in which the transition of $q$ from zero to nonzero is discontinuous. However, even in the large $k$ limit, the region of high $q$ spans a nonvanishing range of $\alpha$ below the critical population, and the picture of percolation transition has to be refined in the next approximation.
[*2. First step RSB approximation ([RSB]{}$_1$, superscript $^{(1)}$):*]{} Here the $n$ replicas are organized into clusters, each with $m$ replicas. | : 7}$$ [ * b) The first order transition line take over the AT credit line :* ] { } The paradox in [ * a) * ] { } is resolved by noting that for a given $ k$, multiple RS solution of $ q$ may coexist for a given $ \alpha$. Indeed, $ \alpha^{(0)}(k, q)$ is monotonic in $ q$ for $ k\!\le\!k^{(0)}_c\!=\!2.89 $, and not indeed differently. Among the coexisting stable RS solutions for $ k > k^{(0)}_c$, the one with the lowest randomness is relevant. Hence, there is a beginning club transition in $ q$ at a value $ \alpha^{(0)}_1(k)$, determine by the vanishing of the entropy difference between the coexist stable solutions. The jump in $ q$ across the conversion widens from 0 at $ k = k^{(0)}_c$, and when $ k$ increases above $ k^{(0)}_0=3.11 $, $ q$ jumps immediately from zero to nonzero at the transition compass point. The line of $ \alpha^{(0)}_1(k)$ starts from $ k = k^{(0)}_c$; it cross the line of $ \alpha_{\rm AT}(k)$ at $ k^{(0)}_0 $, replacing it to become the physically observable phase transition line. In the limit of large $ k$, the first order transition line is given by $ \alpha^{(0)}_1(k)\!=\!0.94 \alpha^{(0)}_c(k)$, where $ q$ jumps from 0 to a high value of $ 1\!-\!0.27k^{-2}$.
The phase diagram illustrates the nature of the phase transitions. For low permissiveness $ k$, each constraint results in a meaning decrease in the version space, and there is a continuous transition to a phase of nonzero value of $ q$. For eminent tolerance $ k$, each constraint introduces a volume decrease which is less significant, resembling a percolation process, in which the transition of $ q$ from zero to nonzero is discontinuous. However, even in the large $ k$ limit, the region of high $ q$ spans a nonvanishing range of $ \alpha$ below the critical population, and the picture of percolation transition receive to be refined in the following approximation.
[ * 2. First step RSB approximation ([ RSB]{}$_1 $, superscript $ ^{(1)}$ ): * ] { } Here the $ n$ replicas are organized into bunch, each with $ m$ replicas. | :7}$$ [*b) Hhe first order transitiun line takes orwr the AT lihe:*]{} The pxradox in [*a)*]{} is resolved by nltung tyat for a given $k$, multkple RS sllutions of $w$ may coexist for a given $\wlphc$. Mndeed, $\alpha^{(0)}(k,q)$ ls monotoniw in $q$ for $k\!\le\!n^{(0)}_c\!=\!2.89$, aud not so otherwise. Among the coexiseing stsbpe RS solutionf fog $h > k^{(0)}_d$, the one with the lowest entropy js relenant. Hence, there os a first order transitioj in $q$ at a value $\alphw^{(0)}_1(k)$, determings br the vanishivg of the tncropy diffetence between the coexisting stacle splutions. Tye julk in $q$ acrosw the transition widens fsom 0 at $k=k^{(0)}_c$, and when $l$ iicreqses above $k^{(0)}_0=3.11$, $q$ jumps virectly from zero tj nonzero ac the transition poinr. Rhe lhne mf $\aupha^{(0)}_1(y)$ suarvs rrom $k=n^{(0)}_c$; mt crosses fhe line of $\alpha_{\rm AT}(k)$ at $k^{(0)}_0$, rtplwbong it to bedome tre physically observable phase transition lihe. In the limit of largw $k$, the first order ttansition jine is given by $\alpha^{(0)}_1(k)\!=\!0.94 \alpha^{(0)}_c(k)$, where $q$ jumps frok 0 to x hnnm vauye of $1\!-\!0.27k^{-2}$.
The phase diagram illustrates the nature kf tme phase transitlons. For low tolersnfe $h$, each constrxint rzaumts in a significajt redustion in the vtrsiom space, and there is a contunuous transptiob to a phase of nouzero value uf $q$. For nigh tolerance $k$, each cunstdaint introfuces a vkuume reduction wficm iv less significant, resemblyng a perrolatnon procdss, on whish the trajsition of $q$ from zero tl noneero iv discontijuous. However, even in the large $k$ limit, the rggimn mf high $z$ spaks a nonvanishigg range of $\alkha$ below the zritical pkpulatimn, and the [icture of pesfolation traisition hws ti be refinea in the next alproximatpou.
[*2. First srep RSB approximatlon ([RRG]{}$_1$, superscript $^{(1)}$):*]{} Ktre the $n$ replicas ard owgwnmzed ytto clusters, eacf wkyh $m$ feplicas. | :7}$$ [*b) The first order transition line the line:*]{} The in [*a)*]{} is a $k$, multiple RS of $q$ may for a given $\alpha$. Indeed, $\alpha^{(0)}(k,q)$ monotonic in $q$ for $k\!\le\!k^{(0)}_c\!=\!2.89$, and not so otherwise. Among the coexisting stable solutions for $k > k^{(0)}_c$, the one with the lowest entropy is relevant. there a order in $q$ at a value $\alpha^{(0)}_1(k)$, determined by the vanishing of the entropy difference between the stable solutions. The jump in $q$ across the widens from 0 at and when $k$ increases above $q$ directly from to at transition point. The of $\alpha^{(0)}_1(k)$ starts from $k=k^{(0)}_c$; it crosses the line of $\alpha_{\rm AT}(k)$ at $k^{(0)}_0$, replacing it to the physically transition line. the of $k$, the first line is given by $\alpha^{(0)}_1(k)\!=\!0.94 \alpha^{(0)}_c(k)$, from 0 to a high value of $1\!-\!0.27k^{-2}$. phase diagram the nature of the phase transitions. low tolerance $k$, each constraint results in a reduction in the version space, and there is a continuous transition to a phase of of $q$. For high $k$, each constraint a reduction is significant, resembling percolation process, in which the transition of $q$ from zero to is discontinuous. However, even in the large $k$ limit, the high spans a nonvanishing of $\alpha$ below the population, the picture of percolation to refined approximation. First RSB approximation ([RSB]{}$_1$, superscript Here the $n$ replicas are into clusters, each with | :7}$$ [*b) The first order transition lIne takes ovEr the aT lIne:*]{} thE parAdox In [*a)*]{} is resolved bY NotiNg that for a given $k$, multipLe RS sOlUTionS Of $Q$ may cOexist fOR a GIVen $\AlPhA$. InDeED, $\aLpha^{(0)}(k,Q)$ is MonotonIc in $q$ for $k\!\lE\!k^{(0)}_c\!=\!2.89$, AnD not so otherwISe. among the coExiSting stable Rs soLutionS fOr $k > K^{(0)}_C$, the oNe wIth thE lowesT EntropY is relevaNt. hEnce, thERe is a fiRST oRder Transition in $q$ at a vALuE $\Alpha^{(0)}_1(k)$, determinEd by thE vANiSHIng Of tHe entropy dIfFerenCE betweeN ThE COExiSTing stable solUtions. The juMP in $Q$ acrosS tHe tRAnsitiOn widEnS FroM 0 at $k=k^{(0)}_c$, and whEn $k$ iNcreases aBove $k^{(0)}_0=3.11$, $q$ JUmps dirECtly froM zero tO noNzeRo at THe TrAnsItIOn pOInT. ThE LinE of $\alpha^{(0)}_1(K)$ sTaRts frOm $k=k^{(0)}_C$; IT CRossEs tHe liNe of $\aLpha_{\rm AT}(k)$ at $k^{(0)}_0$, rEplAcinG It tO becoMe the PhysIcAlly oBservaBle phAsE transition line. in thE limit of lArgE $k$, The FiRst orDEr tranSitIon Line is gIven by $\aLPha^{(0)}_1(K)\!=\!0.94 \aLPHA^{(0)}_c(K)$, where $q$ jumps from 0 to A hIGH vAlue of $1\!-\!0.27k^{-2}$.
THe phasE DiAgRAm illustRaTes The nATUre of The pHAsE transitIons. FoR LoW tOlerancE $k$, Each coNsTraInt ResulTS in a SignifIcant redUctioN In the version spACe, and there is a COnTINuOUs trAnsItion to a phaSe of NOnzeRo vaLUe Of $q$. fOr higH toleRaNCe $K$, Each constraint introDuCes a voLume rEduction which Is less signIFICant, reseMbliNG a PErcolation procEss, in Which the trANsition oF $q$ froM zero to nOnzero is dISContinuoUs. HOweVer, EveN IN tHe large $k$ limit, THE regIoN of high $Q$ spAns a nonVanIshIng RanGe Of $\alpha$ beLow the crItIcAl PoPulAtion, ANd the picTuRe oF pErcOlatiON transItion Has tO bE rEFinEd in the NExT APproXiMaTion.
[*2. firSt Step RsB apPRoxImation ([rSB]{}$_1$, supersCriPT $^{(1)}$):*]{} HerE tHe $N$ replicAs are organizeD iNto clusterS, eAch With $m$ rEPLicas. | :7}$$ [*b) The first order transitio n lin e t ake sover the AT line:*]{}T he p aradox in [*a)*]{} isresol ve d byn ot ing t hat for ag i ven $ k$ , m ul t ip le RS so lutions of $q$ ma y c oe xist for a g i ve n $\alpha$ . I ndeed, $\alp ha^ {(0)}( k, q)$ is mo not onicin $q$ for $k \!\le\!k^ {( 0 )}_c\! = \!2.89$ , an d no t so otherwise. A m on g the coexistin g stab le RS s olu tio ns for $k>k^{(0 ) }_c$, t h eo n e wi t h the lowestentropy isr ele vant.He nce , there is a f i rst order tran siti on in $q$ at av alue $\ a lpha^{( 0)}_1( k)$ , d eter m in ed by t h e v a ni shi n g o f the en tr op y dif fere n c e betw een the coex isting stable so luti o ns. Thejumpin $ q$ acro ss the tran si tion widens fro m 0at $k=k^{ (0) }_ c$, a nd wh e n $k$inc rea ses abo ve $k^{ ( 0)} _0 = 3 . 11 $, $q$ jumps direc tl y fr om zeroto non z er oa t the tr an sit ionp o int.Thel in e of $\a lpha^{ ( 0) }_ 1(k)$ s ta rts fr om $k =k^ {(0)} _ c$;it cro sses the line of $\alpha_{\r m AT}(k)$ at $ k ^{ ( 0 )} _ 0$,rep lacing it t o be c omethep hy sic a lly o bserv ab l ep hase transition lin e. In th e lim it of large $ k$, the fi r s t order t rans i ti o n line is give n by$\alpha^{( 0 )}_1(k)\ !=\!0 .94 \alp ha^{(0)}_ c ( k)$, whe re$q$ ju mps f ro m 0 to a high v alue o f $1\!- \!0 .27k^{- 2}$ .
The ph as e diagram illustr at es t he na tureo f the ph as e t ra nsi tions . For l ow to lera nc e$ k$, each c o ns t r aint r es ults in a sign ific a ntreducti on in the ve r sion s pa ce, and there is a c on tinuous tr an sit ion to a phase o f nonzero value of $q$. For hig h t olera nce$k$, each co nstrai nti ntrodu ces avolum ered u c tionw h ic h i sless signi f i can t, re se mbli ng a pe rcolation process, inwhich the tra nsi tion o f$q$ fr o m z er o to n onzero is disco ntinuous.Ho w ev er, even i n th elarge $ k$ limi t, th e region of high$q$ spans a non v a nis hing range of $\al pha$ belo w thec ri tical po pulati on , a nd th e pict u reof pe rcolat io n tran sitio nhas to b e refined in the next a pproxi matio n.
[*2. Fir sts tep RSB appr oxim ation ([RS B]{ }$_ 1$, s upe r scrip t $^ { (1 )}$ ) :*]{} Her e the $n$r ep lic a s a re organize d i nto clus ter s , each wit h $m$ replicas. | :7}$$ [*b)_The first_order transition line takes_over the_AT_line:*]{} The_paradox_in [*a)*]{} is_resolved by noting_that for a given_$k$, multiple RS_solutions_of $q$ may coexist for a given $\alpha$. Indeed, $\alpha^{(0)}(k,q)$ is monotonic in $q$_for_$k\!\le\!k^{(0)}_c\!=\!2.89$, and_not_so_otherwise. Among the coexisting stable_RS solutions for $k >_k^{(0)}_c$, the_one with the lowest entropy is relevant. Hence,_there_is a first_order transition in $q$ at a value $\alpha^{(0)}_1(k)$, determined_by the vanishing of the entropy_difference between the_coexisting_stable_solutions. The jump in_$q$ across the transition widens from_0 at $k=k^{(0)}_c$, and when $k$_increases above $k^{(0)}_0=3.11$, $q$ jumps directly from_zero to nonzero at the transition_point. The line of $\alpha^{(0)}_1(k)$_starts from_$k=k^{(0)}_c$; it crosses the line_of $\alpha_{\rm AT}(k)$_at $k^{(0)}_0$,_replacing it to_become the physically observable phase transition_line. In the_limit of large $k$, the first_order_transition line is_given_by_$\alpha^{(0)}_1(k)\!=\!0.94 \alpha^{(0)}_c(k)$,_where $q$ jumps_from_0 to_a_high value of $1\!-\!0.27k^{-2}$.
The phase diagram_illustrates_the nature of the phase transitions. For_low tolerance $k$, each_constraint_results in a significant_reduction in the version space,_and there is a continuous transition_to a_phase of_nonzero value of $q$. For high tolerance $k$, each constraint introduces_a volume reduction which is less_significant, resembling a percolation_process, in_which_the transition of_$q$_from zero_to nonzero is discontinuous. However, even in_the large_$k$ limit, the region of high_$q$ spans a nonvanishing_range_of $\alpha$ below the critical population,_and the picture of percolation transition_has to be refined in_the_next_approximation.
[*2. First step RSB approximation_([RSB]{}$_1$, superscript $^{(1)}$):*]{} Here the $n$_replicas are organized_into clusters, each with $m$ replicas. |
872
GLP 3.1375 2.3051 1.4284 0.9860 0.9831
P+XS 1.4588 1.0546 1.3259 0.9967 0.9940
NLV [**1.2073**]{} [**0.8702**]{} [**1.0079**]{} [**0.9977**]{} [**0.9959**]{}
NLVD 1.3580 0.9823 1.3062 0.9967 0.9945
: Quantitative evaluation of the fused products on simulated data from RGB aerial images at resolution of 60 cm per pixel. For these experiments, the low-resolution spectral components were registered and the panchro-spectral constraint fulfilled with $\alpha_B=\alpha_G=\alpha_R=\frac{1}{3}$. The numerical results are less competitive than with data at 30 cm, which was highly expected. In general terms, the same conclusions than in Table \[table\_30cm\_RGB\_regist\_linear\] can be drawn for almost all metrics. However, the increasing in spatial distortion because of the reduction in resolution is more noticeable in CS-based methods.[]{data-label="table_60cm_RGB_regist_linear"}
Figure \[fig\_60cm\_s13\_RGB\_regist\_linear\] displays close-ups of the fusion products on the first image from the proposed dataset at a resolution of 60 cm per pixel. The Gaussian standard deviation used for the simulation of the low-resolution spectral components was $\sigma=1.3$. In order to help the visual analysis, Figure \[fig\_60cm\_s13\_RGB\_regist\_linear\_dif\] shows the difference images between the reference one and the result provided by each method. The first conclusion that can be drawn from the visual inspection is the superiority of the variational models under comparison, particularly NLV and NLVD, over CS-based as well as MRA-based techniques. Although P+XS performed well from a quantitive point of view, the pansharpened image shows some annoying color spots. We further observe that CS and MRA families lead to greater spectral and spatial quality loss than NLV and NLVD. For example, see how the color of the fireplaces in the fused products provided by these two classes of approaches is not preserved | 872
GLP 3.1375 2.3051 1.4284 0.9860 0.9831
P+XS 1.4588 1.0546 1.3259 0.9967 0.9940
NLV [ * * 1.2073 * * ] { } [ * * 0.8702 * * ] { } [ * * 1.0079 * * ] { } [ * * 0.9977 * * ] { } [ * * 0.9959 * * ] { }
NLVD 1.3580 0.9823 1.3062 0.9967 0.9945
: Quantitative evaluation of the fused products on simulated data from RGB aeriform image at resolution of 60 centimeter per pixel. For these experiment, the low - resolution apparitional components were registered and the panchro - spectral restraint satisfy with $ \alpha_B=\alpha_G=\alpha_R=\frac{1}{3}$. The numerical resultant role are less competitive than with data at 30 cm, which was highly have a bun in the oven. In cosmopolitan terms, the same stopping point than in Table \[table\_30cm\_RGB\_regist\_linear\ ] can be drawn for almost all prosody. However, the increasing in spatial distortion because of the reduction in solution is more noticeable in CS - based methods.[]{data - label="table_60cm_RGB_regist_linear " }
Figure \[fig\_60cm\_s13\_RGB\_regist\_linear\ ] displays near - ups of the fusion products on the first image from the proposed dataset at a settlement of 60 cm per pixel. The Gaussian standard deviation used for the simulation of the low - resolution spectral components was $ \sigma=1.3$. In order to help the visual analysis, Figure \[fig\_60cm\_s13\_RGB\_regist\_linear\_dif\ ] shows the difference images between the reference one and the result supply by each method. The first ending that can be describe from the visual inspection is the superiority of the variational model under comparison, particularly NLV and NLVD, over CS - based as well as MRA - based technique. Although P+XS performed well from a quantitive point of view, the pansharpened image shows some annoying color spot. We further observe that CS and MRA families lead to great apparitional and spatial timbre personnel casualty than NLV and NLVD. For example, see how the color of the fireplace in the fused products provided by these two classes of approaches is not preserved | 872
GLP 3.1375 2.3051 1.4284 0.9860 0.9831
P+ZS 1.4588 1.0546 1.3259 0.9967 0.9940
KLV [**1.2073**]{} [**0.8702**]{} [**1.0079**]{} [**0.9977**]{} [**0.9959**]{}
NLVD 1.3580 0.9823 1.3062 0.9967 0.9945
: Quameitafpvt evaluation of the fused producta on sikulated data grom RGB aerial images at gesopution of 60 cm per oixel. For tyese wxperiments, ghe low-resolution specjral components were registered xnd tke panchro-skzxtrwn constrainv fulfplled with $\alina_B=\alpva_G=\alphs_R=\frac{1}{3}$. The numcrican rwsults are less compevitive than with datw at 30 cm, fhnch was highly expectwd. In ggneran tefns, ghe seme conclksikns than ih Table \[taboe\_30cm\_RGB\_regist\_linear\] cwb be drawn fod almoft all metrics. However, the increasing in vpafial distortion because of the reduction in tesolution is more noticeable in CS-based methods.[]{data-label="tatle_60cm_CGC_reylfg_oijear"}
Figure \[fig\_60cm\_s13\_RGB\_regist\_linear\] displays clofs-uks pf the fusion iroducts on the fitsh oiage from the proposed dataset at a resopution jf 60 cn per pixtl. Thr Gaussian standard deviatiin used for nhe wimulation of the pow-resolutiun skectrak components was $\sigma=1.3$. Nn ordsr to help hhe visuam analysis, Figure \[fin\_60cm\_v13\_RGB\_regist\_linear\_dif\] shows ehe diffecence images beteeen tre referenfe onc and the result prlvideb by aach methof. The first conclusion that can be drawn from tve nisual infpectlon is the supewiority of the variatnonal oodels undvr comparmson, particujarly NLV and JLVD, over CS-uased as rell as NRA-basea techniques. Alyhough P+XS performee well from a quanbitivg loint of view, tkt pqnsharpened imabe rhors spmq annoying conor rpogx. We wurther obscrvd thst CS and MRA familias lsad to greater spevtval and skatial quwlity loss thsn NLV and NLVD. Fog exakplx, see now the color of the fireplaces ih the fusfd iroducts provyded by these twp classes of approaches is not preservev | 872 GLP 3.1375 2.3051 1.4284 0.9860 0.9831 1.0546 0.9967 0.9940 [**1.2073**]{} [**0.8702**]{} [**1.0079**]{} 1.3062 0.9945 : Quantitative of the fused on simulated data from RGB aerial at resolution of 60 cm per pixel. For these experiments, the low-resolution spectral were registered and the panchro-spectral constraint fulfilled with $\alpha_B=\alpha_G=\alpha_R=\frac{1}{3}$. The numerical results are competitive with at cm, which was highly expected. In general terms, the same conclusions than in Table \[table\_30cm\_RGB\_regist\_linear\] can drawn for almost all metrics. However, the increasing spatial distortion because of reduction in resolution is more in methods.[]{data-label="table_60cm_RGB_regist_linear"} Figure displays of fusion products on first image from the proposed dataset at a resolution of 60 cm per pixel. The Gaussian standard used for of the spectral was In order to visual analysis, Figure \[fig\_60cm\_s13\_RGB\_regist\_linear\_dif\] shows the the reference one and the result provided by method. The conclusion that can be drawn from visual inspection is the superiority of the variational under comparison, particularly NLV and NLVD, over CS-based as well as MRA-based techniques. Although P+XS from a quantitive point view, the pansharpened shows annoying spots. further observe CS and MRA families lead to greater spectral and spatial quality than NLV and NLVD. For example, see how the color fireplaces the fused products by these two classes approaches not preserved | 872
GLP 3.1375 2.3051 1.4284 0.9860 0.9831
P+XS 1.4588 1.0546 1.3259 0.9967 0.9940
NLV [**1.2073**]{} [**0.8702**]{} [**1.0079**]{} [**0.9977**]{} [**0.9959**]{}
NLVD 1.3580 0.9823 1.3062 0.9967 0.9945
: Quantitative Evaluation Of the FusEd pRoDuctS on sImulated data frOM RGB Aerial images at resolutiOn of 60 cM pER pixEL. FOr theSe experIMeNTS, thE lOw-ResOlUTiOn speCtrAl compoNents were rEgiStEred and the paNChRo-spectral ConStraint fulfiLleD with $\aLpHa_B=\ALpha_G=\AlpHa_R=\frAc{1}{3}$. The nUMericaL results aRe LEss comPEtitive THAn With Data at 30 cm, which was hIGhLY expected. In genEral teRmS, ThE SAme ConClusions thAn In TabLE \[table\_30cM\_rGb\_REGisT\_Linear\] can be drAwn for almosT All MetricS. HOweVEr, the iNcreaSiNG in Spatial distOrtiOn because Of the rEDuction IN resoluTion is MorE noTiceABlE iN CS-BaSEd mEThOds.[]{DAta-Label="tabLe_60Cm_rGB_reGist_LINEAr"}
FiGurE \[fig\_60Cm\_s13\_RGb\_regist\_linear\] DisPlayS CloSe-ups Of the FusiOn ProduCts on tHe firSt Image from the proPoseD dataset aT a rEsOluTiOn of 60 cM Per pixEl. THe GAussian StandarD DevIaTION uSed for the simulatioN oF THe Low-resolUtion sPEcTrAL componeNtS waS $\sigMA=1.3$. in ordEr to HElP the visuAl analYSiS, FIgure \[fiG\_60cM\_s13\_RGB\_rEgIst\_LinEar\_diF\] ShowS the diFference ImageS Between the refeREnce one and the REsULT pROvidEd bY each method. the fIRst cOnclUSiOn tHAt can Be draWn FRoM The visual inspection Is The supEriorIty of the variaTional modeLS UNder compArisON, pARticularly NLV aNd NLVd, over CS-basED as well aS MRA-bAsed techNiques. AltHOUgh P+XS peRfoRmeD weLl fROM a Quantitive poiNT Of viEw, The pansHarPened imAge ShoWs sOme AnNoying colOr spots. WE fUrThEr ObsErve tHAt CS and MrA FamIlIes Lead tO GreateR specTral AnD sPAtiAl qualiTY lOSS thaN NlV And NlVD. foR examPle, sEE hoW the colOr of the fiRepLAces In ThE fused pRoducts providEd By these two ClAssEs of apPROaches is Not preserved | 872
GLP 3 .1375 2 .305 1 1. 4 284 0.9860 0. 9 831
P + X S 1 . 45 88 1. 0546 1.3259 0.9967 0.9940
N LV [**1.2 0 73**]{ } [**0. 87 0 2**]{} [**1. 0 0 79 **]{ } [**0.9977**]{ } [**0.9959**]{} NL VD 1 .3 580 0. 9 8 2 3 1.306 2 0.9 967 0.99 45
: Qu anti tative ev aluati o n of th e fusedproduc tsonsimu l at ed da ta fro m R GBa eri al image sat reso luti o n o f 60 cm per pixe l. For theseexp erim e nts , the low- reso lu tionspectr al co mp onents were reg iste red and t hepa nch ro -spec t ral co nst rai nt fulf illed w i th$\ a l p ha _B=\alpha_G=\alpha _R = \ fr ac{1}{3} $. The nu me r ical res ul tsarel e ss co mpet i ti ve thanwith d a ta a t 30 cm ,whichwa s h igh ly ex p ecte d. Ingeneralterms , the same conc l usions than i n T a b le \[ta ble \_30cm\_RGB \_re g ist\ _lin e ar \]c an be draw nf or almost all metrics. H owever , the increasing i n spatiald i s tortionbeca u se of the reducti on in resolutio n is more noti ceable i n CS-base d methods. []{ dat a-l abe l = "t able_60cm_RGB _ r egis t_ linear" }
Figure\[f ig\ _60 cm\ _s 13\_RGB\_ regist\_ li ne ar \] di splay s close-u ps of t hefusio n produ cts o n th efi r stimage f r om t he p ro po seddat as et at a r e sol ution o f 60 cm p erp ixel .Th e Gauss ian standardde viation us ed fo r thes i mulation of the low-resolutions pectral co mpone ntswas $\sig ma= 1.3$.Ino rder t o help thevi sua l analy s i s, Fi gu re \[fig\_ 6 0 cm\ _s13\ _R GB\_ regist\ _linear\_dif\] sho w s t he difference im ages b et wee n t h e r ef e ren c e one and the re sult provi de d b y each met h od. T he firs t concl usion that ca n be draw n from th evisu a l in spection i s the su periority of th e v ariat ion al mod el s u ndercompar i son , par ticula rl y NLVand N LV D, overCS-based as well as MRA -based tech niq ues. Alth oug h P+ XS perfor medwell froma q uan titiv e p o int o f vi e w, th e pans harp e ned image sh ows s om e annoyingc o l orspots . W e furth er o bserve that CS an d MRA familieslead t o g rea t er s pe ctral and spat ial q u a lity los sthan NLV an d NLVD.Fo r exam ple, s ee how the co l o ro f thefire pla ces in th e f us e d produ ct sp rovide d by t hese t wo cla s seso f approaches is n ot pr e s erved | 872 _ _ _ __
__ _ GLP _ _ __3.1375 2.3051 __ ___ 1.4284 _ _ _ 0.9860 __ _ 0.9831 _
_ _P+XS__ _ 1.4588 _ _ 1.0546 _ _ 1.3259 _ _ _0.9967 _ _ _ 0.9940 _ _ _
_ ___NLV _ __[**1.2073**]{} __[**0.8702**]{} [**1.0079**]{} _[**0.9977**]{}_ [**0.9959**]{} _
__ NLVD_ _ 1.3580 _ _ _ 0.9823 _ 1.3062_ _ __ _0.9967_ _ _ 0.9945_ _ __
: Quantitative evaluation of_the fused products on simulated data_from RGB aerial images at_resolution_of_60 cm per pixel. For_these experiments, the low-resolution spectral components_were registered and_the panchro-spectral constraint fulfilled with $\alpha_B=\alpha_G=\alpha_R=\frac{1}{3}$. The_numerical_results are less competitive than with_data_at 30 cm, which was highly_expected._In_general terms, the same conclusions_than in Table \[table\_30cm\_RGB\_regist\_linear\] can be_drawn for almost all metrics. However, the increasing in_spatial distortion because_of the reduction in resolution_is_more_noticeable in CS-based methods.[]{data-label="table_60cm_RGB_regist_linear"}
Figure \[fig\_60cm\_s13\_RGB\_regist\_linear\] displays close-ups of the fusion_products on_the first image_from the proposed dataset at a resolution of 60 cm_per pixel. The Gaussian standard deviation used_for the simulation of the low-resolution spectral components was $\sigma=1.3$. In_order to help the visual analysis, Figure \[fig\_60cm\_s13\_RGB\_regist\_linear\_dif\]_shows the difference images between the reference_one and the result_provided_by each method. The first conclusion that_can_be_drawn_from the visual_inspection is the_superiority of the_variational models_under comparison, particularly_NLV and_NLVD, over CS-based_as_well as MRA-based techniques. Although P+XS performed well from a quantitive point of view, the pansharpened image shows_some annoying color spots. We further_observe that CS and MRA_families_lead to greater_spectral and spatial quality_loss than NLV and NLVD. For example, see_how the color of the fireplaces in the fused_products_provided by_these two classes of_approaches is_not preserved |
fixed number $H$ of steps is exceeded. Such a dynamics properly reflects the chain of contacts between companies as well as the avoidance of repeated contacts. Its main limitation is that it does not take into account possible preferential interactions between specific companies. Though preferential self-avoiding walks can be implemented in principle, we lack information about the preferential links between companies.
[**Dynamical Measurements:**]{} By simulating a large number of self-avoiding random walks initiating at the input node $i$, it is possible to estimate the transition probabilities $P_(j,i)$, parametrized by the path length $h$, between the input and output ($j$) nodes. Such probabilities supply important information about the interactions resulting from the chosen dynamics, which may or not be related to the topological connectivity quantified by the superedges. Additional measurements of the outward accessibility of the nodes complements the characterization of the dynamics of communication between the companies with respect to the uniformity of contacts. More specifically, a company with high outward accessibility $Acc\_out_{h}$ will be able to access all nodes at a distance $h$ in the shortest period of time. In addition, the robustness of the communication of a company $i$ and other companies at distance $h$ tends to increase with the accessibility of $i$.
Figure \[fig:exemp\] presents a hypothetical relationship between 11 tourism companies, where each column corresponds to a different *region*. The number of paths between these companies provides an important information which is not revealed by traditional measurements such as degrees, clustering coefficient and shortest paths. For instance, the fact that there are four paths (3 of length 3 and 1 of length 4) between companies 1 and 10 implies that searches for resources only found in the latter agency are much more likely to be well-succeeded than if the requested resource can only be found at company 11 (only one path exists between companies 1 and 11). Interestingly, though company 2 is linked to a single other company, in the same region, because it has access to company 1 through the path through companies 3 and 5, it will also have good chances of obtaining the services from company 10. However, a more complete picture of the interconnectivity and probability of visits between these companies requires the more systematic and comprehensive characterization provided by the superedges approach.
{width="1\linewidth"}\
\[fig:exemp\]
Table \[tab:ex | fixed number $ H$ of steps is exceeded. Such a moral force by rights reflect the chain of contacts between company as well as the avoidance of repeated contact. Its main limitation is that it does not carry into account possible discriminatory interactions between specific companies. Though discriminatory self - avoiding walks can be implemented in principle, we lack information about the preferential link between companies.
[ * * Dynamical Measurements :* * ] { } By simulating a big number of self - avoiding random walks initiating at the input signal node $ i$, it is possible to estimate the transition probabilities $ P_(j, i)$, parametrized by the path length $ h$, between the stimulation and output ($ j$) nodes. Such probabilities supply important information about the interactions resulting from the chosen dynamics, which may or not be related to the topological connectivity quantified by the superedges. Additional measurements of the outward accessibility of the nodes complement the word picture of the dynamics of communication between the ship's company with obedience to the uniformity of contacts. More specifically, a company with high outbound accessibility $ Acc\_out_{h}$ will be able to access all node at a distance $ h$ in the shortest period of meter. In addition, the robustness of the communication of a company $ i$ and other companies at distance $ h$ tends to increase with the accessibility of $ i$.
design \[fig: exemp\ ] presents a hypothetical relationship between 11 tourism companies, where each column correspond to a different * area *. The numeral of paths between these companies provides an important data which is not revealed by traditional measurements such as degrees, clustering coefficient and shortest paths. For instance, the fact that there are four paths (3 of distance 3 and 1 of length 4) between party 1 and 10 implies that searches for resource only find in the latter agency are a lot more likely to be well - succeeded than if the requested resource can entirely be found at caller 11 (only one way exists between companies 1 and 11). Interestingly, though company 2 is linked to a individual other company, in the same area, because it has access to company 1 through the path through companies 3 and 5, it will also have beneficial chances of obtaining the services from company 10. However, a more complete picture of the interconnectivity and probability of visits between these company requires the more systematic and comprehensive characterization provided by the superedges approach.
! [ Hypothetical web of joining between tourism companies. ] (exemp.eps " fig:"){width="1\linewidth"}\
\[fig: exemp\ ]
Table \[tab: ex | fided number $H$ of steps is exceeded. Such c dynammcs prolerly rewlects the chain of contacts bwtweeb companies as well as the avoifance of reptated contacts. Its main lliitaflon iv that it does kot take indo account posviclz preferential interactions between fpecifiv fompanies. Thoudh pgeserehnicl self-avoiding walks can be imllementtd in principle, we lack information about thf prfferential links bftween compqnief.
[**Eynamical Mexsurements:**]{} By simulatihg a large number of self-avoidivg raudom walks unutiwjing at the mnput gode $i$, it is possibla to esyimate the traksitimn probabilities $P_(j,i)$, paremetrized by the patr length $v$, yetween the input and oytput ($j$) nmdes. Wucf pdouabjlitied snpply imporfant informqtion about the inttrasnoons resultihg froi ehe chosen dynamics, which may or not be remated to the topologicao connectivity quantivied by tre superedges. Additional measurements of the outwdrd arcdssnnllith ov the nodes complements the characterization jr uhe dynamics of cjmmunicatiom heyreen the compxnies cjtg respect to the ujiformijy of xontacts. Iore specifically, a company wity high outwagd axcessibility $Acc\_ouc_{h}$ will be ayle to accexs all nodes at a distauce $h$ jn the shorhest perika of time. In addktipn, the robustness of the comiunicatioi of c companh $i$ snd otrer companles ab distance $h$ tends ho inereasa with the accessibility of $i$.
Figure \[fig:exemp\] presents a ny[otvetical xelatipnship betweeg 11 tourism comkanies, whzre eazh column borresponvs to a diffqrent *region*. Dje number of paths bqtwewn tyese coooanies providex an impogtcnt infornation which is nob revgamed by traditioucl measurements sich as dvgrxes, cjgstering coexfickeng and rhortesu pctfs. Fpr instance, the fact thaf there are four pstms (3 of lebgth 3 anq 1 of length 4) between companies 1 and 10 iopliex trat searches for resources onlg found ij tme latter agegcy qre much morz likely to be well-succeeded than if the requested resourcw can only be found st company 11 (only one path exists between compqnies 1 and 11). Intertstingly, though company 2 is lhnked to a single other company, in the same region, because it has access to conpany 1 through the patn thrmuyh eompaniqs 3 env 5, it will also hsve good chances of obtaining tie servicev yrom company 10. However, a more vooplete picturg of the interconnectivity and pronability of visits between thesr companies requires the more sywtemafic and cooprehensive characyerizacion peovided by bhe superedges aplrlach.
{width="1\linewlduh"}\
\[fig:exgmp\]
Table \[jab:ex | fixed number $H$ of steps is exceeded. dynamics reflects the of contacts between avoidance repeated contacts. Its limitation is that does not take into account possible interactions between specific companies. Though preferential self-avoiding walks can be implemented in principle, lack information about the preferential links between companies. [**Dynamical Measurements:**]{} By simulating a number self-avoiding walks at the input node $i$, it is possible to estimate the transition probabilities $P_(j,i)$, parametrized by path length $h$, between the input and output nodes. Such probabilities supply information about the interactions resulting the dynamics, which or be to the topological quantified by the superedges. Additional measurements of the outward accessibility of the nodes complements the characterization of dynamics of the companies respect the of contacts. More company with high outward accessibility $Acc\_out_{h}$ to access all nodes at a distance $h$ the shortest of time. In addition, the robustness the communication of a company $i$ and other at distance $h$ tends to increase with the accessibility of $i$. Figure \[fig:exemp\] presents a between 11 tourism companies, each column corresponds a *region*. number paths between companies provides an important information which is not revealed by traditional such as degrees, clustering coefficient and shortest paths. For instance, that are four paths of length 3 and of 4) between companies 1 implies searches found the agency are much more to be well-succeeded than if requested resource can only (only one path exists between companies 1 and Interestingly, though company 2 is linked to single other company, in the same region, because it has access to 1 through through companies 3 and 5, it will also good chances of obtaining services from company 10. However, a more complete picture the and probability visits between these requires the more and comprehensive characterization the superedges {width="1\linewidth"}\ \[fig:exemp\] Table \[tab:ex | fixed number $H$ of steps is exceEded. Such a dYnamiCs pRopErLy reFlecTs the chain of coNTactS between companies as welL as thE aVOidaNCe Of repEated coNTaCTS. ItS mAiN liMiTAtIon is ThaT it does Not take intO acCoUnt possible pREfErential inTerActions betweEn sPecifiC cOmpANies. THouGh preFerentIAl self-Avoiding wAlKS can be IMplemenTED iN priNciple, we lack inforMAtIOn about the prefErentiAl LInKS BetWeeN companies.
[**dyNamicAL MeasurEMeNTS:**]{} by sIMulating a largE number of seLF-avOiding RaNdoM Walks iNitiaTiNG at The input nodE $i$, it Is possiblE to estIMate the TRansitiOn probAbiLitIes $P_(J,I)$, pArAmeTrIZed BY tHe pATh lEngth $h$, beTwEeN the iNput AND OUtpuT ($j$) nOdes. such pRobabilities sUppLy imPOrtAnt inFormaTion AbOut thE interActioNs Resulting from thE choSen dynamiCs, wHiCh mAy Or not BE relatEd tO thE topoloGical coNNecTiVITY qUantified by the supeReDGEs. additionAl measUReMeNTs of the oUtWarD accESSibilIty oF ThE nodes coMplemeNTs ThE characTeRizatiOn Of tHe dYnamiCS of cOmmuniCation beTween THe companies witH Respect to the uNIfORMiTY of cOntActs. More speCifiCAlly, A comPAnY wiTH high OutwaRd ACcESsibility $Acc\_out_{h}$ wilL bE able tO acceSs all nodes at a Distance $h$ iN THE shortesT perIOd OF time. In additioN, the rObustness oF The commuNicatIon of a coMpany $i$ and OTHer compaNieS at DisTanCE $H$ tEnds to increasE WIth tHe AccessiBilIty of $i$.
FIguRe \[fIg:eXemP\] pResents a hYpothetiCaL rElAtIonShip bETween 11 touRiSm cOmPanIes, whERe each ColumN corReSpONds To a diffEReNT *RegiOn*. thE numBer Of Paths BetwEEn tHese comPanies proVidES an iMpOrTant infOrmation which Is Not revealeD bY trAditioNAL measureMents such as degrees, clustERing coeFfiCient And sHortest paThs. for insTanCE, the faCt that There ArE foUR Paths (3 OF LeNgtH 3 aNd 1 of length 4) BETweEn comPaNies 1 And 10 implIes that searches for REsoUrces only founD in The lATTeR agENcY Are MuCH moRE Likely to be well-sUcceeded thAn IF tHe requesteD ResOuRce can oNly be foUnd at COmpany 11 (oNly one patH exists beTwEen cOMPanIes 1 and 11). InteRestinglY, though coMPany 2 iS LiNked tO a sIngle oThEr cOmpanY, in the SAme RegioN, becauSe It has aCcess To Company 1 tHrough the path through comPanies 3 And 5, it WilL also have GooD ChaNces of obtAiniNg the serviCes FroM compAny 10. hOweveR, a moRE cOmpLEte piCturE Of the inteRCoNneCTIvIty and probaBILIty Of visIts BEtween ThesE companies requireS The more systemaTic aND ComPreHEnsiVe CharacterizatiOn pRoVIDed by the SuPeredges appRoach.
{WIdth="1\linEwIdTH"}\
\[fig:exEmp\]
TAbLe \[tab:eX | fixed number $H$ of steps is exceed ed. S uch ady nami cs p roperly reflec t s th e chain of contacts be tween c o mpan i es as w ell ast he a voi da nc e o fr ep eated co ntacts. Its mainlim it ation is tha t i t does not ta ke into acco unt possi bl e p r efere nti al in teract i ons be tween spe ci f ic com p anies.T h ou gh p referential self- a vo i ding walks can be im pl e me n t edinprinciple, w e lac k inform a ti o n abo u t the prefere ntial links bet ween c om pan i es.
[ **Dyn am i cal Measuremen ts:* *]{} By s imulat i ng a la r ge numb er ofsel f-a void i ng r and om wal k sini t iat ing at t he i nputnode $ i $ , it is pos sible to estimatethe tra n sit ion p robab ilit ie s $P_ (j,i)$ , par am etrized by thepath length $ h$, b etw ee n the inputand ou tput ($ j$) nod e s.Su c h pr obabilities supply i m p or tant inf ormati o nab o ut the i nt era ctio n s resu ltin g f rom thechosen dy na mics, w hi ch may o r n otbe re l ated to th e topolo gical connectivity q u antified by t h es u pe r edge s.Additionalmeas u reme ntso fthe outwa rd ac ce s si b ility of the nodesco mpleme nts t he characteri zation oft h e dynamic s of co m munication bet weenthe compan i es withrespe ct to th e uniform i t y of con tac ts. Mo res p ec ifically, a c o m pany w ith hig h o utwardacc ess ibi lit y$Acc\_out _{h}$ wi ll b eab leto ac c ess allno des a t a dist a nce $h $ inthesh or t est period of t ime. I naddi tio n, therobu s tne ss of t he commun ica t ionof a compan y $i$ and oth er companies a t d istanc e $h$ tend s to increase with thea ccessib ili ty of $i$ .
Figure \[ fig:ex emp \ ] pres ents a hypo th eti c a l rel a t io nsh ip between 1 1 tou rismco mpan ies, wh ere each column co r res ponds to a di ffe rent * re gio n *. The n u mbe r of paths betwee n these co mp a ni es provide s an i mportan t infor matio n whichis not re vealed by t radi t i ona l measurem ents suc h as degr e es, c l us terin g c oeffic ie ntand s hortes t pa ths.For in st ance,the f ac t that t here are four paths (3of len gth 3 an d 1 of le ngt h 4) betweencomp anies 1 an d 1 0 i mplie s t h at se arch e sfor resou rces only foun d i n t h e l atter agenc y a remuchmor e likel y to be well-succeede d than if the r eque s t edres o urce c an only be fou ndat c ompany 1 1(only one p ath exis ts betwe en com panies 1 and1 1 ). Intere stin gly , thoughcom pa n y 2 isli nk e d to a sin gl e othe r comp a ny,i n the same region , bec a u se it has acce ss to com p any1 throughthe path th roughcomp anies 3 and5, it wi llal so have go o d chances of o btainin gtheser vicesfrom c ompan y 10 .How ever, a m o r ec om pl e tepict ure o ftheinterconn e ctivityand probabi li tyo f visit s b e t ween these co mpani e s requirest he m o re syste maticand co mprehen s ive c haracte riz a t ion provi ded by th e su pe redg es appro ac h.
 {w i dth="1\ l i new idth"}\
\ [fi g:e xemp \]
Tab l e \[tab:ex | fixed_number $H$_of steps is exceeded._Such a_dynamics_properly reflects_the_chain of contacts_between companies as_well as the avoidance_of repeated contacts._Its_main limitation is that it does not take into account possible preferential interactions between_specific_companies. Though_preferential_self-avoiding_walks can be implemented in_principle, we lack information about_the preferential_links between companies.
[**Dynamical Measurements:**]{} By simulating a large_number_of self-avoiding random_walks initiating at the input node $i$, it is_possible to estimate the transition probabilities_$P_(j,i)$, parametrized by_the_path_length $h$, between the_input and output ($j$) nodes. Such_probabilities supply important information about the_interactions resulting from the chosen dynamics, which_may or not be related to_the topological connectivity quantified by_the superedges._Additional measurements of the outward_accessibility of the_nodes complements_the characterization of_the dynamics of communication between the_companies with respect_to the uniformity of contacts. More_specifically,_a company with_high_outward_accessibility $Acc\_out_{h}$_will be able_to_access all_nodes_at a distance $h$ in the_shortest_period of time. In addition, the robustness_of the communication of_a_company $i$ and other_companies at distance $h$ tends_to increase with the accessibility of_$i$.
Figure \[fig:exemp\] presents_a hypothetical_relationship between 11 tourism companies, where each column corresponds to a_different *region*. The number of paths_between these companies provides_an important_information_which is not_revealed_by traditional_measurements such as degrees, clustering coefficient and_shortest paths._For instance, the fact that there_are four paths (3_of_length 3 and 1 of length_4) between companies 1 and 10_implies that searches for resources_only_found_in the latter agency are_much more likely to be well-succeeded_than if the_requested resource can only be found at_company_11 (only one path exists between_companies_1 and 11). Interestingly, though company_2_is_linked to a single other_company, in the same region, because_it has access to company 1 through the path_through companies 3_and 5, it will also_have_good_chances of obtaining the services from company 10. However, a_more complete_picture of the_interconnectivity and probability of visits between these companies requires the_more systematic and comprehensive characterization provided by_the superedges approach.
{width="1\linewidth"}\
\[fig:exemp\]
Table \[tab:ex |
the imaginary $k_i$ axes, but not all of them are located on the positive imaginary axis. Finally, we call the solution a *resonance* if the zero is not lying on any of the imaginary $k_i$ axes, hence the corresponding energy is complex or if real then it is above at least one of the thresholds.
The derivation of the Cox potential and how to solve it exactly are given in [@pup08; @spa06; @sam07]. To make the paper self-contained we collect some formulas. The Cox [@1964Cox] interaction matrix $$\label{cox_v}
V(r) = \left(
\begin{array}{cc}
v_{11}(r) & v_{12}(r) \\
v_{21(r)} & v_{22}(r) \\
\end{array}
\right)~.$$ is given by $$V(r)=-{\cal K}+2 {\cal K}^{1/2}(I+X(r))^{-1}
{\cal K}^{1/2},$$ where $I$ is the 2x2 unit matrix and $$\begin{aligned}
&X(r)={\rm diag}(\exp(-\kappa_1r),\exp(-\kappa_2 r)) X_0\nonumber\\
&\times {\rm diag} (\exp(-\kappa_1 r),\exp(-\kappa_2 r)).\end{aligned}$$
The symmetric 2x2 matrix $X_0$ contains the parameters of the potential. The factorization wave numbers $\kappa_1$ and $\kappa_2$ are positive parameters and they satisfy the condition $\kappa_2^2-\kappa_1^2=\Delta.$ (we will use $\kappa_1=\kappa$ as independent parameter). The matrix $\cal K$ is related to the factorization energies by ${\cal K}={\rm diag}(\kappa_1,\kappa_2)$. Of course, the interaction matrix is symmetric.
The determinant of the Jost matrix is given by [@pup08] $$\label{eq.f}
f(k,p)=\frac{(k+i\, \alpha_1)(p+i\, \alpha_2)+\beta^2}{(k+i\, \kappa_1)(p+i\, \kappa_2)}.$$ Here we denote for convenience the channel wave numbers as $k_1=k$ and $k_2=p | the imaginary $ k_i$ axes, but not all of them are located on the positive fanciful bloc. last, we call the solution a * plangency * if the zero is not lie on any of the imaginary $ k_i$ axes, therefore the corresponding energy is complex or if real then it is above at least one of the threshold.
The derivation of the Cox potential and how to clear it exactly are given in [ @pup08; @spa06; @sam07 ]. To seduce the paper self - contained we collect some formulas. The Cox [ @1964Cox ] interaction matrix $ $ \label{cox_v }
V(r) = \left (
\begin{array}{cc }
v_{11}(r) & v_{12}(r) \\
v_{21(r) } & v_{22}(r) \\
\end{array }
\right)~.$$ is given by $ $ V(r)=-{\cal K}+2 { \cal K}^{1/2}(I+X(r))^{-1 }
{ \cal K}^{1/2},$$ where $ I$ is the 2x2 unit matrix and $ $ \begin{aligned }
& X(r)={\rm diag}(\exp(-\kappa_1r),\exp(-\kappa_2 roentgen) ) X_0\nonumber\\
& \times { \rm diag } (\exp(-\kappa_1 r),\exp(-\kappa_2 r)).\end{aligned}$$
The symmetric 2x2 matrix $ X_0 $ contains the parameters of the electric potential. The factorization wave numbers $ \kappa_1 $ and $ \kappa_2 $ are positive parameter and they satisfy the circumstance $ \kappa_2 ^ 2-\kappa_1 ^ 2=\Delta.$ (we will use $ \kappa_1=\kappa$ as autonomous parameter). The matrix $ \cal K$ is related to the factorization energies by $ { \cal K}={\rm diag}(\kappa_1,\kappa_2)$. Of course, the interaction matrix is symmetrical.
The determinant of the Jost matrix is given by [ @pup08 ] $ $ \label{eq.f }
f(k, p)=\frac{(k+i\, \alpha_1)(p+i\, \alpha_2)+\beta^2}{(k+i\, \kappa_1)(p+i\, \kappa_2)}.$$ Here we denote for convenience the channel wave numbers as $ k_1 = k$ and $ k_2 = p | thf imaginary $k_i$ axes, but kot all of them cee loceted on the posktive imaginary axis. Finally, ww calo the solution a *resonxnce* if tje zero us nir lying on any of bke imzninarv $j_i$ axes, hence jhe correspotding energy iv zolplex or if real then it is above ae least oje of the threfholcf.
The derivation of the Cox potential ahd how uo solve it exactlu are given in [@pup08; @spa06; @sam07]. To lake the paper selv-contained qe cjolect some furmulas. Tht Eox [@1964Cox] intgraction matrix $$\label{cox_v}
V(r) = \lewt(
\bzgin{array}{cc}
v_{11}(g) & v_{12}(r) \\
v_{21(r)} & v_{22}(r) \\
\end{array}
\vpght)~.$$ is given ny $$V(r)=-{\cal K}+2 {\cal K}^{1/2}(I+E(r))^{-1}
{\cao K}^{1/2},$$ where $I$ is the 2x2 nnit matrix and $$\begig{aligned}
&X(s)={\rj diag}(\exp(-\kappa_1r),\exp(-\kqppa_2 t)) X_0\notumbde\\
&\tioes {\rk siag} (\edp(-\keppa_1 r),\exp(-\kalpa_2 r)).\end{alitned}$$
The symmetric 2x2 mwntix $X_0$ contaihs the pwrameters of the potential. The factorizdtikn wave numbers $\kappa_1$ abd $\kappa_2$ are positive parametews and they satisfy the condition $\kappa_2^2-\kappa_1^2=\Delta.$ (we wmlu uwe $\yqpoa_1=\kappa$ as independent parameter). The matrix $\cwm L$ ps related to the factorizatooj rgergies by ${\cau K}={\rm bjaf}(\kappa_1,\kappa_2)$. Of cougse, the inteeaction mwtric is symmetric.
The determinabt of the Jofr matrix is given yy [@pup08] $$\label{zq.f}
f(k,k)=\frac{(k+o\, \alpha_1)(p+i\, \alpha_2)+\beta^2}{(k+i\, \kcppa_1)(p+i\, \kappa_2)}.$$ Here we denots for convenience thv chdnnel wave numbers as $k_1=k$ agd $k_2=p | the imaginary $k_i$ axes, but not all are on the imaginary axis. Finally, *resonance* the zero is lying on any the imaginary $k_i$ axes, hence the energy is complex or if real then it is above at least one the thresholds. The derivation of the Cox potential and how to solve it are in @spa06; To make the paper self-contained we collect some formulas. The Cox [@1964Cox] interaction matrix $$\label{cox_v} V(r) \left( \begin{array}{cc} v_{11}(r) & v_{12}(r) \\ v_{21(r)} & \\ \end{array} \right)~.$$ is by $$V(r)=-{\cal K}+2 {\cal K}^{1/2}(I+X(r))^{-1} K}^{1/2},$$ $I$ is 2x2 matrix $$\begin{aligned} &X(r)={\rm diag}(\exp(-\kappa_1r),\exp(-\kappa_2 X_0\nonumber\\ &\times {\rm diag} (\exp(-\kappa_1 r),\exp(-\kappa_2 r)).\end{aligned}$$ The symmetric 2x2 matrix $X_0$ contains the parameters of the The factorization $\kappa_1$ and are parameters they satisfy the (we will use $\kappa_1=\kappa$ as independent $\cal K$ is related to the factorization energies ${\cal K}={\rm Of course, the interaction matrix is The determinant of the Jost matrix is given [@pup08] $$\label{eq.f} f(k,p)=\frac{(k+i\, \alpha_1)(p+i\, \alpha_2)+\beta^2}{(k+i\, \kappa_1)(p+i\, \kappa_2)}.$$ Here we denote for convenience the channel wave $k_1=k$ and $k_2=p | the imaginary $k_i$ axes, but not aLl of them arE locaTed On tHe PosiTive Imaginary axis. FINallY, we call the solution a *resOnancE* iF The zERo Is not Lying on ANy OF The ImAgInaRy $K_I$ aXes, heNce The corrEsponding eNerGy Is complex or iF ReAl then it is AboVe at least one Of tHe threShOldS.
the deRivAtion Of the COX potenTial and hoW tO Solve iT Exactly ARE gIven In [@pup08; @spa06; @sam07]. To make THe PAper self-contaiNed we cOlLEcT SOme ForMulas. The CoX [@1964COx] intERaction MAtRIX $$\LabEL{cox_v}
V(r) = \left(
\beGin{array}{cc}
v_{11}(R) & V_{12}(r) \\
v_{21(R)} & v_{22}(r) \\
\end{ArRay}
\RIght)~.$$ is Given By $$v(R)=-{\caL K}+2 {\cal K}^{1/2}(I+X(r))^{-1}
{\caL K}^{1/2},$$ whEre $I$ is the 2X2 unit mATrix and $$\BEgin{aliGned}
&X(r)={\Rm dIag}(\Exp(-\kAPpA_1r),\Exp(-\KaPPa_2 r)) x_0\NoNumBEr\\
&\tImes {\rm diAg} (\ExP(-\kappA_1 r),\exP(-\KAPPa_2 r)).\eNd{aLignEd}$$
The Symmetric 2x2 matRix $x_0$ conTAinS the pArameTers Of The poTentiaL. The fAcTorization wave nUmbeRs $\kappa_1$ anD $\kaPpA_2$ arE pOsitiVE paramEteRs aNd they sAtisfy tHE coNdITIOn $\Kappa_2^2-\kappa_1^2=\Delta.$ (we wIlL USe $\Kappa_1=\kapPa$ as inDEpEnDEnt paramEtEr). THe maTRIx $\cal k$ is rELaTed to the FactorIZaTiOn energIeS by ${\cal k}={\rM diAg}(\kAppa_1,\kAPpa_2)$. OF coursE, the inteRactiON matrix is symmeTRic.
The determiNAnT OF tHE JosT maTrix is given By [@puP08] $$\LabeL{eq.f}
F(K,p)=\FraC{(K+i\, \alpHa_1)(p+i\, \aLpHA_2)+\bETa^2}{(k+i\, \kappa_1)(p+i\, \kappa_2)}.$$ HerE wE denotE for cOnvenience the Channel wavE NUMbers as $k_1=K$ and $K_2=P | the imaginary $k_i$ axes, but not a ll of th emar e lo cate d on the posit i ve i maginary axis. Finally , weca l l th e s oluti on a *r e so n a nce *if th ez er o isnot lyingon any ofthe i maginary $k_ i $axes, henc e t he correspon din g ener gy is compl exor if realt hen it is above a t least one oft h ethre sholds.
The deri v at i on of the Coxpotent ia l a n d ho w t o solve it e xactl y are gi v en i n [@ p up08; @spa06; @sam07]. T o ma ke the p ape r self- conta in e d w e collect s omeformulas. The C o x [@196 4 Cox] in teract ion ma trix $$ \l abe l{ c ox_ v } V( r ) = \left( \b egin{ arra y } { c c}
v_{1 1}(r) & v_{12}(r)\\ v _{2 1(r)} & v_ {22} (r ) \\ \end {arra y}
\right)~.$$is g iven by $ $V( r) =-{ \c al K} + 2 {\ca l K }^{ 1/2}(I+ X(r))^{ - 1}{\ c a l K }^{1/2},$$ where $ I$ i sthe 2x2unit m a tr ix and $$\b eg in{ alig n e d}
&X (r)= { \r m diag}( \exp(- \ ka pp a_1r),\ ex p(-\ka pp a_2 r) ) X_0 \ nonu mber\\
&\times {\rm diag} (\exp(-\ k appa_1 r),\ex p (- \ k ap p a_2r)) .\end{align ed}$ $
Th e sy m me tri c 2x2matri x$ X_ 0 $ contains the para me ters o f the potential. T he factori z a t ion wave num b er s $\kappa_1$ an d $\k appa_2$ ar e positiv e par ametersand theys a tisfy th e c ond iti on$ \ ka ppa_2^2-\kapp a _ 1^2= \D elta.$(we will u se$\k app a_1 =\ kappa$ as indepen de nt p ar ame ter). The matr ix $\ ca l K $ isr elated to t he f ac to r iza tion en e rg i e s by $ {\ calK}= {\ rm di ag}( \ kap pa_1,\k appa_2)$. Of cour se ,the int eraction matr ix is symmet ri c.
The d e t erminant of the Jost matrix isg iven by [@ pup08 ] $$ \label{eq .f}
f(k, p)= \ frac{( k+i\,\alph a_ 1)( p + i\, \ a l ph a_2 )+ \beta^2}{( k + i\, \kap pa _1)( p+i\, \ kappa_2)}.$$ Herew e d enote for con ven ienc e th e c h an n elwa v e n u m bers as $k_1=k$ and $k_2= p | the_imaginary $k_i$_axes, but not all_of them_are_located on_the_positive imaginary axis._Finally, we call_the solution a *resonance*_if the zero_is_not lying on any of the imaginary $k_i$ axes, hence the corresponding energy is_complex_or if_real_then_it is above at least_one of the thresholds.
The derivation_of the_Cox potential and how to solve it exactly_are_given in [@pup08;_@spa06; @sam07]. To make the paper self-contained we collect_some formulas. The Cox [@1964Cox] interaction_matrix $$\label{cox_v}
V(r)_=_\left(
_ \begin{array}{cc}
_ v_{11}(r) & v_{12}(r) \\
_ v_{21(r)} & v_{22}(r) \\
_ \end{array}
\right)~.$$ is given by_$$V(r)=-{\cal K}+2 {\cal K}^{1/2}(I+X(r))^{-1}
{\cal K}^{1/2},$$ where_$I$ is the 2x2 unit_matrix and_$$\begin{aligned}
&X(r)={\rm diag}(\exp(-\kappa_1r),\exp(-\kappa_2 r)) X_0\nonumber\\
&\times {\rm_diag} (\exp(-\kappa_1 r),\exp(-\kappa_2_r)).\end{aligned}$$
The symmetric_2x2 matrix $X_0$_contains the parameters of the potential._The factorization wave_numbers $\kappa_1$ and $\kappa_2$ are positive_parameters_and they satisfy_the_condition_$\kappa_2^2-\kappa_1^2=\Delta.$ (we_will use $\kappa_1=\kappa$_as_independent parameter)._The_matrix $\cal K$ is related to_the_factorization energies by ${\cal K}={\rm diag}(\kappa_1,\kappa_2)$. Of_course, the interaction matrix_is_symmetric.
The determinant of the_Jost matrix is given by_[@pup08] $$\label{eq.f}
f(k,p)=\frac{(k+i\, \alpha_1)(p+i\, \alpha_2)+\beta^2}{(k+i\, \kappa_1)(p+i\,_\kappa_2)}.$$ Here_we denote_for convenience the channel wave numbers as $k_1=k$ and $k_2=p |
$. One is $\beta_1=\beta_2=\beta_3=1$, and the corresponding mean vectors are generally selected as ${{\bm \mu}}_1=1, {{\bm \mu}}_2=1, {{\bm \mu}}_3=-2$. The other one is $\beta_1=\beta_2=-\frac{1}{2}, \beta_3=1$, and the corresponding mean vectors are given as ${{\bm \mu}}_1=1, {{\bm \mu}}_2=3, {{\bm \mu}}_3=2$ without loss of generality. The alternative hypotheses are designed that the ${{\bm \mu}}_3$ is the value under the null hypothesis added $\Delta{{\bm \mu}}$ described as above, while ${{\bm \mu}}_1$ and ${{\bm \mu}}_2$ remain unchanged.
For each set of the scenarios, we report both empirical Type I errors and powers with 10,000 replications at $\alpha=0.05$ significance level. Different pair values of $p, n_1, n_2, n_3$ are selected, and $\epsilon$ is varying from 0 to 0.9 or 1 to show the empirical sizes and powers. The mean parameter is supposed to be unknown and substituted by the sample mean during the calculations. Simulation results of empirical Type I errors and powers for the three group tests are listed in the Table \[tab:1\] and Table \[tab:2\]. Simulation results of empirical Type I errors and powers for the two-sample test are represented in Table \[tab:3\].
[14cm]{}[ccXXXX]{} [$(p, n_1, n_2, n_3)$ ]{} & & &\
&&Ours & TNT&Ours & TNT\
& & &\
Normal & $\epsilon=0$ (size) &0.0647 &0.0697 &0.0614 &0.0707\
& 0.2 &0.1056 &0.0772 &0.1059 &0.0721\
& 0.4 &0.2568 &0.0872 &0.3042 &0.0855\
& 0.6 &0.5795 &0.1235 &0.6888 &0.1181\
& 0.8 &0.8739 &0.2055 &0.9496 &0.2038\
| $. One is $ \beta_1=\beta_2=\beta_3=1 $, and the corresponding mean vectors are generally selected as $ { { \bm \mu}}_1=1, { { \bm \mu}}_2=1, { { \bm \mu}}_3=-2$. The early one is $ \beta_1=\beta_2=-\frac{1}{2 }, \beta_3=1 $, and the comparable mean vector are given as $ { { \bm \mu}}_1=1, { { \bm \mu}}_2=3, { { \bm \mu}}_3=2 $ without personnel casualty of generalization. The alternative hypotheses are designed that the $ { { \bm \mu}}_3 $ is the value under the null guess add $ \Delta{{\bm \mu}}$ described as above, while $ { { \bm \mu}}_1 $ and $ { { \bm \mu}}_2 $ stay unchanged.
For each set of the scenario, we report both empirical Type I error and powers with 10,000 replications at $ \alpha=0.05 $ significance level. Different couple values of $ p, n_1, n_2, n_3 $ are selected, and $ \epsilon$ is deviate from 0 to 0.9 or 1 to show the empirical size and powers. The average parameter is supposed to be unknown and substituted by the sample think of during the calculations. Simulation results of empirical Type I errors and powers for the three group tests are listed in the Table \[tab:1\ ] and Table \[tab:2\ ]. Simulation results of empirical Type I errors and powers for the two - sample test are represent in Table \[tab:3\ ].
[ 14cm]{}[ccXXXX ] { } [ $ (p, n_1, n_2, n_3)$ ] { } & & & \
& & Ours & TNT&Ours & TNT\
& & & \
Normal & $ \epsilon=0 $ (size) & 0.0647 & 0.0697 & 0.0614 & 0.0707\
& 0.2 & 0.1056 & 0.0772 & 0.1059 & 0.0721\
& 0.4 & 0.2568 & 0.0872 & 0.3042 & 0.0855\
& 0.6 & 0.5795 & 0.1235 & 0.6888 & 0.1181\
& 0.8 & 0.8739 & 0.2055 & 0.9496 & 0.2038\ | $. Onf is $\beta_1=\beta_2=\beta_3=1$, and tht corresponding mgab vectmrs ars generauly selected as ${{\bm \mu}}_1=1, {{\bm \mu}}_2=1, {{\bn \mu}}_3=-2$. Uke other one is $\beta_1=\ceta_2=-\frac{1}{2}, \heta_3=1$, and the xorrespondmhg mean vectodd arz jiven as ${{\bm \mu}}_1=1, {{\nm \mu}}_2=3, {{\bm \mg}}_3=2$ without loss ow yenerality. The alternative hypothesef are drslgned that the ${{\bm \kt}}_3$ is nht value under the null hypothesis added $\Velta{{\bm \mu}}$ desctibed as above, while ${{\bm \mu}}_1$ and ${{\bm \mu}}_2$ remain unchwnged.
For eaxh sqr of the scevarios, we gzport both gmpirical Type I errors and powefs wich 10,000 replicajnins dt $\alpha=0.05$ sijnificwnce level. Dlgferend pair falues of $p, n_1, k_2, n_3$ ace swlected, and $\epsilon$ iv varying from 0 to 0.9 or 1 to vhkw the empirical wizes atd pmwerr. Thd mtan pzrametfr ms supposed to be unkniwn and substituted br the sample msan duwigg the calculations. Simulation results mf smpirical Type I errors and powers for the tjree grou[ tests are listed in the Table \[tab:1\] and Table \[tab:2\]. Simumxtiin fwsklts of empirical Type I errors and powers fow tne two-sample tesb are represented on Tsfle \[tab:3\].
[14cm]{}[ccXXBX]{} [$(p, n_1, n_2, h_3)$ ]{} & & &\
&&Ours & TNT&Ours & TNT\
& & &\
Gormao & $\epsilog=0$ (siae) &0.0647 &0.0697 &0.0614 &0.0707\
& 0.2 &0.1056 &0.0772 &0.1059 &0.0721\
& 0.4 &0.2568 &0.0872 &0.3042 &0.0855\
& 0.6 &0.5795 &0.1235 &0.6888 &0.1181\
& 0.8 &0.8739 &0.2055 &0.9496 &0.2038\
| $. One is $\beta_1=\beta_2=\beta_3=1$, and the corresponding are selected as \mu}}_1=1, {{\bm \mu}}_2=1, is \beta_3=1$, and the mean vectors are as ${{\bm \mu}}_1=1, {{\bm \mu}}_2=3, {{\bm without loss of generality. The alternative hypotheses are designed that the ${{\bm \mu}}_3$ the value under the null hypothesis added $\Delta{{\bm \mu}}$ described as above, while \mu}}_1$ ${{\bm remain For each set of the scenarios, we report both empirical Type I errors and powers with replications at $\alpha=0.05$ significance level. Different pair values $p, n_1, n_2, n_3$ selected, and $\epsilon$ is varying 0 0.9 or to the sizes and powers. mean parameter is supposed to be unknown and substituted by the sample mean during the calculations. Simulation of empirical errors and for three tests are listed Table \[tab:1\] and Table \[tab:2\]. Simulation Type I errors and powers for the two-sample are represented Table \[tab:3\]. [14cm]{}[ccXXXX]{} [$(p, n_1, n_2, ]{} & & &\ &&Ours & TNT&Ours & & & &\ Normal & $\epsilon=0$ (size) &0.0647 &0.0697 &0.0614 &0.0707\ & 0.2 &0.1056 &0.0772 & 0.4 &0.2568 &0.0872 &0.0855\ & 0.6 &0.1235 &0.1181\ 0.8 &0.2055 &0.9496 | $. One is $\beta_1=\beta_2=\beta_3=1$, and the coRrespondinG mean VecTorS aRe geNeraLly selected as ${{\bM \Mu}}_1=1, {{\bm \Mu}}_2=1, {{\bm \mu}}_3=-2$. The other one is $\betA_1=\beta_2=-\FrAC{1}{2}, \betA_3=1$, AnD the cOrrespoNDiNG MeaN vEcTorS aRE gIven aS ${{\bm \Mu}}_1=1, {{\bm \mu}}_2=3, {{\bM \mu}}_3=2$ without LosS oF generality. THE aLternative HypOtheses are deSigNed thaT tHe ${{\bM \Mu}}_3$ is tHe vAlue uNder thE Null hyPothesis aDdED $\Delta{{\BM \mu}}$ descRIBeD as aBove, while ${{\bm \mu}}_1$ and ${{\bM \Mu}}_2$ REmain unchanged.
for eacH sET oF THe sCenArios, we repOrT both EMpiricaL tyPE i ErrORs and powers wiTh 10,000 replicatiONs aT $\alpha=0.05$ SiGniFIcance Level. diFFerEnt pair valuEs of $P, n_1, n_2, n_3$ are seLected, ANd $\epsilON$ is varyIng froM 0 to 0.9 Or 1 tO shoW ThE eMpiRiCAl sIZeS anD PowErs. The meAn PaRametEr is SUPPOsed To bE unkNown aNd substituted By tHe saMPle Mean dUring The cAlCulatIons. SiMulatIoN results of empirIcal type I erroRs aNd PowErS for tHE three GroUp tEsts are Listed iN The taBLE \[TaB:1\] and Table \[tab:2\]. SimulaTiON ReSults of eMpiricAL TYpE i errors aNd PowErs fOR The twO-samPLe Test are rEpreseNTeD iN Table \[tAb:3\].
[14Cm]{}[ccXXxX]{} [$(P, n_1, n_2, N_3)$ ]{} & & &\
&&OuRs & TNT&oUrs & TnT\
& & &\
NormAl & $\epsiloN=0$ (size) &0.0647 &0.0697 &0.0614 &0.0707\
& 0.2 &0.1056 &0.0772 &0.1059 &0.0721\
& 0.4 &0.2568 &0.0872 &0.3042 &0.0855\
& 0.6 &0.5795 &0.1235 &0.6888 &0.1181\
& 0.8 &0.8739 &0.2055 &0.9496 &0.2038\
| $. One is $\beta_1=\beta_2 =\beta_3=1 $, an d t heco rres pond ing mean vecto r s ar e generally selected a s ${{ \b m \mu } }_ 1=1,{{\bm \ m u} } _ 2=1 , { {\b m\ mu }}_3= -2$ . The o ther one i s $ \b eta_1=\beta_ 2 =- \frac{1}{2 },\beta_3=1$,and the c or res p ondin g m ean v ectors are gi ven as ${ {\ b m \mu} } _1=1, { { \ bm \mu }}_2=3, {{\bm \m u }} _ 3=2$ without l oss of g e ne r a lit y.The altern at ive h y pothese s a r e des i gned that the ${{\bm \mu } }_3 $ is t he va l ue und er th en ull hypothesis add ed $\Delt a{{\bm \mu}}$d escribe d as a bov e,whil e $ {{ \bm \ m u}} _ 1$ an d ${ {\bm \mu }} _2 $ rem ainu n c h ange d.
For each set of the s cen ario s , w e rep ort b othem piric al Typ e I e rr ors and powerswith 10,000 r epl ic ati on s at$ \alpha =0. 05$ signif icancel eve l. D i ff erent pair valuesof $ p, n_1, n_ 2, n_3 $ a re selected ,and $\e p s ilon$ isv ar ying fro m 0 to 0. 9or 1 to s how th eemp iri cal s i zesand po wers. Th e mea n parameter iss upposed to be un k n ow n and su bstituted b y th e sam plem ea n d u ringthe c al c ul a tions. Simulation r es ults o f emp irical Type I errors an d p owers fo r th e t h ree group test s are listed in the Tabl e \[t ab:1\] a nd Table\ [ tab:2\]. Si mul ati onr e su lts of empiri c a l Ty pe I erro rsand pow ers fo r t hetw o-sampletest are r ep re se nte d inT able \[t ab :3\ ].
[ 14cm] { }[ccXX XX]{} [$( p, n _ 1,n_2, n_ 3 )$ ] {} & & & \
&& Our s& TNT &Our s &TNT\
&& &\
Norm al& $\e ps il on=0$ ( size) &0.0647 & 0.0697 &0. 06 14&0.070 7 \
& 0.2 & 0.1056 &0.0772 &0.1059& 0.0721\
&0.4 & 0.25 68 &0.087 2 & 0.3042 &0 . 0855\& 0.6&0.57 95 &0 . 1 235 & 0 . 68 88&0 .1181\
& 0 . 8 &0 .8739 & 0.20 55 &0.9 496 &0.2038\
| $. One_is $\beta_1=\beta_2=\beta_3=1$,_and the corresponding mean_vectors are_generally_selected as_${{\bm_\mu}}_1=1, {{\bm \mu}}_2=1,_ {{\bm \mu}}_3=-2$._The other one is_$\beta_1=\beta_2=-\frac{1}{2}, \beta_3=1$, and_the_corresponding mean vectors are given as ${{\bm \mu}}_1=1, {{\bm \mu}}_2=3, {{\bm \mu}}_3=2$ without_loss_of generality._The_alternative_hypotheses are designed that the_${{\bm \mu}}_3$ is the value_under the_null hypothesis added $\Delta{{\bm \mu}}$ described as above,_while_${{\bm \mu}}_1$ and_${{\bm \mu}}_2$ remain unchanged.
For each set of the scenarios,_we report both empirical Type I_errors and powers_with_10,000_replications at $\alpha=0.05$ significance_level. Different pair values of $p,_n_1, n_2, n_3$ are selected, and_$\epsilon$ is varying from 0 to 0.9_or 1 to show the empirical_sizes and powers. The mean_parameter is_supposed to be unknown and_substituted by the_sample mean_during the calculations._Simulation results of empirical Type I_errors and powers_for the three group tests are_listed_in the Table_\[tab:1\]_and_Table \[tab:2\]._Simulation results of_empirical_Type I_errors_and powers for the two-sample test_are_represented in Table \[tab:3\].
[14cm]{}[ccXXXX]{} [$(p, n_1, n_2,_n_3)$ ]{} & &_&\
&&Ours_& TNT&Ours & TNT\
&_& &\
Normal & $\epsilon=0$ (size)_&0.0647 &0.0697 &0.0614 &0.0707\
& 0.2 &0.1056_&0.0772 &0.1059_&0.0721\
& 0.4_&0.2568 &0.0872 &0.3042 &0.0855\
& 0.6 &0.5795 &0.1235 &0.6888 &0.1181\
& 0.8 &0.8739_&0.2055 &0.9496 &0.2038\
|
model.
First inelastic channel
-----------------------
The last point to be presented in this work concerns the reaction $$\begin{aligned}
\mbox{p+$^3$H} &\rightarrow& \mbox{p+$^3$H} \cr
&\rightarrow& \mbox{n+$^3$He} \cr
&\rightarrow& \mbox{d+d} \end{aligned}$$ In the isospin approximation employed throughout this work, both n+$^3$He and p+t thresholds are degenerate and correspond to the N+NNN of the preceding section. The Pauli principle imposes for the deuteron-deuteron channel $(-)^{L+S+T}=+1$. If we assume the final state d+d to be in a relative S-wave one has $L=T=0$ and $J^\pi=0^+,2^+$. The $J^\pi=2^+$ state requires a relative angular momentum $l_z=2$ in the initial N+NNN channel and it is expected to give small contribution at very low d+d energy. We will consider only the $(J^\pi=0^+, T=0)$ state.
The S-matrix is defined according to (\[smatrix\]). We present in Table \[Smat\] the S-matrix elements at several energies, chosen with respect to the NN+NN threshold. N+NNN is referred to as channel 1, NN+NN as channel 2. The symmetry and unitarity properties are there fulfilled at the level of $10^{-3}$, what corresponds to the accuracy of our results.
With the conventions used above, the scattering cross sections in a given partial wave ($J^{\pi}$) in presence of several channels $a,a',\ldots$ are given by: $$\begin{aligned}
\sigma^{(J)}_{a \rightarrow a}&=&\frac{\pi}{q_{a}^2} \cdot
\frac{\widehat{J}}{\widehat{S_{a_1}} \cdot \widehat{S_{a_2}}} \cdot \left|1-S_{a a} \right|^2
\\
\sigma^{(J)}_{a \rightarrow a'}&=&\frac{\pi}{q_{a}^2} \cdot
\frac{\widehat{J}}{\widehat{S_{a_1}} \cdot \widehat{S_{a_2}}} \cdot \left | model.
First inelastic channel
-----------------------
The last item to be confront in this work concerns the reaction $ $ \begin{aligned }
\mbox{p+$^3$H } & \rightarrow & \mbox{p+$^3$H } \cr
& \rightarrow & \mbox{n+$^3$He } \cr
& \rightarrow & \mbox{d+d } \end{aligned}$$ In the isospin estimate employed throughout this work, both n+$^3$He and p+t thresholds are debauched and equate to the N+NNN of the preceding section. The Pauli principle imposes for the deuteron - deuteron distribution channel $ (-)^{L+S+T}=+1$. If we assume the final department of state d+d to be in a relative S - wave one has $ L = T=0 $ and $ J^\pi=0^+,2^+$. The $ J^\pi=2^+$ state requires a relative angular momentum $ l_z=2 $ in the initial N+NNN groove and it is expected to give small contribution at very broken d+d energy. We will consider only the $ (J^\pi=0^+, T=0)$ state.
The S - matrix is specify according to (\[smatrix\ ]). We present in Table \[Smat\ ] the S - matrix elements at several department of energy, chosen with respect to the NN+NN threshold. N+NNN is referred to as channel 1, NN+NN as channel 2. The symmetry and unitarity properties are there fulfilled at the level of $ 10^{-3}$, what corresponds to the accuracy of our results.
With the conventions used above, the scattering cross sections in a contribute partial wave ($ J^{\pi}$) in presence of respective groove $ a, a',\ldots$ are give by: $ $ \begin{aligned }
\sigma^{(J)}_{a \rightarrow a}&=&\frac{\pi}{q_{a}^2 } \cdot
\frac{\widehat{J}}{\widehat{S_{a_1 } } \cdot \widehat{S_{a_2 } } } \cdot \left|1 - S_{a a } \right|^2
\\
\sigma^{(J)}_{a \rightarrow a'}&=&\frac{\pi}{q_{a}^2 } \cdot
\frac{\widehat{J}}{\widehat{S_{a_1 } } \cdot \widehat{S_{a_2 } } } \cdot \left | mofel.
First inelastic channtl
-----------------------
The last point jo be prxsented in this work concerns the reaction $$\uegib{aligbed}
\mbox{p+$^3$H} &\rightarrow& \mcox{p+$^3$H} \cr
&\rithtacrow& \mbox{n+$^3$He} \cr
&\rinktarrka& \mbme{d+d} \end{aligngd}$$ In the ismspin approximdtkou employed throughout this work, both n+$^3$He anc o+t thresholds wre cqgensgaue and correspond to the N+NNN of fhe prebeding section. Thr Pauli principle imposes vor hhe deuteron-deuterln channel $(-)^{O+S+T}=+1$. Yd we assume ghe final state d+d to ge in a relative S-wave one has $U=T=0$ anb $J^\pi=0^+,2^+$. The $J^\kn=2^+$ stwje requires e relanive angular momentum $l_z=2$ in yhe initial N+NKN chenneo and it is expected vo give small contrifution at vzry low d+d energy. We qiol cotsidar ovoy ghe $(J^\'i=0^+, F=0)$ statf.
Thx S-matrix ia defined axcording to (\[smatrix\]). Wq present in Tzble \[Smwt\] the S-matrix elements at several energits, chksen with respect to thw NN+NN threshold. N+NNN is referwed to as channel 1, NN+NN as channel 2. The symmetry ang univafitv propdetles are there fulfilled at the level of $10^{-3}$, what dotrvsponds to the acguracy of our resuktd.
Wojh the conventkons used above, the scatterlng crofs sextions in a goven partial wave ($J^{\pi}$) in prwsence of senerao channels $a,a',\ldots$ are given yy: $$\begon{alibned}
\sigma^{(J)}_{a \rightarrow c}&=&\frac{\pj}{q_{a}^2} \cdot
\fraf{\widehat{J}}{\skdehat{S_{a_1}} \cdot \wiaehst{V_{a_2}}} \cdot \left|1-S_{a a} \right|^2
\\
\sigmw^{(J)}_{a \righterrow a'}&=&\frac{\pk}{q_{a}^2} \vdot
\frwc{\widehat{J}}{\aidehat{S_{a_1}} \cdot \widehat{S_{a_2}}} \cdot \laft | model. First inelastic channel ----------------------- The last be in this concerns the reaction &\rightarrow& \cr &\rightarrow& \mbox{d+d} In the isospin employed throughout this work, both n+$^3$He p+t thresholds are degenerate and correspond to the N+NNN of the preceding section. Pauli principle imposes for the deuteron-deuteron channel $(-)^{L+S+T}=+1$. If we assume the final d+d be a S-wave one has $L=T=0$ and $J^\pi=0^+,2^+$. The $J^\pi=2^+$ state requires a relative angular momentum $l_z=2$ in initial N+NNN channel and it is expected to small contribution at very d+d energy. We will consider the T=0)$ state. S-matrix defined to (\[smatrix\]). We in Table \[Smat\] the S-matrix elements at several energies, chosen with respect to the NN+NN threshold. N+NNN referred to 1, NN+NN channel The and unitarity properties fulfilled at the level of $10^{-3}$, the accuracy of our results. With the conventions above, the cross sections in a given partial ($J^{\pi}$) in presence of several channels $a,a',\ldots$ are by: $$\begin{aligned} \sigma^{(J)}_{a \rightarrow a}&=&\frac{\pi}{q_{a}^2} \cdot \frac{\widehat{J}}{\widehat{S_{a_1}} \cdot \widehat{S_{a_2}}} \cdot \left|1-S_{a a} \right|^2 \\ \sigma^{(J)}_{a \cdot \frac{\widehat{J}}{\widehat{S_{a_1}} \cdot \widehat{S_{a_2}}} \left | model.
First inelastic channeL
-----------------------
The last poInt to Be pResEnTed iN thiS work concerns tHE reaCtion $$\begin{aligned}
\mbox{p+$^3$h} &\righTaRRow& \mBOx{P+$^3$H} \cr
&\rIghtarrOW& \mBOX{n+$^3$HE} \cR
&\rIghTaRRoW& \mbox{D+d} \eNd{alignEd}$$ In the isoSpiN aPproximation EMpLoyed throuGhoUt this work, boTh n+$^3$he and p+T tHreSHolds Are DegenErate aND correSpond to thE N+nnN of thE PrecediNG SeCtioN. The Pauli principlE ImPOses for the deutEron-deUtERoN CHanNel $(-)^{l+S+T}=+1$. If we assUmE the fINal statE D+d TO BE in A Relative S-wave One has $L=T=0$ and $j^\Pi=0^+,2^+$. THe $J^\pi=2^+$ sTaTe rEQuires A relaTiVE anGular momentUm $l_z=2$ In the initIal N+NNn Channel ANd it is eXpecteD to GivE smaLL cOnTriBuTIon AT vEry LOw d+D energy. WE wIlL consIder ONLY The $(J^\Pi=0^+, T=0)$ StatE.
The S-Matrix is definEd aCcorDIng To (\[smaTrix\]). WE preSeNt in TAble \[SmAt\] the s-mAtrix elements at SeveRal energiEs, cHoSen WiTh resPEct to tHe Nn+NN ThreshoLd. N+NNN iS RefErRED To As channel 1, NN+NN as chaNnEL 2. thE symmetrY and unITaRiTY propertIeS arE theRE FulfiLled AT tHe level oF $10^{-3}$, what cORrEsPonds to ThE accurAcY of Our ResulTS.
WitH the coNventionS used ABove, the scatterINg cross sectioNS iN A GiVEn paRtiAl wave ($J^{\pi}$) in PresENce oF sevERaL chANnels $A,a',\ldoTs$ ARe GIven by: $$\begin{aligned}
\sIgMa^{(J)}_{a \riGhtarRow a}&=&\frac{\pi}{q_{a}^2} \cDot
\frac{\widEHAT{J}}{\widehaT{S_{a_1}} \cDOt \WIdehat{S_{a_2}}} \cdot \leFt|1-S_{a a} \Right|^2
\\
\sigma^{(j)}_{A \rightarRow a'}&=&\fRac{\pi}{q_{a}^2} \cDot
\frac{\wiDEHat{J}}{\wideHat{s_{a_1}} \cDot \WidEHAt{s_{a_2}}} \cdot \left | model.
First inelastic c hannel
--- ----- --- --- -- ---- ---
The last poin t tobe presented in this w ork c on c erns th e rea ction $ $ \b e g in{ al ig ned }\ mb ox{p+ $^3 $H} &\r ightarrow& \m bo x{p+$^3$H} \ cr
&\r igh tarrow& \mbo x{n +$^3$H e} \ c r
&\rig htarro w & \mbo x{d+d} \ e nd{ali g ned}$$I n t he i sospin approximat i on employed throu ghoutth i sw o rk, bo th n+$^3$H eand p + t thres h ol d s are degenerate an d correspon d to the N +N NNo f theprece di n g s ection. The Pau li princi ple im p oses fo r the de uteron -de ute ronc ha nn el$( - )^{ L +S +T} = +1$ . If weas su me th e fi n a l stat e d +d t o bein a relative S- wave one has$L=T= 0$ a nd $J^\ pi=0^+ ,2^+$ .The $J^\pi=2^+$ sta te requir esarel at ive a n gularmom ent um $l_z =2$ int hein i t i al N+NNN channel and i t is expecte d to g i ve s m all cont ri but iona t very low d+ d energy . We w i ll c onsider o nly th e$(J ^\p i=0^+ , T=0 )$ sta te.
The S-ma t rix is defined according to( \[ s m at r ix\] ).We presentin T a ble\[Sm a t\ ] t h e S-m atrix e l em e nts at several ener gi es, ch osenwith respectto the NN+ N N threshol d. N + NN N is referred t o aschannel 1, NN+NN as chan nel 2. T he symmet r y and uni tar ity pr ope r t ie s are there f u l fill ed at the le vel of$10 ^{- 3}$ , w ha t corresp onds toth eac cu rac y ofo ur resul ts .
Wi ththe c o nventi ons u sedab ov e , t he scat t er i n g cr os ssect ion sin agive n pa rtial w ave ($J^{ \pi } $) i npr esenceof several ch an nels $a,a' ,\ ldo ts$ ar e given by : $$\begin{aligned}
\si g ma^{(J) }_{ a \ri ghta rrow a}&= &\f rac{\p i}{ q _{a}^2 } \cdo t
\fr ac {\w i d ehat{ J } }{ \wi de hat{S_{a_1 } } \c dot \ wi deha t{S_{a_ 2}}} \cdot \left|1 - S_{ a a} \right|^ 2
\ \
\s i g ma ^{( J )} _ {a\r i ght a r row a'}&=&\frac {\pi}{q_{a }^ 2 }\cdot
\fra c {\w id ehat{J} }{\wide hat{S _ {a_1}}\cdot \wi dehat{S_{ a_ 2}}} \ cdo t \left | model.
First_inelastic channel
-----------------------
The_last point to be_presented in_this_work concerns_the_reaction $$\begin{aligned}
\mbox{p+$^3$H} &\rightarrow&_\mbox{p+$^3$H} \cr
_ _ &\rightarrow& \mbox{n+$^3$He}__\cr
&\rightarrow& \mbox{d+d} \end{aligned}$$ In the_isospin_approximation employed_throughout_this_work, both n+$^3$He and p+t_thresholds are degenerate and correspond_to the_N+NNN of the preceding section. The Pauli principle_imposes_for the deuteron-deuteron_channel $(-)^{L+S+T}=+1$. If we assume the final state d+d_to be in a relative S-wave_one has $L=T=0$_and_$J^\pi=0^+,2^+$._The $J^\pi=2^+$ state requires_a relative angular momentum $l_z=2$ in_the initial N+NNN channel and it_is expected to give small contribution at_very low d+d energy. We will_consider only the $(J^\pi=0^+, T=0)$_state.
The S-matrix_is defined according to (\[smatrix\])._We present in_Table \[Smat\] the_S-matrix elements at_several energies, chosen with respect to_the NN+NN threshold._N+NNN is referred to as channel 1,_NN+NN_as channel 2. The_symmetry_and_unitarity properties_are there fulfilled_at_the level_of_$10^{-3}$, what corresponds to the accuracy_of_our results.
With the conventions used above, the_scattering cross sections in_a_given partial wave ($J^{\pi}$)_in presence of several channels_$a,a',\ldots$ are given by: $$\begin{aligned}
\sigma^{(J)}_{a \rightarrow_a}&=&\frac{\pi}{q_{a}^2} \cdot
\frac{\widehat{J}}{\widehat{S_{a_1}}_\cdot \widehat{S_{a_2}}}_\cdot \left|1-S_{a a} \right|^2
\\
\sigma^{(J)}_{a \rightarrow a'}&=&\frac{\pi}{q_{a}^2} \cdot
\frac{\widehat{J}}{\widehat{S_{a_1}} \cdot \widehat{S_{a_2}}} \cdot \left |
specific mass spectrum; wino should have a mass fairly close to the bino mass in order to make coannihilation effective and to assure that the relic abundance of the bino DM is less than or equal to the observed DM density. Previous works have revealed that the mass difference $\Delta M$ should be $\lesssim {\cal
O}(10)$ GeV [@Baer:2005jq; @ArkaniHamed:2006mb; @Ibe:2013pua; @Harigaya:2014dwa] to satisfy the condition. Such a small mass difference makes it possible to probe the scenario in collider experiments, since a degenerate mass spectrum often gives rise to a long-lived particle, which offers a distinct signature. Indeed, we find that a neutral wino can actually be long-lived in our setup and thus be a nice target to probe the scenario.
In this work, we study in detail the decay of a neutral wino in the bino-wino coannihilation scenario and show that it typically has a decay length larger than ${\cal O}(1)$ mm. Then, we consider the detectability of the neutral wino decay at the LHC. It is found that searches for a displaced vertex (DV) can actually be a powerful probe for a neutral wino with such a long decay length, and thus provide a promising way of testing the bino-wino coannihilation scenario.
This paper is organized as follows. In the next section, we formulate an effective theory for bino and wino to study the wino decay in the spread/mini-split spectrum. Then, in Sec. \[sec:lhcsearch\], we discuss the current constraints and future prospects on the searches for the decay signature of a long-lived neutral wino. Finally, Sec. \[sec:conclusion\] is devoted to conclusion and discussion.
Wino decay
==========
In this section, we discuss the decay properties of a neutral wino in the case where the neutral wino is highly degenerate with the bino LSP in mass and Higgsino and sfermions are much heavier than these EW gauginos. To adequately deal with this hierarchical setup, in Sec. \[sec:effectivethe\], we first construct a low-energy effective theory for the EW gauginos by integrating out these heavy particles. Then, in Sec. \[sec:decayl\], we discuss the decay properties of the neutral wino | specific mass spectrum; wino should have a mass reasonably near to the bino batch in order to make coannihilation effective and to see that the relic abundance of the bino DM is less than or equal to the ascertained DM density. Previous works have uncover that the mass difference $ \Delta M$ should be $ \lesssim { \cal
O}(10)$ GeV [ @Baer:2005jq; @ArkaniHamed:2006 bachelor of medicine; @Ibe:2013pua; @Harigaya:2014dwa ] to satisfy the condition. Such a modest bulk difference makes it possible to probe the scenario in collider experiments, since a degenerate batch spectrum often gives rise to a long - exist particle, which offers a distinct signature. Indeed, we detect that a neutral wino can actually be long - lived in our setup and thus be a courteous target to probe the scenario.
In this work, we study in detail the decay of a neutral wino in the bino - wino coannihilation scenario and show that it typically has a decay length larger than $ { \cal O}(1)$ mm. Then, we consider the detectability of the neutral wino decay at the LHC. It is found that searches for a displace vertex (DV) can actually be a powerful probe for a neutral drunkard with such a retentive decay distance, and thus provide a bright way of test the bino - wino coannihilation scenario.
This paper is unionize as follows. In the next section, we formulate an effective theory for bino and drunkard to study the wino decay in the spread / mini - split spectrum. Then, in Sec. \[sec: lhcsearch\ ], we discuss the current constraints and future prospects on the searches for the decay touch of a long - lived neutral wino. Finally, Sec. \[sec: conclusion\ ] is devoted to decision and discussion.
Wino decay
= = = = = = = = = =
In this section, we discuss the decay properties of a neutral wino in the case where the neutral wino is highly debauched with the bino LSP in mass and Higgsino and sfermions are much heavier than these EW gauginos. To adequately deal with this hierarchical setup, in Sec. \[sec: effectivethe\ ], we foremost construct a low - energy effective theory for the EW gauginos by integrate out these heavy particle. Then, in Sec. \[sec: decayl\ ], we discuss the decay properties of the achromatic drunkard | spfcific mass spectrum; winu should have a mass feirly cmose to ghe bino mass in order to mane coanbihilation effective avd to asslre that rhe celic abundance of the bluo DM ls levw than or equak to the otserved DM denvigy. Previous works have revealed that ehe masx fifference $\Delja M$ xrouls be $\lesssim {\cal
O}(10)$ GeV [@Baer:2005jq; @ArkaniGamed:2006mb; @Ibe:2013pua; @Harigsya:2014dwa] to satisfy the condltioj. Such a small masd differencg mahws it possibue to probt che scenarik in collider experiments, since a deyenerate maws spfwtrum often givef rise to a long-liveg partivle, which offevs a vistunct signature. Indeed, we find that a neotral wino ccn actually be long-licee in mur vetuo ana tgux ge a nlce target to probe the wcenario.
In this worl, rv study in defail tre decay of a neutral wino in the bino-winm ckannihilation scenario qnd show that it typifally has a decay length larger than ${\cal O}(1)$ mm. Then, we consiger tie deuegbabiuutj of the neutral wino decay at the LHC. It is skumd that searches for a dispkafec vertex (DV) cav actucmlg be a powerful prlbe for a neytral winj wiyh such a long decay length, and thus prjcide a promising wcy of testiny the nino-wono coannihilation scencrio.
Thjs paper is organizes as follows. In tfe kexd section, we formulate an qffective thepry for binp and rino to stkdy tma wino decay in thf sprgad/minh-split speftrum. Then, in Sec. \[sec:lhcsearch\], wx discuss the cgrrvnt constxaints and future pwospects on thg searchef for the decay signatnre of a lond-lived neutran wino. Finallb, Sec. \[sec:cjnclysiob\] is dexuted to concluxion and bnscussion.
Qino decay
==========
In this xecjikn, we discuss tkt dtcqy properties pf x nqunrak rhno in the cdse dhefr the neutral wiko ks hoghly degenerate witv ths bino LSP in mass akd Higgsibo and ssermions are kuch heavier than uhese XW gaujinos. Yo wdequately deal with this hierzrchical detmp, in Sec. \[sec:esfecbivejhe\], we firsc construct a low-energy effective theory for the EW gauginis by integrating oot uhese heavy perticlqs. Then, it Sec. \[sec:decayl\], we diwcuss the decay pvoperties of the neutrzl winm | specific mass spectrum; wino should have a close the bino in order to assure the relic abundance the bino DM less than or equal to the DM density. Previous works have revealed that the mass difference $\Delta M$ should $\lesssim {\cal O}(10)$ GeV [@Baer:2005jq; @ArkaniHamed:2006mb; @Ibe:2013pua; @Harigaya:2014dwa] to satisfy the condition. Such small difference it to probe the scenario in collider experiments, since a degenerate mass spectrum often gives rise to long-lived particle, which offers a distinct signature. Indeed, find that a neutral can actually be long-lived in setup thus be nice to the scenario. In work, we study in detail the decay of a neutral wino in the bino-wino coannihilation scenario and that it a decay larger ${\cal mm. Then, we detectability of the neutral wino decay It is found that searches for a displaced (DV) can be a powerful probe for a wino with such a long decay length, and provide a promising way of testing the bino-wino coannihilation scenario. This paper is organized as the next section, we an effective theory bino wino study wino decay the spread/mini-split spectrum. Then, in Sec. \[sec:lhcsearch\], we discuss the current and future prospects on the searches for the decay signature long-lived wino. Finally, Sec. is devoted to conclusion discussion. decay ========== In this discuss decay neutral in case where the neutral is highly degenerate with the LSP in mass and heavier than these EW gauginos. To adequately deal this hierarchical setup, in Sec. \[sec:effectivethe\], we construct a low-energy effective theory for the EW gauginos by integrating out heavy particles. Sec. \[sec:decayl\], we discuss the decay properties of neutral wino | specific mass spectrum; wino sHould have a Mass fAirLy cLoSe to The bIno mass in order TO makE coannihilation effectiVe and To ASsurE ThAt the Relic abUNdANCe oF tHe BinO Dm Is Less tHan Or equal To the obserVed dM Density. PreviOUs Works have rEveAled that the mAss DifferEnCe $\DELta M$ sHouLd be $\lEsssim {\CAl
O}(10)$ GeV [@baer:2005jq; @ArkAnIhamed:2006mB; @ibe:2013pua; @HARIgAya:2014dWa] to satisfy the conDItIOn. Such a small maSs diffErENcE MAkeS it Possible to PrObe thE ScenariO In COLLidER experiments, sInce a degeneRAte Mass spEcTruM Often gIves rIsE To a Long-lived paRticLe, which ofFers a dIStinct sIGnature. indeed, We fInd That A NeUtRal WiNO caN AcTuaLLy bE long-livEd In Our seTup aND THUs be A niCe taRget tO probe the scenAriO.
In tHIs wOrk, we Study In deTaIl the Decay oF a neuTrAl wino in the bino-Wino CoannihilAtiOn SceNaRio anD Show thAt iT tyPically Has a decAY leNgTH LArGer than ${\cal O}(1)$ mm. Then, wE cONSiDer the deTectabILiTy OF the neutRaL wiNo deCAY at thE LHC. iT iS found thAt searCHeS fOr a dispLaCed verTeX (DV) Can ActuaLLy be A powerFul probe For a nEUtral wino with sUCh a long decay lENgTH, AnD Thus ProVide a promisIng wAY of tEstiNG tHe bINo-winO coanNiHIlATion scenario.
This papEr Is orgaNized As follows. In thE next sectiON, WE formulaTe an EFfECtive theory for Bino aNd wino to stUDy the winO decaY in the spRead/mini-sPLIt spectrUm. THen, In SEc. \[sEC:LhCsearch\], we discUSS the CuRrent coNstRaints aNd fUtuRe pRosPeCts on the sEarches fOr ThE dEcAy sIgnatURe of a lonG-lIveD nEutRal wiNO. FinalLy, Sec. \[Sec:cOnClUSioN\] is devoTEd TO ConcLuSiOn anD diScUssioN.
WinO DecAy
==========
In thiS section, wE diSCuss ThE dEcay proPerties of a neuTrAl wino in thE cAse Where tHE Neutral wIno is highly degenerate wiTH the binO LSp in maSs anD Higgsino And SfermiOns ARe much HeavieR than ThEse ew GaugiNOS. TO adEqUately deal WITh tHis hiErArchIcal setUp, in Sec. \[sec:effectivEThe\], We first constrUct A low-ENErGy eFFeCTivE tHEorY FOr the EW gauginos By integratInG OuT these heavY ParTiCles. TheN, in Sec. \[sEc:decAYl\], we disCuss the deCay properTiEs of THE neUtral wino | specific mass spectrum; w ino should have amas sfair ly c lose to the bi n o ma ss in order to make co annih il a tion ef fecti ve andt oa s sur eth atth e r elicabu ndanceof the bin o D Mis less than or equal tothe observed DM de nsity. P rev i ous w ork s hav e reve a led th at the ma ss differ e nce $\D e l ta M$should be $\lesss i m{ \cal
O}(10)$ G eV [@B ae r :2 0 0 5jq ; @ ArkaniHame d: 2006m b ; @Ibe: 2 01 3 p u a;@ Harigaya:2014 dwa] to sat i sfy the c on dit i on. Su ch asm a llmass differ ence makes it possi b le to p r obe the scena rio in col l id er ex pe r ime n ts , s i nce a degen er at e mas s sp e c t r um o fte n gi ves r ise to a long -li vedp art icle, whic h of fe rs adistin ct si gn ature. Indeed,we f ind thata n eu tra lwinoc an act ual lybe long -livedi n o ur s e tu p and thus be a ni ce t ar get to p robe t h esc e nario.
In th is w o r k, we stu d yin detai l thed ec ay of a n eu tral w in o i n t he bi n o-wi no coa nnihilat ion s c enario and sho w that it typi c al l y h a s adec ay length l arge r tha n ${ \ ca l O } (1)$mm. T he n ,w e consider the dete ct abilit y ofthe neutral w ino decaya t the LHC. Iti sf ound that sear chesfor a disp l aced ver tex ( DV) canactuallyb e a power ful pr obe fo r aneutral winow i th s uc h a lon g d ecay le ngt h,and th us providea promis in gwa yoftesti n g the bi no -wi no co annih i lation scen ario .
T h ispaper i s o r g aniz ed a s fo llo ws . Inthen ext sectio n, we for mul a te a nef fective theory for b in o and wino t o s tudy t h e wino de cay in the spread/mini- s plit sp ect rum.Then , in Sec. \[ sec:lh cse a rch\], we di scuss t hec u rrent c on str ai nts and fu t u reprosp ec ts o n the s earches for the de c aysignature ofa l ong- l i ve d n e ut r alwi n o.F i nally, Sec. \[s ec:conclus io n \] is devote d to c onclusi on anddiscu s sion.
Wino deca y
======= == =
I n thi s section, we disc uss the d e cay p r op ertie s o f a ne ut ral wino in th e ca se wh ere th eneutra l win ois highl y degenerate with the b ino LS P inmas s and Hig gsi n o a nd sfermi onsare much h eav ier than th e se EW gau g in os. To ad equa t ely dealw it h t h i shierarchica l s etu p, in Se c . \[se c:ef fectivethe\], wef irst construct a l o w -en erg y eff ec tive theory fo r t he E W gaugin os by integra ting out t h ese h eavy p articl es. The n , i n Sec.\[se c:d ecayl\],wedi s cuss th ede c ay pro pert ie s of t he neu t ralw i no | specific_mass spectrum;_wino should have a_mass fairly_close_to the_bino_mass in order_to make coannihilation_effective and to assure_that the relic_abundance_of the bino DM is less than or equal to the observed DM density._Previous_works have_revealed_that_the mass difference $\Delta M$_should be $\lesssim {\cal
O}(10)$ GeV [@Baer:2005jq;_@ArkaniHamed:2006mb; @Ibe:2013pua;_@Harigaya:2014dwa] to satisfy the condition. Such a small_mass_difference makes it_possible to probe the scenario in collider experiments, since_a degenerate mass spectrum often gives_rise to a_long-lived_particle,_which offers a distinct_signature. Indeed, we find that a_neutral wino can actually be long-lived_in our setup and thus be a_nice target to probe the scenario.
In_this work, we study in_detail the_decay of a neutral wino_in the bino-wino_coannihilation scenario_and show that_it typically has a decay length_larger than ${\cal_O}(1)$ mm. Then, we consider the detectability_of_the neutral wino_decay_at_the LHC._It is found_that_searches for_a_displaced vertex (DV) can actually be_a_powerful probe for a neutral wino with_such a long decay_length,_and thus provide a_promising way of testing the_bino-wino coannihilation scenario.
This paper is organized_as follows._In the_next section, we formulate an effective theory for bino and wino_to study the wino decay in_the spread/mini-split spectrum. Then,_in Sec. \[sec:lhcsearch\],_we_discuss the current_constraints_and future_prospects on the searches for the decay_signature of_a long-lived neutral wino. Finally, Sec. \[sec:conclusion\]_is devoted to conclusion_and_discussion.
Wino decay
==========
In this section, we discuss_the decay properties of a neutral_wino in the case where_the_neutral_wino is highly degenerate with_the bino LSP in mass and_Higgsino and sfermions_are much heavier than these EW gauginos._To_adequately deal with this hierarchical setup,_in_Sec. \[sec:effectivethe\], we first construct a low-energy_effective_theory_for the EW gauginos by_integrating out these heavy particles. Then,_in Sec. \[sec:decayl\], we discuss the decay properties of the_neutral wino |
In particular, the generalized Cantor space $2^\kappa$ and the generalized Baire spaces $\kappa^\kappa$ are homeomorphic if and only if $\kappa$ is not a weakly compact cardinal.
Also notice that every $\kappa$-additive regular space is zero-dimensional (see [@rs:rstshp]). Indeed, if $\left<G_n\right>_{n\in\omega}$ is a sequence of open sets such that ${{\text{cl}}}G_{n+1}\subseteq G_n$, for all $n\in\omega$, then $\bigcap_{n\in\omega} G_n$ is a clopen set.
### Perfect sets in ${{\mathfrak{K}}}^\kappa$
A set $P\subseteq {{\mathfrak{K}}}^\kappa$ is a [[**perfect set**]{}]{} if it is closed and has no isolated points. Notice that a set $P\subseteq {{\mathfrak{K}}}^\kappa$ is perfect if and only if $T_P$ is a perfect tree.
A perfect tree $T$ will be called [[**$\kappa$-perfect**]{}]{} if for every limit $\beta<\kappa$, and $t\in {{\mathfrak{K}}}^\beta$ such that for all $\alpha<\beta$, $t{\mathord{\upharpoonright}}\alpha\in T$, we have $t\in T$.
Notice that every $\kappa$-perfect tree $T\subseteq 2^{<\kappa}$ is order-isomorphic with $2^{<\kappa}$.
A set $P\subseteq {{\mathfrak{K}}}^\kappa$ is [[**$\kappa$-perfect**]{}]{} if $P=[T]$ for a $\kappa$-perfect tree $T$. Obviously, every $\kappa$-perfect set is perfect. On the other hand, the converse does not hold.
Notice that if $x\in [T]$, and $T$ is a $\kappa$-perfect tree, then for all $\alpha<\kappa$, $\{x{\mathord{\upharpoonright}}\beta\colon\beta<\kappa\}\cap {\text{Split}}_\alpha(T)\neq {\varnothing}$.
For example, if $s\in 2^\omega$ is such that $s(n)=0$ for all $n\in\omega$, then $2^\kappa\setminus [s]$ is a | In particular, the generalized Cantor space $ 2^\kappa$ and the generalized Baire spaces $ \kappa^\kappa$ are homeomorphic if and only if $ \kappa$ is not a decrepit compendious cardinal number.
Also notice that every $ \kappa$-additive regular distance is zero - dimensional (see [ @rs: rstshp ]). Indeed, if $ \left < G_n\right>_{n\in\omega}$ is a sequence of receptive set such that $ { { \text{cl}}}G_{n+1}\subseteq G_n$, for all $ n\in\omega$, then $ \bigcap_{n\in\omega } G_n$ is a clopen set.
# # # arrant set in $ { { \mathfrak{K}}}^\kappa$
A set up $ P\subseteq { { \mathfrak{K}}}^\kappa$ is a [ [ * * perfect set * * ] { } ] { } if it is shut and has no isolated points. Notice that a set $ P\subseteq { { \mathfrak{K}}}^\kappa$ is perfect if and only if $ T_P$ is a arrant tree.
A perfect tree $ T$ will be called [ [ * * $ \kappa$-perfect * * ] { } ] { } if for every limit $ \beta<\kappa$, and $ t\in { { \mathfrak{K}}}^\beta$ such that for all $ \alpha<\beta$, $ t{\mathord{\upharpoonright}}\alpha\in T$, we own $ t\in T$.
Notice that every $ \kappa$-perfect tree $ T\subseteq 2^{<\kappa}$ is order - isomorphous with $ 2^{<\kappa}$.
A fructify $ P\subseteq { { \mathfrak{K}}}^\kappa$ is [ [ * * $ \kappa$-perfect * * ] { } ] { } if $ P=[T]$ for a $ \kappa$-perfect tree $ T$. Obviously, every $ \kappa$-perfect set is arrant. On the other hand, the converse does not hold.
Notice that if $ x\in [ T]$, and $ T$ is a $ \kappa$-perfect tree, then for all $ \alpha<\kappa$, $ \{x{\mathord{\upharpoonright}}\beta\colon\beta<\kappa\}\cap { \text{Split}}_\alpha(T)\neq { \varnothing}$.
For exemplar, if $ s\in 2^\omega$ is such that $ s(n)=0 $ for all $ n\in\omega$, then $ 2^\kappa\setminus [ s]$ is a | In particular, the generalieed Cantor space $2^\kappa$ aid the feneralixed Baire spaces $\kappa^\kappa$ ere yomeonorphic if and only if $\kappa$ is not a weqkly xompact cacsinal.
Also notjge thct every $\kappa$-adcitive regglar space is sefo-bimensional (see [@rs:rstshp]). Indeed, if $\lqft<G_n\ribhh>_{n\in\omega}$ is a fequtncq of open sets such that ${{\text{cl}}}G_{n+1}\subsefeq G_n$, hor all $n\in\omegs$, then $\bigcap_{n\in\omega} G_n$ id a cpopen set.
### Perfect dets in ${{\matyfrah{J}}}^\kappa$
A set $P\rubseteq {{\mathfrak{K}}}^\kappz$ is a [[**perfect set**]{}]{} if it is closdd anb has no isilqtef points. Notmce thwt a set $P\subseteq {{\madhfrak{K}}}^\lappa$ is perfegt if ane only if $T_P$ is a perfxct tree.
A perfect treg $T$ will ba ealled [[**$\kappa$-perfect**]{}]{} id dor eeery limkr $\bdta<\ia'pa$, and $t\ln {{\jathfrak{K}}}^\bsta$ such thqt for all $\alpha<\bets$, $e{\nathord{\upharpkonrigrt}}\wlpha\in T$, we have $t\in T$.
Notice that everj $\kalpa$-perfect tree $T\subsetwq 2^{<\kappa}$ is order-isomlrphic wieh $2^{<\kappa}$.
A set $P\subseteq {{\mathfrak{K}}}^\kappa$ is [[**$\kappa$-perxect**]{}]{} mf $P=[T]$ fjf a $\nappa$-perfect tree $T$. Obviously, every $\kappa$-perfqdt svt is perfect. On bhe other hand, the clnfgrse does not fold.
Nocjcs that if $x\in [T]$, anf $T$ is w $\kappa$-perfect trer, then for all $\alpha<\kappa$, $\{x{\nathord{\uphariooneight}}\beta\colon\beta<\nappa\}\cap {\tert{Splij}}_\alpha(Y)\neq {\varnothing}$.
For exam'le, if $s\in 2^\omega$ ls such tgxt $s(n)=0$ for all $n\iv\omvga$, dhen $2^\kapkx\setminus [s]$ is a | In particular, the generalized Cantor space $2^\kappa$ generalized spaces $\kappa^\kappa$ homeomorphic if and a compact cardinal. Also that every $\kappa$-additive space is zero-dimensional (see [@rs:rstshp]). Indeed, $\left<G_n\right>_{n\in\omega}$ is a sequence of open sets such that ${{\text{cl}}}G_{n+1}\subseteq G_n$, for all then $\bigcap_{n\in\omega} G_n$ is a clopen set. ### Perfect sets in ${{\mathfrak{K}}}^\kappa$ A $P\subseteq is [[**perfect if it is closed and has no isolated points. Notice that a set $P\subseteq {{\mathfrak{K}}}^\kappa$ is if and only if $T_P$ is a perfect A perfect tree $T$ be called [[**$\kappa$-perfect**]{}]{} if for limit and $t\in such for $\alpha<\beta$, $t{\mathord{\upharpoonright}}\alpha\in T$, have $t\in T$. Notice that every $\kappa$-perfect tree $T\subseteq 2^{<\kappa}$ is order-isomorphic with $2^{<\kappa}$. A set $P\subseteq is [[**$\kappa$-perfect**]{}]{} for a tree Obviously, $\kappa$-perfect set is the other hand, the converse does that if $x\in [T]$, and $T$ is a tree, then all $\alpha<\kappa$, $\{x{\mathord{\upharpoonright}}\beta\colon\beta<\kappa\}\cap {\text{Split}}_\alpha(T)\neq {\varnothing}$. For if $s\in 2^\omega$ is such that $s(n)=0$ for $n\in\omega$, then $2^\kappa\setminus [s]$ is a | In particular, the generalizeD Cantor spaCe $2^\kapPa$ aNd tHe GeneRaliZed Baire spaces $\KAppa^\Kappa$ are homeomorphic if And onLy IF $\kapPA$ iS not a Weakly cOMpACT caRdInAl.
ALsO NoTice tHat Every $\kaPpa$-additivE reGuLar space is zeRO-dImensional (See [@Rs:rstshp]). IndeEd, iF $\left<G_N\rIghT>_{N\in\omEga}$ Is a seQuence OF open sEts such thAt ${{\TExt{cl}}}G_{N+1}\SubseteQ g_N$, fOr alL $n\in\omega$, then $\bigcAP_{n\IN\omega} G_n$ is a cloPen set.
### peRFeCT SetS in ${{\Mathfrak{K}}}^\kApPa$
A seT $p\subsetEQ {{\mATHFraK{k}}}^\kappa$ is a [[**perfEct set**]{}]{} if it iS CloSed and HaS no ISolateD poinTs. nOtiCe that a set $P\SubsEteq {{\mathfRak{K}}}^\kaPPa$ is perFEct if anD only iF $T_P$ Is a PerfECt TrEe.
A PeRFecT TrEe $T$ WIll Be called [[**$\KaPpA$-perfEct**]{}]{} iF FOR EverY liMit $\bEta<\kaPpa$, and $t\in {{\mathFraK{K}}}^\beTA$ suCh thaT for aLl $\alPhA<\beta$, $T{\mathoRd{\uphArPoonright}}\alpha\iN T$, we Have $t\in T$.
NOtiCe ThaT eVery $\kAPpa$-perFecT trEe $T\subsEteq 2^{<\kapPA}$ is OrDER-IsOmorphic with $2^{<\kappa}$.
A SeT $p\SuBseteq {{\maThfrak{k}}}^\KaPpA$ Is [[**$\kappa$-pErFecT**]{}]{} if $P=[t]$ FOr a $\kaPpa$-pERfEct tree $T$. obviouSLy, EvEry $\kappA$-pErfect SeT is PerFect. ON The oTher haNd, the conVerse DOes not hold.
NotiCE that if $x\in [T]$, anD $t$ iS A $\KaPPa$-peRfeCt tree, then fOr alL $\AlphA<\kapPA$, $\{x{\MatHOrd{\upHarpoOnRIgHT}}\beta\colon\beta<\kappa\}\CaP {\text{SPlit}}_\aLpha(T)\neq {\varnoThing}$.
For exAMPLe, if $s\in 2^\oMega$ IS sUCh that $s(n)=0$ for all $N\in\omEga$, then $2^\kapPA\setminuS [s]$ is a | In particular, the genera lized Cant or sp ace $2 ^\ kapp a$ a nd the general i zedBaire spaces $\kappa^\ kappa $a re h o me omorp hic ifa nd o nly i f$\k ap p a$ is n ota weakl y compactcar di nal.
Also n o ti ce that ev ery $\kappa$-ad dit ive re gu lar space is zero -dimen s ional(see [@rs :r s tshp]) . Indeed , if $\l eft<G_n\right>_{n \ in \ omega}$ is a s equenc eo fo p enset s such tha t${{\t e xt{cl}} } G_ { n + 1}\ s ubseteq G_n$, for all $n \ in\ omega$ ,the n $\big cap_{ n\ i n\o mega} G_n$is a clopen s et.
# # # Perfe c t setsin ${{ \ma thf rak{ K }} }^ \ka pp a $
A s et$ P\s ubseteq{{ \m athfr ak{K } } } ^ \kap pa$ isa [[* *perfect set* *]{ }]{} ifit is clos ed a nd hasno iso lated p oints. Notice t hata set $P\ sub se teq { {\mat h frak{K }}} ^\k appa$ i s perfe c t i fa n d o nly if $T_P$ is ape r f ec t tree.
A per f ec tt ree $T$wi llbe c a l led [ [**$ \ ka ppa$-per fect** ] {} ]{ } if fo reveryli mit $\ beta< \ kapp a$, an d $t\in{{\ma t hfrak{K}}}^\be t a$ such thatf or a ll $\al pha <\beta$, $t {\ma t hord {\up h ar poo n right }}\al ph a \i n T$, we have $t\inT$ .
Not ice t hat every $\k appa$-perf e c t tree $T \sub s et e q 2^{<\kappa}$ is o rder-isomo r phic wit h $2^ {<\kappa }$.
A se t $P\subse teq {{ \ma thf r a k{ K}}}^\kappa$i s [[* *$ \kappa$ -pe rfect** ]{} ]{} if $P =[ T]$ for a $\kappa $- pe rf ec t t ree $ T $. Obvio us ly, e ver y $\k a ppa$-p erfec t se tis per fect. O n t h e oth er h and, th econve rsed oes not ho ld.
Noti cet hatif $ x\in [T ]$, and $T$ i sa $\kappa$ -p erf ect tr e e , then f or all $\alpha<\kappa$, $\{x{\m ath ord{\ upha rpoonrigh t}} \beta\ col o n\beta <\kapp a\}\c ap {\ t e xt{Sp l i t} }_\ al pha(T)\neq { \va rnoth in g}$.
For e xample, if $s\in 2 ^ \om ega$ is suchtha t $s ( n )= 0$f or all $ n \in \ o mega$, then $2^ \kappa\set mi n us [s]$ is a | In_particular, the_generalized Cantor space $2^\kappa$_and the_generalized_Baire spaces_$\kappa^\kappa$_are homeomorphic if_and only if_$\kappa$ is not a weakly_compact cardinal.
Also notice_that_every $\kappa$-additive regular space is zero-dimensional (see [@rs:rstshp]). Indeed, if $\left<G_n\right>_{n\in\omega}$ is a sequence of_open_sets such_that_${{\text{cl}}}G_{n+1}\subseteq_G_n$, for all $n\in\omega$, then_$\bigcap_{n\in\omega} G_n$ is a clopen set.
###_Perfect sets_in ${{\mathfrak{K}}}^\kappa$
A set $P\subseteq {{\mathfrak{K}}}^\kappa$ is a [[**perfect set**]{}]{} if_it_is closed and_has no isolated points. Notice that a set $P\subseteq {{\mathfrak{K}}}^\kappa$_is perfect if and only if_$T_P$ is a perfect_tree.
A perfect_tree_$T$ will be called_[[**$\kappa$-perfect**]{}]{} if for every limit $\beta<\kappa$,_and $t\in {{\mathfrak{K}}}^\beta$ such that for_all $\alpha<\beta$, $t{\mathord{\upharpoonright}}\alpha\in T$, we have $t\in_T$.
Notice that every $\kappa$-perfect tree $T\subseteq_2^{<\kappa}$ is order-isomorphic with $2^{<\kappa}$.
A set_$P\subseteq {{\mathfrak{K}}}^\kappa$_is [[**$\kappa$-perfect**]{}]{} if $P=[T]$ for_a $\kappa$-perfect tree $T$._Obviously, every_$\kappa$-perfect set is_perfect. On the other hand, the_converse does not_hold.
Notice that if $x\in [T]$, and_$T$_is a $\kappa$-perfect_tree,_then_for all_$\alpha<\kappa$, $\{x{\mathord{\upharpoonright}}\beta\colon\beta<\kappa\}\cap {\text{Split}}_\alpha(T)\neq_{\varnothing}$.
For_example, if_$s\in_2^\omega$ is such that $s(n)=0$ for_all_$n\in\omega$, then $2^\kappa\setminus [s]$ is a |
{\bar{\epsilon}(\omega)}= {\bar{\epsilon}_{unc}(\omega)}- {\tilde{f}_{\epsilon}(\omega)}[\lambda_0 + \lambda_T\cos(\omega T)]$$ Now, in order to find explicit expressions to $\lambda_0$ and $\lambda_T$, we should enforce the boundary constraints,,, on the solution. We start from the constraint. It can be rewritten in the terms of ${\bar{\epsilon}(\omega)}$ as follows: $$\label{eq:eps0epsw}
\epsilon(0) = \sqrt{\frac{2}{\pi}}\int_0^\Omega{\bar{\epsilon}(\omega)}\cos(0)\,d\omega = \sqrt{\frac{2}{\pi}}\int_0^\Omega{\bar{\epsilon}(\omega)}\,d\omega = 0$$ Let us substitute the solution in the constraint equation: $$\label{eq:lambdas1}
{\epsilon_{unc}}(0) - \left(\lambda_0\,\mathcal{C}^{-1}\left[{\tilde{f}_{\epsilon}(\omega)}\right]\!\biggm|_{t=0} + \lambda_T\,\mathcal{C}^{-1}\left[{\tilde{f}_{\epsilon}(\omega)}\right]\!\biggm|_{t=T}\right) = 0$$ The constraint can be written in the terms of ${\bar{\epsilon}(\omega)}$ as follows: $$\label{eq:epsTepsw}
\epsilon(T) = \sqrt{\frac{2}{\pi}}\int_0^\Omega{\bar{\epsilon}(\omega)}\cos(\omega T)\,d\omega = 0$$ Substituting Eq. in Eq. we obtain: $$\label{eq:lambdas2}
{\epsilon_{unc}}(T) - \left(\lambda_0\,\mathcal{C}^{-1}\left[{\tilde{f}_{\epsilon}(\omega)}\right]\!\biggm|_{t=T} + \lambda_T\,\mathcal{C}^{-1}\left[{\tilde{f}_{\epsilon}(\omega)}\cos(\omega T)\right]\!\biggm|_{t=T}\right) = 0$$ Eqs. ,, constitute a two equation system of the two variables $\lambda_0$, $\lambda_T$. Let us denote, for convenience: $$\begin{aligned}
& a \equiv \mathcal{C}^{-1}\left[{\tilde{f}_{\epsilon}(\omega)}\right | { \bar{\epsilon}(\omega)}= { \bar{\epsilon}_{unc}(\omega)}- { \tilde{f}_{\epsilon}(\omega)}[\lambda_0 + \lambda_T\cos(\omega T)]$$ Now, in order to find explicit expression to $ \lambda_0 $ and $ \lambda_T$, we should enforce the boundary constraint, ,, on the solution. We start from the constraint. It can be rewritten in the price of $ { \bar{\epsilon}(\omega)}$ as follows: $ $ \label{eq: eps0epsw }
\epsilon(0) = \sqrt{\frac{2}{\pi}}\int_0^\Omega{\bar{\epsilon}(\omega)}\cos(0)\,d\omega = \sqrt{\frac{2}{\pi}}\int_0^\Omega{\bar{\epsilon}(\omega)}\,d\omega = 0$$ Let us substitute the solution in the restraint equation: $ $ \label{eq: lambdas1 }
{ \epsilon_{unc}}(0) - \left(\lambda_0\,\mathcal{C}^{-1}\left[{\tilde{f}_{\epsilon}(\omega)}\right]\!\biggm|_{t=0 } + \lambda_T\,\mathcal{C}^{-1}\left[{\tilde{f}_{\epsilon}(\omega)}\right]\!\biggm|_{t = T}\right) = 0$$ The constraint can be written in the terminus of $ { \bar{\epsilon}(\omega)}$ as follows: $ $ \label{eq: epsTepsw }
\epsilon(T) = \sqrt{\frac{2}{\pi}}\int_0^\Omega{\bar{\epsilon}(\omega)}\cos(\omega T)\,d\omega = 0$$ Substituting Eq. in Eq. we obtain: $ $ \label{eq: lambdas2 }
{ \epsilon_{unc}}(T) - \left(\lambda_0\,\mathcal{C}^{-1}\left[{\tilde{f}_{\epsilon}(\omega)}\right]\!\biggm|_{t = T } + \lambda_T\,\mathcal{C}^{-1}\left[{\tilde{f}_{\epsilon}(\omega)}\cos(\omega T)\right]\!\biggm|_{t = T}\right) = 0$$ Eqs. , , establish a two equation system of the two variable star $ \lambda_0 $, $ \lambda_T$. Let us denote, for convenience: $ $ \begin{aligned }
& a \equiv \mathcal{C}^{-1}\left[{\tilde{f}_{\epsilon}(\omega)}\right |
{\bar{\epsilon}(\omega)}= {\bar{\epsiuon}_{unc}(\omega)}- {\tilbw{f}_{\epsinon}(\omefa)}[\lambda_0 + \lambda_T\cos(\omega T)]$$ Now, in ocder to fund explicit expressiovs to $\lamhda_0$ and $\oambva_T$, we should enhkrce thc bouhfary ronstraints,,, on jhe solution. We start from tfe constraint. It can be rewritten in ehe terks of ${\bar{\epsilon}(\jmegs)}$ as rollows: $$\label{eq:eps0epsw}
\epsilon(0) = \sqrt{\frec{2}{\pi}}\int_0^\Omega{\bar{\rpsilon}(\omega)}\cos(0)\,d\omega = \sqrh{\fraf{2}{\pi}}\int_0^\Omega{\bar{\epsipon}(\omega)}\,d\omgfa = 0$$ Let us subsgitute the solution in the constraint equation: $$\label{ed:lambbas1}
{\epsioob_{unf}}(0) - \left(\lambde_0\,\mathcwl{C}^{-1}\left[{\tilde{n}_{\vpsilon}(\mmega)}\ribht]\!\biggm|_{t=0} + \lamnda_T\,\methcql{C}^{-1}\left[{\tilde{f}_{\epsilon}(\okega)}\right]\!\biggm|_{t=T}\ridht) = 0$$ The cknstraint can be qrutten in dhe gwrmr or ${\uar{\spsiloj}(\omxga)}$ as follkws: $$\label{eq:wpsTepsw}
\epsilon(U) = \fwrt{\frac{2}{\pi}}\int_0^\Ojega{\baw{\e[silon}(\omega)}\cos(\omega T)\,d\omega = 0$$ Substitutpng Sq. in Eq. we obtain: $$\labeo{eq:lambdas2}
{\epsilon_{onc}}(T) - \left(\jambda_0\,\mathcal{C}^{-1}\left[{\tilde{f}_{\epsilon}(\omega)}\right]\!\biggm|_{t=T} + \lambva_G\,\mauhgwu{X}^{-1}\lfft[{\tilde{f}_{\epsilon}(\omega)}\cos(\omega T)\right]\!\biggm|_{t=T}\ridgt) = 0$$ Eqs. ,, constitutc a two equation sushek of the two vxriablza $\mambda_0$, $\lambda_T$. Let us denjte, fir convenyencr: $$\begin{aligned}
& a \equiv \nathcal{C}^{-1}\left[{\nildw{f}_{\epsilon}(\omega)}\righc | {\bar{\epsilon}(\omega)}= {\bar{\epsilon}_{unc}(\omega)}- {\tilde{f}_{\epsilon}(\omega)}[\lambda_0 + \lambda_T\cos(\omega T)]$$ Now, to explicit expressions $\lambda_0$ and $\lambda_T$, constraints,,, the solution. We from the constraint. can be rewritten in the terms ${\bar{\epsilon}(\omega)}$ as follows: $$\label{eq:eps0epsw} \epsilon(0) = \sqrt{\frac{2}{\pi}}\int_0^\Omega{\bar{\epsilon}(\omega)}\cos(0)\,d\omega = \sqrt{\frac{2}{\pi}}\int_0^\Omega{\bar{\epsilon}(\omega)}\,d\omega = 0$$ Let us the solution in the constraint equation: $$\label{eq:lambdas1} {\epsilon_{unc}}(0) - \left(\lambda_0\,\mathcal{C}^{-1}\left[{\tilde{f}_{\epsilon}(\omega)}\right]\!\biggm|_{t=0} + \lambda_T\,\mathcal{C}^{-1}\left[{\tilde{f}_{\epsilon}(\omega)}\right]\!\biggm|_{t=T}\right) = The can written the terms of ${\bar{\epsilon}(\omega)}$ as follows: $$\label{eq:epsTepsw} \epsilon(T) = \sqrt{\frac{2}{\pi}}\int_0^\Omega{\bar{\epsilon}(\omega)}\cos(\omega T)\,d\omega = 0$$ Substituting Eq. in we obtain: $$\label{eq:lambdas2} {\epsilon_{unc}}(T) - \left(\lambda_0\,\mathcal{C}^{-1}\left[{\tilde{f}_{\epsilon}(\omega)}\right]\!\biggm|_{t=T} + \lambda_T\,\mathcal{C}^{-1}\left[{\tilde{f}_{\epsilon}(\omega)}\cos(\omega = 0$$ Eqs. ,, a two equation system of two $\lambda_0$, $\lambda_T$. us for $$\begin{aligned} & a \mathcal{C}^{-1}\left[{\tilde{f}_{\epsilon}(\omega)}\right |
{\bar{\epsilon}(\omega)}= {\bar{\epsilon}_{Unc}(\omega)}- {\tiLde{f}_{\ePsiLon}(\OmEga)}[\lAmbdA_0 + \lambda_T\cos(\omeGA T)]$$ NoW, in order to find explicit ExpreSsIOns tO $\LaMbda_0$ aNd $\lambdA_t$, wE SHouLd EnForCe THe BoundAry ConstraInts,,, on the sOluTiOn. We start froM ThE constrainT. It Can be rewrittEn iN the teRmS of ${\BAr{\epsIloN}(\omegA)}$ as folLOws: $$\labEl{eq:eps0epSw}
\EPsilon(0) = \SQrt{\frac{2}{\PI}}\InT_0^\OmeGa{\bar{\epsilon}(\omega)}\COs(0)\,D\Omega = \sqrt{\frac{2}{\pI}}\int_0^\OmEgA{\BaR{\EPsiLon}(\Omega)}\,d\omegA = 0$$ LEt us sUBstitutE ThE SOLutIOn in the constrAint equatioN: $$\LabEl{eq:laMbDas1}
{\EPsilon_{Unc}}(0) - \leFt(\LAmbDa_0\,\mathcal{C}^{-1}\lEft[{\tIlde{f}_{\epsiLon}(\omeGA)}\right]\!\bIGgm|_{t=0} + \lamBda_T\,\maThcAl{C}^{-1}\Left[{\TIlDe{F}_{\epSiLOn}(\oMEgA)}\riGHt]\!\bIggm|_{t=T}\riGhT) = 0$$ THe conStraINT CAn be WriTten In the Terms of ${\bar{\epsIloN}(\omeGA)}$ as FolloWs: $$\labEl{eq:EpSTepsW}
\epsilOn(T) = \sqRt{\Frac{2}{\pi}}\int_0^\Omega{\bAr{\epSilon}(\omegA)}\coS(\oMegA T)\,D\omegA = 0$$ substiTutIng eq. in Eq. wE obtain: $$\LAbeL{eQ:LAMbDas2}
{\epsilon_{unc}}(T) - \left(\LaMBDa_0\,\Mathcal{C}^{-1}\Left[{\tiLDe{F}_{\ePSilon}(\omeGa)}\RigHt]\!\biGGM|_{t=T} + \laMbda_t\,\MaThcal{C}^{-1}\leFt[{\tildE{F}_{\ePsIlon}(\omeGa)}\Cos(\omeGa t)\riGht]\!\Biggm|_{T=t}\rigHt) = 0$$ Eqs. ,, cOnstitutE a two EQuation system oF The two variablES $\lAMBdA_0$, $\LambDa_T$. let us denote, For cONvenIencE: $$\BeGin{ALigneD}
& a \equIv \MAtHCal{C}^{-1}\left[{\tilde{f}_{\epsilOn}(\Omega)}\rIght |
{\bar{\epsilon}(\omeg a)}= {\bar {\eps ilo n}_ {u nc}( \ome ga)}- {\tilde{ f }_{\ epsilon}(\omega)}[\lam bda_0 + \lam b da _T\co s(\omeg a T ) ] $$No w, in o r de r tofin d expli cit expres sio ns to $\lambda _ 0$ and $\lam bda _T$, we shou ldenforc ethe bound ary cons traint s ,,, on the solu ti o n. Wes tart fr o m t he c onstraint. It can be rewritten in t he ter ms of $ {\b ar{ \epsilon}( \o mega) } $ as fo l lo w s : $$ \ label{eq:eps0 epsw}
\ e psi lon(0) = \s q rt{\fr ac{2} {\ p i}} \int_0^\Ome ga{\ bar{\epsi lon}(\ o mega)}\ c os(0)\, d\omeg a = \s qrt{ \ fr ac {2} {\ p i}} \ in t_0 ^ \Om ega{\bar {\ ep silon }(\o m e g a )}\, d\o mega = 0$ $ Let us subs tit utet hesolut ion i n th econst raintequat io n: $$\label{eq: lamb das1}
{\ ep sil on _{unc } }(0) - \l eft (\lambd a_0\,\m a thc al { C } ^{ -1}\left[{\tilde{f }_ { \ ep silon}(\ omega) } \r ig h t]\!\big gm |_{ t=0} + \lam bda_ T \, \mathcal {C}^{- 1 }\ le ft[{\ti ld e{f}_{ \e psi lon }(\om e ga)} \right ]\!\bigg m|_{t = T}\right) = 0$ $ The constrai n tc a nb e wr itt en in the t erms of $ {\ba r {\ eps i lon}( \omeg a) } $a s follows: $$\label {e q:epsT epsw}
\epsilon (T) = \sqr t { \ frac{2}{ \pi} } \i n t_0^\Omega{\ba r{\ep silon}(\om e ga)}\cos (\ome ga T)\,d \omega =0 $ $ Substi tut ing Eq . i n E q. we obtain : $$\l ab el{eq:l amb das2}
{\e psi lon _{ unc}}(T)- \left( \l am bd a_ 0\, \math c al{C}^{- 1} \le ft [{\ tilde { f}_{\e psilo n}(\ om eg a )}\ right]\ ! \b i g gm|_ {t =T } +\la mb da_T\ ,\ma t hca l{C}^{- 1}\left[{ \ti l de{f }_ {\ epsilon }(\omega)}\co s( \omega T)\ ri ght ]\!\bi g g m|_{t=T} \right) = 0$$ Eqs. ,, c o nstitut e a twoequa tion syst emof the tw o varia bles $ \lamb da _0$ , $\lam b d a_ T$. L et us deno t e , f or co nv enie nce: $$ \begin{aligned}
&a \equiv \mat hca l{C} ^ { -1 }\l e ft [ {\t il d e{f } _ {\epsilon}(\ome ga)}\right |
_ _{\bar{\epsilon}(\omega)}= {\bar{\epsilon}_{unc}(\omega)}- {\tilde{f}_{\epsilon}(\omega)}[\lambda_0 +_\lambda_T\cos(\omega T)]$$_Now,_in order_to_find explicit expressions_to $\lambda_0$ and_$\lambda_T$, we should enforce_the boundary constraints,,,_on_the solution. We start from the constraint. It can be rewritten in the terms_of_${\bar{\epsilon}(\omega)}$ as_follows:_$$\label{eq:eps0epsw}
_ \epsilon(0) =_\sqrt{\frac{2}{\pi}}\int_0^\Omega{\bar{\epsilon}(\omega)}\cos(0)\,d\omega = \sqrt{\frac{2}{\pi}}\int_0^\Omega{\bar{\epsilon}(\omega)}\,d\omega = 0$$_Let us_substitute the solution in the constraint equation: $$\label{eq:lambdas1}
__ {\epsilon_{unc}}(0)_- \left(\lambda_0\,\mathcal{C}^{-1}\left[{\tilde{f}_{\epsilon}(\omega)}\right]\!\biggm|_{t=0} + \lambda_T\,\mathcal{C}^{-1}\left[{\tilde{f}_{\epsilon}(\omega)}\right]\!\biggm|_{t=T}\right) = 0$$ The constraint can_be written in the terms of_${\bar{\epsilon}(\omega)}$ as follows:_$$\label{eq:epsTepsw}
__ \epsilon(T) =_\sqrt{\frac{2}{\pi}}\int_0^\Omega{\bar{\epsilon}(\omega)}\cos(\omega T)\,d\omega = 0$$ Substituting Eq. _in Eq. we obtain: $$\label{eq:lambdas2}
_ {\epsilon_{unc}}(T) - \left(\lambda_0\,\mathcal{C}^{-1}\left[{\tilde{f}_{\epsilon}(\omega)}\right]\!\biggm|_{t=T} + \lambda_T\,\mathcal{C}^{-1}\left[{\tilde{f}_{\epsilon}(\omega)}\cos(\omega_T)\right]\!\biggm|_{t=T}\right) = 0$$ Eqs. ,, constitute a_two equation system of the_two variables_$\lambda_0$, $\lambda_T$. Let us denote,_for convenience: $$\begin{aligned}
_ _ & a_\equiv \mathcal{C}^{-1}\left[{\tilde{f}_{\epsilon}(\omega)}\right |
0} \frac{\mathcal{O}(\sigma)}{\left( \det_{a,b} \partial_b J^a_{\rm eff} \right) H_X H_Y},
\\
& = & \frac{1}{|\mathcal{W}|} \sum_{J^a_{\rm eff}=0} \frac{\mathcal{O}(\sigma) \prod_{\tilde{\mu}} \left( \sum_{a=1}^r \sigma_a \alpha_{\tilde{\mu}}^a \right)}{\left( \det_{a,b} \partial_b J^a_{\rm eff} \right) \left(\prod_i E_{i}(\sigma)\right)},\end{aligned}$$ which is the same as the A/2 correlation function computed from the original GLSM [@Closset:2015ohf equ’n (3.63)]. (The factor of $1/| {\mathcal W}|$ reflects the Weyl orbifold, which acts freely on the critical locus, as in [@Gu:2018fpm], so that twisted sectors do not enter this computation, at least for mirrors to theories with connected gauge groups.)
Specialization to abelian theories {#sect:specialize}
==================================
Let’s consider a GLSM with gauge group $U(1)^r$. The chiral field $\Phi_i$ and Fermi field $\Psi_i$ have charge $Q^a_i$ under the $a$-th $U(1)$, for $i=1, \cdots, N$. Assuming linear and diagonal (0,2) deformations, as discussed before, these fields satisfy $$\overline{D}_+ \Psi_i \: = \: \sum_{a=1}^r E_i^a \sigma_a \Phi_i,$$ where $E_i^a=Q_i^a$ on the (2,2) locus.
In the abelian case, the fields $X_\mu$ and $\Lambda_\mu$ are absent in the mirror theory. The matter content of the mirror Landau-Ginzburg model thus consists of chiral fields $\sigma_a, Y_i$ and Fermi fields $\Upsilon_a, F_i$, $a=1,\cdots,r, i=1,\cdots,N$. The superpotential is $$W \: = \:
\sum_{a=1}^r \Upsilon_a \left( \sum_{i= | 0 } \frac{\mathcal{O}(\sigma)}{\left (\det_{a, b } \partial_b J^a_{\rm eff } \right) H_X H_Y },
\\
& = & \frac{1}{|\mathcal{W}| } \sum_{J^a_{\rm eff}=0 } \frac{\mathcal{O}(\sigma) \prod_{\tilde{\mu } } \left (\sum_{a=1}^r \sigma_a \alpha_{\tilde{\mu}}^a \right)}{\left (\det_{a, b } \partial_b J^a_{\rm eff } \right) \left(\prod_i E_{i}(\sigma)\right)},\end{aligned}$$ which is the same as the A/2 correlation affair calculate from the original GLSM [ @Closset:2015ohf equ’n (3.63) ]. (The factor of $ 1/| { \mathcal W}|$ reflects the Weyl orbifold, which act freely on the critical venue, as in [ @Gu:2018fpm ], so that twisted sectors do not enter this calculation, at least for mirrors to theories with affiliated gauge groups .)
Specialization to abelian theory { # sect: specialize }
= = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = =
lease ’s consider a GLSM with gauge group $ U(1)^r$. The chiral field $ \Phi_i$ and Fermi field $ \Psi_i$ have tear $ Q^a_i$ under the $ a$-th $ U(1)$, for $ i=1, \cdots, N$. Assuming linear and diagonal (0,2) deformations, as hash out before, these fields satisfy $ $ \overline{D}_+ \Psi_i \: = \: \sum_{a=1}^r E_i^a \sigma_a \Phi_i,$$ where $ E_i^a = Q_i^a$ on the (2,2) locus.
In the abelian event, the fields $ X_\mu$ and $ \Lambda_\mu$ are absent in the mirror theory. The matter content of the mirror Landau - Ginzburg exemplar thus consists of chiral fields $ \sigma_a, Y_i$ and Fermi fields $ \Upsilon_a, F_i$, $ a=1,\cdots, r, i=1,\cdots, N$. The superpotential is $ $ W \: = \:
\sum_{a=1}^r \Upsilon_a \left (\sum_{i= | 0} \frwc{\mathcal{O}(\sigma)}{\left( \det_{a,n} \partial_b J^a_{\rm gfd} \righv) H_X H_Y},
\\
& = & \frac{1}{|\mxthcal{W}|} \sum_{J^a_{\rm eff}=0} \frac{\mathral{O}(\wigma) \prod_{\tilde{\mu}} \left( \sum_{a=1}^f \sigma_a \wlpha_{\tilee{\mu}}^e \right)}{\left( \det_{a,u} \partial_b J^a_{\rj eff} \cight) \left(\prod_i E_{i}(\sigma)\richt)},\end{aligned}$$ fhkck is the same as the A/2 correlation ftnction clmputed from tre ogidinam GLSM [@Closset:2015ohf equ’n (3.63)]. (The factor of $1/| {\mauhcal W}|$ reflects tne Weyl orbifold, which actd frfely on the criticwl locus, as in [@Dy:2018fpm], so that twisted stccors do not enter this computation, at leasg for mirrors ti rhelties with coinecteq gauge grouix.)
Specidlizatipn to abelian bheormes {#wect:specialize}
==================================
Let’s coisider a GLSM with gwuge grou[ $B(1)^r$. The chiral field $\Pyi_u$ and Ferki fkwld $\Psj_i$ hzve chwrgx $Q^a_i$ under the $a$-th $U(1)$, dor $i=1, \cdots, N$. Assumond linear and djagonaj (0,2) deformations, as discussed before, these fislds satisfy $$\overline{D}_+ \Psi_i \: = \: \sum_{a=1}^r E_i^a \sigla_a \Phi_i,$$ rhere $E_i^a=Q_i^a$ on the (2,2) locus.
In the abelian case, the fielvs $X_\mb$ and $\Uqmhda_\mu$ are absent in the mirror theory. The matesr cpntent of the iirror Landsu-Himeburg model thos consnata of chiral fields $\sigma_a, Y_i$ abd Fermi sielcs $\Upsilon_a, F_i$, $a=1,\cdots,r, i=1,\cdors,N$. The supegpotwntial is $$W \: = \:
\sum_{a=1}^x \Upsilon_a \lzft( \suk_{i= | 0} \frac{\mathcal{O}(\sigma)}{\left( \det_{a,b} \partial_b J^a_{\rm eff} \right) \\ = & \sum_{J^a_{\rm eff}=0} \frac{\mathcal{O}(\sigma) \right)}{\left( \partial_b J^a_{\rm eff} \left(\prod_i E_{i}(\sigma)\right)},\end{aligned}$$ which the same as the A/2 correlation computed from the original GLSM [@Closset:2015ohf equ’n (3.63)]. (The factor of $1/| {\mathcal reflects the Weyl orbifold, which acts freely on the critical locus, as in so twisted do enter this computation, at least for mirrors to theories with connected gauge groups.) Specialization to abelian {#sect:specialize} ================================== Let’s consider a GLSM with gauge $U(1)^r$. The chiral field and Fermi field $\Psi_i$ have $Q^a_i$ the $a$-th for \cdots, Assuming linear and (0,2) deformations, as discussed before, these fields satisfy $$\overline{D}_+ \Psi_i \: = \: \sum_{a=1}^r E_i^a \sigma_a \Phi_i,$$ $E_i^a=Q_i^a$ on locus. In abelian the $X_\mu$ and $\Lambda_\mu$ in the mirror theory. The matter mirror Landau-Ginzburg model thus consists of chiral fields Y_i$ and fields $\Upsilon_a, F_i$, $a=1,\cdots,r, i=1,\cdots,N$. The is $$W \: = \: \sum_{a=1}^r \Upsilon_a \left( | 0} \frac{\mathcal{O}(\sigma)}{\left( \det_{a,B} \partial_b J^A_{\rm efF} \riGht) h_X h_Y},
\\
& = & \frAc{1}{|\maThcal{W}|} \sum_{J^a_{\rm eFF}=0} \fraC{\mathcal{O}(\sigma) \prod_{\tildE{\mu}} \leFt( \SUm_{a=1}^r \SIgMa_a \alPha_{\tildE{\Mu}}^A \RIghT)}{\lEfT( \deT_{a,B} \PaRtial_B J^a_{\Rm eff} \riGht) \left(\proD_i E_{I}(\sIgma)\right)},\end{ALiGned}$$ which iS thE same as the A/2 cOrrElatioN fUncTIon coMpuTed frOm the oRIginal gLSM [@ClossEt:2015OHf equ’n (3.63)]. (tHe factoR OF $1/| {\mAthcAl W}|$ reflects the WeyL OrBIfold, which acts Freely On THe CRItiCal Locus, as in [@GU:2018fPm], so tHAt twistED sECTOrs DO not enter this Computation, AT leAst for MiRroRS to theOries WiTH coNnected gaugE groUps.)
SpeciaLizatiON to abelIAn theorIes {#secT:spEciAlizE}
==================================
leT’s ConSiDEr a glSm wiTH gaUge group $u(1)^r$. thE chirAl fiELD $\pHi_i$ aNd FErmi Field $\psi_i$ have chargE $Q^a_I$ undER thE $a$-th $U(1)$, For $i=1, \cDots, n$. ASsumiNg lineAr and DiAgonal (0,2) deformatiOns, aS discusseD beFoRe, tHeSe fieLDs satiSfy $$\OveRline{D}_+ \PSi_i \: = \: \sum_{a=1}^R e_i^a \SiGMA_A \PHi_i,$$ where $E_i^a=Q_i^a$ on thE (2,2) lOCUs.
in the abeLian caSE, tHe FIelds $X_\mu$ AnD $\LaMbda_\MU$ Are abSent IN tHe mirror Theory. tHe MaTter conTeNt of thE mIrrOr LAndau-gInzbUrg modEl thus coNsistS Of chiral fields $\SIgma_a, Y_i$ and FerMI fIELdS $\upsiLon_A, F_i$, $a=1,\cdots,r, i=1,\CdotS,n$. The SupeRPoTenTIal is $$w \: = \:
\sum_{a=1}^R \UPSiLOn_a \left( \sum_{i= | 0} \frac{\mathcal{O}(\sigm a)}{\left( \det _{a ,b} \ part ial_ b J^a_{\rm eff } \ri ght) H_X H_Y},
\\
& =& \fr ac { 1}{| \ ma thcal {W}|} \ s um _ { J^a _{ \r m e ff } =0 } \fr ac{ \mathca l{O}(\sigm a)\p rod_{\tilde{ \ mu }} \left(\su m_{a=1}^r \s igm a_a \a lp ha_ { \tild e{\ mu}}^ a \rig h t)}{\l eft( \det _{ a ,b} \p a rtial_b J ^a _{\r m eff} \right) \l e ft ( \prod_i E_{i}( \sigma )\ r ig h t )}, \en d{aligned} $$ whic h is the sa m e ast he A/2 correl ation funct i oncomput ed fr o m theorigi na l GL SM [@Closse t:20 15ohf equ ’n (3. 6 3)]. (T h e facto r of $ 1/| {\ math c al W }|$ r e fle c ts th e We yl orbif ol d, whic h ac t s f reel y o n th e cri tical locus,asin [ @ Gu: 2018f pm],so t ha t twi sted s ector sdo not enter th is c omputatio n,at le as t for mirror s t o t heories with c o nne ct e d ga uge groups.)
Spec ia l i za tion toabelia n t he o ries {#s ec t:s peci a l ize}==== = == ======== ====== = == == ======= =
Let’s c ons ide r a G L SM w ith ga uge grou p $U( 1 )^r$. The chir a l field $\Phi _ i$ a nd Ferm i f ield $\Psi_ i$ h a ve c harg e $ Q^a _ i$ un der t he $a $ -th $U(1)$, for $i= 1, \cdot s, N$ . Assuming li near and d i a g onal (0, 2) d e fo r mations, as di scuss ed before, these fi eldssatisfy$$\overli n e {D}_+ \P si_ i \ : = \: \ su m_{a=1}^r E_i ^ a \si gm a_a \Ph i_i ,$$ whe re$E_ i^a =Q_ i^ a$ on the (2,2) l oc us .
I n t he ab e lian cas e, th efie lds $ X _\mu$and $ \Lam bd a_ \ mu$ are ab s en t in t he m irro r t he ory.Them att er cont ent of th e m i rror L an dau-Gin zburg model t hu s consists o f c hiralf i elds $\s igma_a, Y_i$ and Fermif ields $ \Up silon _a,F_i$, $a= 1,\ cdots, r,i =1,\cd ots,N$ . The s upe r p otent i a lis$$ W \: = \:\ s um_ {a=1} ^r \Up silon_a \left( \sum_{i= | 0} \frac{\mathcal{O}(\sigma)}{\left(_\det_{a,b} \partial_b_J^a_{\rm eff} \right) H_X_H_Y},
\\
& =_&_\frac{1}{|\mathcal{W}|} \sum_{J^a_{\rm_eff}=0}_\frac{\mathcal{O}(\sigma) \prod_{\tilde{\mu}} \left(_\sum_{a=1}^r \sigma_a \alpha_{\tilde{\mu}}^a_\right)}{\left( \det_{a,b} \partial_b J^a_{\rm_eff} \right) \left(\prod_i_E_{i}(\sigma)\right)},\end{aligned}$$_which is the same as the A/2 correlation function computed from the original GLSM_[@Closset:2015ohf_equ’n (3.63)]._(The_factor_of $1/| {\mathcal W}|$ reflects_the Weyl orbifold, which acts_freely on_the critical locus, as in [@Gu:2018fpm], so that_twisted_sectors do not_enter this computation, at least for mirrors to theories_with connected gauge groups.)
Specialization to abelian_theories {#sect:specialize}
==================================
Let’s consider_a_GLSM_with gauge group $U(1)^r$._The chiral field $\Phi_i$ and Fermi_field $\Psi_i$ have charge $Q^a_i$ under_the $a$-th $U(1)$, for $i=1, \cdots, N$._Assuming linear and diagonal (0,2) deformations,_as discussed before, these fields_satisfy $$\overline{D}_+_\Psi_i \: = \: \sum_{a=1}^r_E_i^a \sigma_a \Phi_i,$$_where $E_i^a=Q_i^a$_on the (2,2)_locus.
In the abelian case, the fields_$X_\mu$ and $\Lambda_\mu$_are absent in the mirror theory._The_matter content of_the_mirror_Landau-Ginzburg model_thus consists of_chiral_fields $\sigma_a,_Y_i$_and Fermi fields $\Upsilon_a, F_i$, $a=1,\cdots,r,_i=1,\cdots,N$._The superpotential is $$W \: = \:
\sum_{a=1}^r_\Upsilon_a \left( \sum_{i= |
4}{3}$ which carry baryon number and lepton number, $B_{J_{1,\alpha}}=+\frac{1}{3}$, and $L_{J_\alpha}=-L_{J_1}=+2$. All masses are generated with four multiplets of scalar fields $$\begin{aligned}
\eta & \sim & ({\bf 1}, {\bf 3}, 0)
\nonumber \\
\rho & \sim & ({\bf 1}, {\bf 3},+1)
\nonumber \\
\chi & \sim & ({\bf 1}, {\bf 3},-1)
\nonumber \\
S_{ij} & \sim & ({\bf 1},{\bf\bar 6}_S,0)\end{aligned}$$ and in the symmetric phase of the theory they are parametrized by 30 real scalar fields. Such unavoidable scalarland is the most desirable field sector to have an experimental comprovation, since in theories with spontaneous symmetry breaking of the gauge symmetry it is essential but the unique field of the standard model which does not present evidences is the Higgs scalar boson. The total number of massless fields in the $G_{331}$ model is 110 and there are not spin-$\frac{3}{2}$ Rarita–Schwinger fields.
Our main purpose is to select the possible color gauge simple groups from the leptoquark-bilepton flavordynamics. Let us remark that in a theory whith the SU($n_c$)$_c$ gauge simple group the ’t Hooft [@thft] limit $n_c\rightarrow \infty$ and the Maldacena [@Mald] conjecture provide the evidence of the gauge to string theories limit. In the four-dimensional super Yang–Mills type IIB string theory arises the color confinement and a mass gap within the 5-brane of the eleven-dimensional M-theory [@Witten98]. The $n_c=3$ standard QCD is an asymptotically free theory including its non-perturbative confinement property. The perturbative strong coupling constant is $$\alpha_{\rm s}(q) = \frac{g^2_{{\rm SU}(n_c)_c}(q)}{4\pi} = 4\pi
\left [\beta_0
%%%%%%\left (\frac{11}{3} n_c - \frac{2}{3} n_f\right )
\ln\left (\frac{q}{\Lambda_{\rm QCD | 4}{3}$ which carry baryon number and lepton number, $ B_{J_{1,\alpha}}=+\frac{1}{3}$, and $ L_{J_\alpha}=-L_{J_1}=+2$. All masses are generate with four multiplets of scalar field $ $ \begin{aligned }
\eta & \sim & ({ \bf 1 }, { \bf 3 }, 0)
\nonumber \\
\rho & \sim & ({ \bf 1 }, { \bf 3},+1)
\nonumber \\
\chi & \sim & ({ \bf 1 }, { \bf 3},-1)
\nonumber \\
S_{ij } & \sim & ({ \bf 1},{\bf\bar 6}_S,0)\end{aligned}$$ and in the symmetric phase of the theory they are parametrized by 30 real scalar battlefield. Such unavoidable scalarland is the most desirable playing field sector to give birth an experimental comprovation, since in theories with spontaneous symmetry breaking of the bore symmetry it is essential but the unique field of the standard exemplar which does not present evidences is the Higgs scalar boson. The total number of massless fields in the $ G_{331}$ exemplar is 110 and there are not spin-$\frac{3}{2}$ Rarita – Schwinger fields.
Our main determination is to select the possible color gauge elementary groups from the leptoquark - bilepton flavordynamics. Let us remark that in a theory whith the SU($n_c$)$_c$ gauge childlike group the ’ t Hooft [ @thft ] limit $ n_c\rightarrow \infty$ and the Maldacena [ @Mald ] conjecture provide the evidence of the gauge to string theories limit. In the four - dimensional super Yang – Mills type IIB string theory arises the color confinement and a mass gap within the 5 - brane of the eleven - dimensional M - hypothesis [ @Witten98 ]. The $ n_c=3 $ standard QCD is an asymptotically complimentary theory admit its non - perturbative parturiency property. The perturbative strong coupling changeless is $ $ \alpha_{\rm s}(q) = \frac{g^2_{{\rm SU}(n_c)_c}(q)}{4\pi } = 4\pi
\left [ \beta_0
% %% %% % \left (\frac{11}{3 } n_c - \frac{2}{3 } n_f\right)
\ln\left (\frac{q}{\Lambda_{\rm QCD | 4}{3}$ whlch carry baryon number xnd lepton numbgr, $B_{J_{1,\alpia}}=+\frac{1}{3}$, znd $L_{J_\aloha}=-L_{J_1}=+2$. All masses are generatxd wuth fiur multiplets of scalxr fields $$\begin{alugnev}
\eta & \sim & ({\bf 1}, {\bh 3}, 0)
\nonumnzr \\
\rhk & \sik & ({\bf 1}, {\bf 3},+1)
\nonumner \\
\chi & \shm & ({\bf 1}, {\bf 3},-1)
\nongmcex \\
S_{ij} & \sim & ({\bf 1},{\bf\bar 6}_S,0)\end{aligned}$$ anq in thr dymmetric phasg of uhe thekgy they are parametrized by 30 rezl scaler fields. Such inavoidable scalarland is hhe lost desirable fiepd sector ti hadw an experimdntal compgmvation, sihce in theories with spontaneour symketry breajibg lx the gauge symmvtry it is essential tut the unique field pf vhe wtandard model which voes not present eviqences is tke Higgs scalar boson. Tye tojal ngmbef of maaskeas fiepds in the $G_{331}$ jodel is 110 abd there are not spon-$\sgsc{3}{2}$ Rarita–Schsinger fyelds.
Our main purpose is to select the kossigle color gauge simple troups from the leptoeuark-bile[ton flavordynamics. Let us remark that in a theorf whivh tht WU($n_c$)$_c$ takge simple group the ’t Hooft [@thft] limit $n_c\righezrtoe \infty$ and thc Maldacena [@Mald] comjfcyore provide thg evideude of the gauge to shring treoriws limit. Yn tne four-dimensional super Yabg–Mills type UIB string theory crises the culor confonement and a mass gap cithin the 5-brane lf the elsxen-dimensional M-gheprf [@Witten98]. Ufe $n_c=3$ standard QCQ is an awympcoticallh frge theowy includijg its non-perturbative clnfingment [roperty. Tje perturbative strong coupling constant is $$\akpva_{\rk s}(q) = \frcc{g^2_{{\rm XU}(n_c)_c}(q)}{4\pi} = 4\pi
\lqft [\beta_0
%%%%%%\left (\ftac{11}{3} n_c - \fxac{2}{3} n_f\fight )
\ln\lert (\frac{x}{\Lambda_{\rm QSD | 4}{3}$ which carry baryon number and lepton and All masses generated with four \eta \sim & ({\bf {\bf 3}, 0) \\ \rho & \sim & ({\bf {\bf 3},+1) \nonumber \\ \chi & \sim & ({\bf 1}, {\bf 3},-1) \nonumber S_{ij} & \sim & ({\bf 1},{\bf\bar 6}_S,0)\end{aligned}$$ and in the symmetric phase of theory are by real scalar fields. Such unavoidable scalarland is the most desirable field sector to have an experimental since in theories with spontaneous symmetry breaking of gauge symmetry it is but the unique field of standard which does present is Higgs scalar boson. total number of massless fields in the $G_{331}$ model is 110 and there are not spin-$\frac{3}{2}$ Rarita–Schwinger Our main to select possible gauge groups from the Let us remark that in a SU($n_c$)$_c$ gauge simple group the ’t Hooft [@thft] $n_c\rightarrow \infty$ the Maldacena [@Mald] conjecture provide the of the gauge to string theories limit. In four-dimensional super Yang–Mills type IIB string theory arises the color confinement and a mass gap 5-brane of the eleven-dimensional [@Witten98]. The $n_c=3$ QCD an free including its confinement property. The perturbative strong coupling constant is $$\alpha_{\rm s}(q) = SU}(n_c)_c}(q)}{4\pi} = 4\pi \left [\beta_0 %%%%%%\left (\frac{11}{3} n_c - \frac{2}{3} \ln\left QCD | 4}{3}$ which carry baryon number and Lepton numbEr, $B_{J_{1,\aLphA}}=+\frAc{1}{3}$, And $L_{j_\alpHa}=-L_{J_1}=+2$. All masses aRE genErated with four multipleTs of sCaLAr fiELdS $$\begiN{aligneD}
\EtA & \SIm & ({\bF 1}, {\bF 3}, 0)
\nOnuMbER \\
\rHo & \sim & ({\Bf 1}, {\bF 3},+1)
\nonumbEr \\
\chi & \sim & ({\bf 1}, {\Bf 3},-1)
\nOnUmber \\
S_{ij} & \sim & ({\bF 1},{\Bf\Bar 6}_S,0)\end{aliGneD}$$ and in the symMetRic phaSe Of tHE theoRy tHey arE paramETrized By 30 real scaLaR Fields. sUch unavOIDaBle sCalarland is the mosT DeSIrable field secTor to hAvE An EXPerImeNtal comproVaTion, sINce in thEOrIES WitH Spontaneous syMmetry breakINg oF the gaUgE syMMetry iT is esSeNTiaL but the uniqUe fiEld of the sTandarD Model whICh does nOt presEnt EviDencES iS tHe HIgGS scALaR boSOn. THe total nUmBeR of maSsleSS FIElds In tHe $G_{331}$ mOdel iS 110 and there are nOt sPin-$\fRAc{3}{2}$ RArita–schwiNger FiElds.
OUr main PurpoSe Is to select the poSsibLe color gaUge SiMplE gRoups FRom the LepToqUark-bilEpton flAVorDyNAMIcS. Let us remark that in A tHEOrY whith thE SU($n_c$)$_c$ GAuGe SImple groUp The ’T HooFT [@Thft] lImit $N_C\rIghtarroW \infty$ ANd ThE MaldacEnA [@Mald] cOnJecTurE provIDe thE evideNce of the Gauge TO string theorieS Limit. In the fouR-DiMENsIOnal SupEr Yang–Mills Type iiB stRing THeOry ARises The coLoR CoNFinement and a mass gap WiThin thE 5-branE of the eleven-dImensional m-THEory [@WittEn98]. ThE $N_c=3$ STandard QCD is an AsympTotically fREe theory IncluDing its nOn-perturbATIve confiNemEnt ProPerTY. thE perturbative STRong CoUpling cOnsTant is $$\aLphA_{\rm S}(q) = \fRac{G^2_{{\rM SU}(n_c)_c}(q)}{4\pi} = 4\Pi
\left [\beTa_0
%%%%%%\LeFt (\FrAc{11}{3} n_C - \frac{2}{3} N_F\right )
\ln\LeFt (\fRaC{q}{\LAmbda_{\RM QCD | 4}{3}$ which carry baryonnumber and lept onnum be r, $ B_{J _{1,\alpha}}=+ \ frac {1}{3}$, and $L_{J_\al pha}= -L _ {J_1 } =+ 2$. A ll mass e sa r e g en er ate dw it h fou r m ultiple ts of scal arfi elds $$\begi n {a ligned}
\e ta& \sim & ({\ bf1}, {\ bf 3} , 0)
\ non umber \\
\r h o & \s im & ({\b f1 }, {\b f 3},+1) \ no numb er \\
\chi & \si m & ({\bf 1}, {\bf 3},-1 )\ no n u mbe r \ \
S_{ij}&\sim& ({\bf1 }, { \ b f\b a r 6}_S,0)\end {aligned}$$ and in th esym m etricphase o f th e theory th ey a re parame trized by 30 r e al scal ar fie lds . S uchu na vo ida bl e sc a la rla n d i s the mo st d esira blef i e l d se cto r to have an experimen tal com p rov ation , sin ce i ntheor ies wi th sp on taneous symmetr y br eaking of th egau ge symm e try it is es sential but th e un iq u e fi eld of the standar dm o de l whichdoes n o tpr e sent evi de nce s is t he Hi ggss ca lar boso n. The to ta l numbe rof mas sl ess fi eldsi n th e $G_{ 331}$ mo del i s 110 and there are not spin- $ \f r a c{ 3 }{2} $ R arita–Schwi nger fiel ds.Ou r m a in pu rpose i s t o select the possibl ecolorgauge simple group s from the l e ptoquark -bil e pt o n flavordynami cs. L et us rema r k that i n a t heory wh ith the S U ( $n_c$)$_ c$gau gesim p l egroup the ’tH o oft[@ thft] l imi t $n_c\ rig hta rro w \ in fty$ andthe Mald ac en a[@ Mal d] co n jecturepr ovi de th e evi d ence o f the gau ge t o st ring th e or i e s li mi t. Inthe f our-d imen s ion al supe r Yang–Mi lls type I IB string theory arise sthe colorco nfi nement a nd a mas s gap within the 5-bran e of the el even- dime nsional M -th eory [ @Wi t ten98] . The$n_c= 3$ st a n dardQ C Disan asymptoti c a lly free t heor y inclu ding its non-pertu r bat ive confineme ntprop e r ty . T h ep ert ur b ati v e strong couplin g constant i s $ $\alpha_{\ r m s }( q) = \f rac{g^2 _{{\r m SU}(n_ c)_c}(q)} {4\pi} =4\ pi
\ l e ft[\beta_0
% %%%%%\le ft (\frac { 11}{3 } n _c -\fr ac{2}{ 3} n_ f\rig ht )
\ l n\l eft ( \frac{ q} {\Lamb da_{\ rm QCD | 4}{3}$ which_carry baryon_number and lepton number,_$B_{J_{1,\alpha}}=+\frac{1}{3}$, and_$L_{J_\alpha}=-L_{J_1}=+2$._All masses_are_generated with four_multiplets of scalar_fields $$\begin{aligned}
\eta & \sim_& ({\bf 1},_{\bf_3}, 0)
\nonumber \\
\rho & \sim & ({\bf 1}, {\bf 3},+1)
\nonumber \\
\chi & \sim_&_({\bf 1},_{\bf_3},-1)
\nonumber_\\
S_{ij} & \sim &_({\bf 1},{\bf\bar 6}_S,0)\end{aligned}$$ and in_the symmetric_phase of the theory they are parametrized by_30_real scalar fields._Such unavoidable scalarland is the most desirable field sector_to have an experimental comprovation, since_in theories with_spontaneous_symmetry_breaking of the gauge_symmetry it is essential but the_unique field of the standard model_which does not present evidences is the_Higgs scalar boson. The total number_of massless fields in the_$G_{331}$ model_is 110 and there are_not spin-$\frac{3}{2}$ Rarita–Schwinger_fields.
Our main_purpose is to_select the possible color gauge simple_groups from the_leptoquark-bilepton flavordynamics. Let us remark that_in_a theory whith_the_SU($n_c$)$_c$_gauge simple_group the ’t_Hooft [@thft]_limit $n_c\rightarrow_\infty$_and the Maldacena [@Mald] conjecture provide the_evidence_of the gauge to string theories limit._In the four-dimensional super_Yang–Mills_type IIB string theory_arises the color confinement and_a mass gap within the 5-brane_of the_eleven-dimensional M-theory [@Witten98]._The $n_c=3$ standard QCD is an asymptotically free theory including its_non-perturbative confinement property. The perturbative strong_coupling constant is $$\alpha_{\rm_s}(q) =_\frac{g^2_{{\rm_SU}(n_c)_c}(q)}{4\pi} = 4\pi
\left_[\beta_0
%%%%%%\left_(\frac{11}{3} n_c_- \frac{2}{3} n_f\right )
\ln\left (\frac{q}{\Lambda_{\rm QCD |
- \cos\bigl(\frac{2m\pi}{p}\bigr)}e^{-\frac{1}{g}\Gamma_{ba}(z_*)}}
{\sin\bigl(\frac{n\pi}{q}\bigr)
\sin\bigl(\frac{m\pi}{p}\bigr)
\sqrt{\sin\bigl(\frac{q-p}{p}m\pi\bigr)
\sin\bigl(\frac{q-p}{q}n \pi\bigr)}}
\label{Dab0}\end{aligned}$$ with $$\begin{aligned}
\Gamma_{ba}(z_*)&=-\Gamma_{ab}(z_*) {\nonumber}\\
&=+\frac{2pq\beta}{(q^2-p^2)}\,
\mu^{(q+p)/2p}\,
\sin\left(\frac{q-p}{q}n\pi\right)\,
\sin\left(\frac{q-p}{p}m\pi\right)
\label{Dab00}.\end{aligned}$$
The D-instanton action $\Gamma_{ba}(z_*)$ at the saddle points can be identified with the $(m,n)$ ZZ brane amplitude $Z_{ZZ}(m,n)$ by using its relation [@Mar] to the FZZT disk amplitudes as $$\begin{aligned}
Z_{ZZ}(m,n)={\left\langle \varphi_0(\zeta(z_{mn}^+)) \right\rangle}^{\!(0)}
-{\left\langle \varphi_0(\zeta(z_{mn}^-)) \right\rangle}^{\!(0)}
={\left\langle \varphi_b(\zeta_*) \right\rangle}^{\!\!(0)}-{\left\langle \varphi_a(\zeta_*) \right\rangle}^{\!\!(0)}, \end{aligned}$$ and thus we obtain $$\begin{aligned}
Z_{ZZ}(m,n)=\Gamma_{ba}(z_*)
=\Bigl< \varphi_b\bigl(\zeta(z_*(a,b;n))\bigr) \Bigr>^{\!(0)}
-\Bigl< \varphi_a\bigl(\zeta(z_*(a,b;n))\bigr) \Bigr>^{\!(0)}.\end{aligned}$$ Note that the expression [(\[Dab0 | - \cos\bigl(\frac{2m\pi}{p}\bigr)}e^{-\frac{1}{g}\Gamma_{ba}(z _ *) } }
{ \sin\bigl(\frac{n\pi}{q}\bigr)
\sin\bigl(\frac{m\pi}{p}\bigr)
\sqrt{\sin\bigl(\frac{q - p}{p}m\pi\bigr)
\sin\bigl(\frac{q - p}{q}n \pi\bigr) } }
\label{Dab0}\end{aligned}$$ with $ $ \begin{aligned }
\Gamma_{ba}(z_*)&=-\Gamma_{ab}(z _ *) { \nonumber}\\
& = + \frac{2pq\beta}{(q^2 - p^2)}\,
\mu^{(q+p)/2p}\,
\sin\left(\frac{q - p}{q}n\pi\right)\,
\sin\left(\frac{q - p}{p}m\pi\right)
\label{Dab00}.\end{aligned}$$
The D - instanton action $ \Gamma_{ba}(z_*)$ at the saddle points can be identified with the $ (m, n)$ ZZ brane amplitude $ Z_{ZZ}(m, n)$ by use its relation back [ @Mar ] to the FZZT disk amplitudes as $ $ \begin{aligned }
Z_{ZZ}(m, n)={\left\langle \varphi_0(\zeta(z_{mn}^+) ) \right\rangle}^{\!(0) }
-{\left\langle \varphi_0(\zeta(z_{mn}^-) ) \right\rangle}^{\!(0) }
= { \left\langle \varphi_b(\zeta _ *) \right\rangle}^{\!\!(0)}-{\left\langle \varphi_a(\zeta _ *) \right\rangle}^{\!\!(0) }, \end{aligned}$$ and thus we prevail $ $ \begin{aligned }
Z_{ZZ}(m, n)=\Gamma_{ba}(z _ *)
= \Bigl < \varphi_b\bigl(\zeta(z_*(a, b;n))\bigr) \Bigr>^{\!(0) }
-\Bigl < \varphi_a\bigl(\zeta(z_*(a, b;n))\bigr) \Bigr>^{\!(0)}.\end{aligned}$$ Note that the expression [ (\[Dab0 | - \cod\bigl(\frac{2m\pi}{p}\bigr)}e^{-\frac{1}{g}\Gxmma_{ba}(z_*)}}
{\sin\bmgl(\frac{h\pi}{q}\bigr)
\sin\bigl(\frac{m\pi}{p}\uigr)
\sqrt{\sin\bigl(\frac{d-p}{p}m\pi\bigg)
\wun\bigl(\frac{q-p}{q}n \pi\nngr)}}
\lznel{Day0}\eid{aligned}$$ with $$\negin{aligneg}
\Gamma_{ba}(z_*)&=-\Gammd_{ac}(z_*) {\nonumber}\\
&=+\frac{2pq\beta}{(q^2-p^2)}\,
\mu^{(q+p)/2p}\,
\sin\lqft(\frac{a-p}{e}n\pi\right)\,
\sin\jeft(\gwac{q-l}{i}m\ki\right)
\label{Dab00}.\end{aligned}$$
The D-inatanton action $\Gamma_{na}(z_*)$ at the saddle points cwn bf identified with hhe $(m,n)$ ZZ btzne qmplitude $Z_{ZX}(m,n)$ by using its relatjon [@Mar] to the FZZT disk amplitjdes cs $$\begin{alitnwd}
Z_{XE}(m,n)={\left\langlx \varpri_0(\zeta(z_{mn}^+)) \rinnt\rangne}^{\!(0)}
-{\lefy\langle \varphi_0(\deta(z_{kn}^-)) \eight\rangle}^{\!(0)}
={\left\langlx \varphi_b(\zeta_*) \right\rwngle}^{\!\!(0)}-{\left\naugle \varphi_a(\zeta_*) \righr\rqngle}^{\!\!(0)}, \eng{aliebed}$$ ans vhua we ohtamn $$\begin{alifned}
Z_{ZZ}(m,n)=\Gqmma_{ba}(z_*)
=\Bigl< \varpho_b\fpbl(\zeta(z_*(a,b;n))\bifr) \Bigw>^{\!(0)}
-\Bigl< \varphi_a\bigl(\zeta(z_*(a,b;n))\bigr) \Bigr>^{\!(0)}.\end{anighed}$$ Note that the exprewsion [(\[Dab0 | - \cos\bigl(\frac{2m\pi}{p}\bigr)}e^{-\frac{1}{g}\Gamma_{ba}(z_*)}} {\sin\bigl(\frac{n\pi}{q}\bigr) \sin\bigl(\frac{m\pi}{p}\bigr) \sqrt{\sin\bigl(\frac{q-p}{p}m\pi\bigr) \sin\bigl(\frac{q-p}{q}n \pi\bigr)}} $$\begin{aligned} {\nonumber}\\ &=+\frac{2pq\beta}{(q^2-p^2)}\, \sin\left(\frac{q-p}{q}n\pi\right)\, \sin\left(\frac{q-p}{p}m\pi\right) \label{Dab00}.\end{aligned}$$ the points can be with the $(m,n)$ brane amplitude $Z_{ZZ}(m,n)$ by using its [@Mar] to the FZZT disk amplitudes as $$\begin{aligned} Z_{ZZ}(m,n)={\left\langle \varphi_0(\zeta(z_{mn}^+)) \right\rangle}^{\!(0)} -{\left\langle \varphi_0(\zeta(z_{mn}^-)) ={\left\langle \varphi_b(\zeta_*) \right\rangle}^{\!\!(0)}-{\left\langle \varphi_a(\zeta_*) \right\rangle}^{\!\!(0)}, \end{aligned}$$ and thus we obtain $$\begin{aligned} Z_{ZZ}(m,n)=\Gamma_{ba}(z_*) =\Bigl< \Bigr>^{\!(0)} \varphi_a\bigl(\zeta(z_*(a,b;n))\bigr) Note the expression [(\[Dab0 | - \cos\bigl(\frac{2m\pi}{p}\bigr)}e^{-\frac{1}{g}\gamma_{ba}(z_*)}}
{\siN\bigl(\FraC{n\pI}{q}\Bigr)
\Sin\bIgl(\frac{m\pi}{p}\bigR)
\Sqrt{\Sin\bigl(\frac{q-p}{p}m\pi\bigr)
\sIn\bigL(\fRAc{q-p}{Q}N \pI\bigr)}}
\Label{DaB0}\EnD{ALigNeD}$$ wIth $$\BeGIn{AlignEd}
\GAmma_{ba}(z_*)&=-\gamma_{ab}(z_*) {\noNumBeR}\\
&=+\frac{2pq\beta}{(q^2-P^2)}\,
\Mu^{(Q+p)/2p}\,
\sin\left(\FraC{q-p}{q}n\pi\right)\,
\Sin\Left(\frAc{Q-p}{p}M\Pi\rigHt)
\lAbel{DAb00}.\end{aLIgned}$$
THe D-instanToN Action $\gAmma_{ba}(z_*)$ AT ThE sadDle points can be ideNTiFIed with the $(m,n)$ ZZ Brane aMpLItUDE $Z_{Zz}(m,n)$ By using its ReLatioN [@mar] to thE fZzt DIsk AMplitudes as $$\beGin{aligned}
Z_{zz}(m,n)={\Left\laNgLe \vARphi_0(\zeTa(z_{mn}^+)) \RiGHt\rAngle}^{\!(0)}
-{\left\laNgle \Varphi_0(\zetA(z_{mn}^-)) \riGHt\ranglE}^{\!(0)}
={\Left\lanGle \varPhi_B(\zeTa_*) \riGHt\RaNglE}^{\!\!(0)}-{\lEFt\lANgLe \vARphI_a(\zeta_*) \riGhT\rAngle}^{\!\!(0)}, \End{aLIGNEd}$$ anD thUs we ObtaiN $$\begin{aligned}
z_{ZZ}(M,n)=\GaMMa_{bA}(z_*)
=\BigL< \varpHi_b\bIgL(\zeta(Z_*(a,b;n))\biGr) \BigR>^{\!(0)}
-\BIgl< \varphi_a\bigl(\zEta(z_*(A,b;n))\bigr) \BiGr>^{\!(0)}.\eNd{AliGnEd}$$ NotE That thE exPreSsion [(\[DaB0 | - \cos\bigl(\frac{2m\pi}{p }\bigr)}e^ {-\fr ac{ 1}{ g} \Gam ma_{ ba}(z_*)}}
{\sin\bigl(\frac{n\ pi}{q }\ b igr)
\ s in \ b igl (\ fr ac{ m\ p i} {p}\b igr )
\sq rt{ \s in\bigl(\fra c {q -p}{p}m\pi \bi gr)
\sin\ bi gl( \ frac{ q-p }{q}n \pi\ b igr)}}
\label{ Da b 0}\end { aligned } $ $with $$\begin{aligned }
\ Gamma_{ba}(z_* )&=-\G am m a_ { a b}( z_* ) {\nonumb er }\\
& =+\frac { 2p q \ b eta } {(q^2-p^2)}\,
\mu^{(q+ p )/2 p}\,
\ sin \ left(\ frac{ q- p }{q }n\pi\right )\, \sin\le ft(\fr a c{q-p}{ p }m\pi\r ight) \l abe l{Da b 00 }. \en d{ a lig n ed }$$
Th e D-inst an to n act ion$ \ G a mma_ {ba }(z_ *)$ a t the saddlepoi ntsc anbe id entif iedwi th th e $(m, n)$ Z Zbrane amplitude $Z_ {ZZ}(m,n) $ b yusi ng itsr elatio n [ @Ma r] to t he FZZT dis ka m p li tudes as $$\begin{ al i g ne d}
Z_{ZZ }(m,n) = {\ le f t\langle \ var phi_ 0 ( \zeta (z_{ m n} ^+)) \ri ght\ra n gl e} ^{\!(0) } -{\l ef t\l ang le \v a rphi _0(\ze ta(z_{mn }^-)) \right\rangle} ^ {\!(0)}
={\l e ft \ l an g le \ var phi_b(\zeta _*)\ righ t\ra n gl e}^ { \!\!( 0)}-{ \l e ft \ langle \varphi_a(\z et a_*) \ right \rangle}^{\!\ !(0)}, \e n d { aligned} $$ a n dt hus we obtain$$\be gin{aligne d }
Z_{ZZ }(m,n )=\Gamma _{ba}(z_* ) =\Bigl < \ var phi _b\ b i gl (\zeta(z_*(a, b ; n))\ bi gr) \Bi gr> ^{\!(0) }
- \Bi gl< \ varphi_a\ bigl(\ze ta (z _* (a ,b; n))\b i gr) \Big r> ^{\ !( 0)} .\end { aligne d}$$Note t ha t th e expre s si o n [(\ [D ab 0 | - \cos\bigl(\frac{2m\pi}{p}\bigr)}e^{-\frac{1}{g}\Gamma_{ba}(z_*)}}
_ _ _ __ {\sin\bigl(\frac{n\pi}{q}\bigr)_
_ _ _ _ \sin\bigl(\frac{m\pi}{p}\bigr)
__ \sqrt{\sin\bigl(\frac{q-p}{p}m\pi\bigr)
__ ___ \sin\bigl(\frac{q-p}{q}n \pi\bigr)}}
_\label{Dab0}\end{aligned}$$ with $$\begin{aligned}
\Gamma_{ba}(z_*)&=-\Gamma_{ab}(z_*) {\nonumber}\\
_&=+\frac{2pq\beta}{(q^2-p^2)}\,
_\mu^{(q+p)/2p}\,
\sin\left(\frac{q-p}{q}n\pi\right)\,
\sin\left(\frac{q-p}{p}m\pi\right)
\label{Dab00}.\end{aligned}$$
The D-instanton action_$\Gamma_{ba}(z_*)$_at the saddle_points can be identified with the $(m,n)$ ZZ brane_amplitude $Z_{ZZ}(m,n)$ by using its relation_[@Mar] to the_FZZT_disk_amplitudes as $$\begin{aligned}
Z_{ZZ}(m,n)={\left\langle \varphi_0(\zeta(z_{mn}^+))_\right\rangle}^{\!(0)}
-{\left\langle \varphi_0(\zeta(z_{mn}^-)) \right\rangle}^{\!(0)}
={\left\langle_\varphi_b(\zeta_*) \right\rangle}^{\!\!(0)}-{\left\langle \varphi_a(\zeta_*) \right\rangle}^{\!\!(0)}, \end{aligned}$$_and thus we obtain $$\begin{aligned}
Z_{ZZ}(m,n)=\Gamma_{ba}(z_*)
_=\Bigl< \varphi_b\bigl(\zeta(z_*(a,b;n))\bigr) \Bigr>^{\!(0)}
-\Bigl<_\varphi_a\bigl(\zeta(z_*(a,b;n))\bigr) \Bigr>^{\!(0)}.\end{aligned}$$ Note that the_expression [(\[Dab0 |
5/1 hybrid amortizing ARM fees and points\
Freddie\_Regional\_NC\_1yrARM\_PriorSale & Freddie Mac 1-Year ARM North Central\
Freddie\_Regional\_NE\_1yrARM\_PriorSale & Freddie Mac 1-Year ARM North East\
Freddie\_Regional\_SE\_1yrARM\_PriorSale & Freddie Mac 1-Year ARM South East\
Freddie\_Regional\_SW\_1yrARM\_PriorSale & Freddie Mac 1-Year ARM South West\
Freddie\_Regional\_W\_1yrARM\_PriorSale & Freddie Mac 1-Year ARM West\
Freddie\_US\_15yr\_FRM\_PriorSale & Freddie Mac 15-Year fixed mortgage rate\
Freddie\_US\_15yr\_feespts\_PriorSale & Freddie Mac 15-Year fixed mortgage rate fees and points\
Freddie\_US\_1yrARM\_Fees\_PriorSale & Freddie Mac 1-Year ARM fees and points\
Freddie\_US\_1yrARM\_Margin\_PriorSale & Freddie Mac 1-Year ARM margin\
Freddie\_US\_1yrARM\_PriorSale & Freddie Mac 1-Year ARM\
Freddie\_US\_30yr\_FRM\_PriorSale & Freddie Mac 30-Year fixed mortgage rate\
Freddie\_US\_30yr\_feespts\_PriorSale & Freddie Mac 30-Year fixed mortgage rate fees and points\
\
N\_ShortSales & Number of short sales that occurred within the same zipcode as the property within the past 12 months.\
N\_PrePayments & Number of loans that have been prepaid\
N\_Outstanding\_Loans & How many loans are currently outstanding\
N\_PIF & Number of loans that were paid in full by reaching their term maturity\
N\_Originations & Number of loans originated at same location and same time period of sale\
Median\_Time\_to\_SAF & Median time between the original foreclosure month and the SAF month of all SAF that occurred in the given month.\
Median\_Time\_to\_NonInst\_SAF & Same as above but for non-institutional / non-corporate owner SAF.\
Median\_Time\_to\_SAF\_l12 & Same as above except that computation rolls | 5/1 hybrid amortizing ARM fees and points\
Freddie\_Regional\_NC\_1yrARM\_PriorSale & Freddie Mac 1 - Year sleeve North Central\
Freddie\_Regional\_NE\_1yrARM\_PriorSale & Freddie Mac 1 - class ARM North East\
Freddie\_Regional\_SE\_1yrARM\_PriorSale & Freddie Mac 1 - Year sleeve South East\
Freddie\_Regional\_SW\_1yrARM\_PriorSale & Freddie Mac 1 - class ARM South West\
Freddie\_Regional\_W\_1yrARM\_PriorSale & Freddie Mac 1 - Year ARM West\
Freddie\_US\_15yr\_FRM\_PriorSale & Freddie Mac 15 - Year pay back mortgage rate\
Freddie\_US\_15yr\_feespts\_PriorSale & Freddie Mac 15 - Year fixed mortgage pace fee and points\
Freddie\_US\_1yrARM\_Fees\_PriorSale & Freddie Mac 1 - Year sleeve fees and points\
Freddie\_US\_1yrARM\_Margin\_PriorSale & Freddie Mac 1 - Year sleeve margin\
Freddie\_US\_1yrARM\_PriorSale & Freddie Mac 1 - Year ARM\
Freddie\_US\_30yr\_FRM\_PriorSale & Freddie Mac 30 - Year fixed mortgage rate\
Freddie\_US\_30yr\_feespts\_PriorSale & Freddie Mac 30 - Year fixed mortgage pace fees and points\
\
N\_ShortSales & Number of short sales that occur within the same zipcode as the property within the past 12 months.\
N\_PrePayments & Number of loan that have been prepaid\
N\_Outstanding\_Loans & How many loans are presently outstanding\
N\_PIF & Number of loans that were yield in full by reaching their term maturity\
N\_Originations & Number of loan originated at same location and same time period of sale\
Median\_Time\_to\_SAF & Median time between the original foreclosure month and the SAF month of all SAF that occurred in the given month.\
Median\_Time\_to\_NonInst\_SAF & Same as above but for non - institutional / non - corporate owner SAF.\
Median\_Time\_to\_SAF\_l12 & Same as above except that computation rolls | 5/1 hjbrid amortizing ARM feer and points\
Frebeie\_Regmonal\_NC\_1grARM\_PriurSale & Freddie Mac 1-Year ARM Nirth Xentral\
Freddie\_Regional\_VE\_1yrARM\_PrporSale & Dredvie Mac 1-Year ARM North East\
Fresfie\_Rzgmonal\_SE\_1yrARM\_PriprSale & Fraddie Mac 1-Year AFM South East\
Freddie\_Regional\_SW\_1yrARM\_PryorSale & Vreddie Mac 1-Yewr AGM Soufh West\
Freddie\_Regional\_W\_1yrARM\_PriorSzle & Frtddie Mac 1-Year ARM West\
Freddie\_US\_15yr\_FRM\_PriorSape & Vreddie Mac 15-Year flxed mortgate rwre\
Freddie\_US\_15yf\_feespts\_PrpmrSale & Frgddie Mac 15-Year fixed mortgage rage fezs and poinjs\
Rrfgdie\_US\_1yrARM\_Hees\_PrporSale & Freddie Mac 1-Fear ARK fees and poikts\
Frxddiw\_US\_1yrARM\_Margin\_PriorSane & Freddie Mac 1-Yewr ARM masgnn\
Freddie\_US\_1yrARM\_PriorWaoe & Fteddia Maz 1-Yexr ZRK\
Fdeddie\_KS\_30yc\_FRM\_PriorSame & Freddie Mac 30-Year fixed morugadv rate\
Freddie\_HS\_30yr\_feqs[ts\_PriorSale & Freddie Mac 30-Year fixed mogtgafe rate fees and points\
\
B\_ShortSales & Number ov short swles that occurred within the same zipcode as the propxrgy clbhin rhf past 12 months.\
N\_PrePayments & Number of loans trzt hsve been prepald\
N\_Outstanding\_Loanx & Hpr many loans xre cuxdehtly outstanding\
N\_PLF & Numfer od loans trat eere paid in full by reachibg their teri maturity\
N\_Originatnons & Number of koans originated at same locction znd same tile period uf sale\
Median\_Timd\_to\_XAX & Median time between the original forzclosure monjh and ehe SAF mojth on all SAF that occugred nn tha given mojth.\
Median\_Time\_to\_NonInst\_SAF & Same as above but gos nmn-institbtionak / non-corporaee owner SAF.\
Mebian\_Time\_co\_SAF\_l12 & Same as zbove eecept that cjmputation ronps | 5/1 hybrid amortizing ARM fees and points\ Freddie 1-Year ARM Central\ Freddie\_Regional\_NE\_1yrARM\_PriorSale & East\ & Freddie Mac ARM South East\ & Freddie Mac 1-Year ARM South Freddie\_Regional\_W\_1yrARM\_PriorSale & Freddie Mac 1-Year ARM West\ Freddie\_US\_15yr\_FRM\_PriorSale & Freddie Mac 15-Year fixed rate\ Freddie\_US\_15yr\_feespts\_PriorSale & Freddie Mac 15-Year fixed mortgage rate fees and points\ Freddie\_US\_1yrARM\_Fees\_PriorSale Freddie 1-Year fees points\ Freddie\_US\_1yrARM\_Margin\_PriorSale & Freddie Mac 1-Year ARM margin\ Freddie\_US\_1yrARM\_PriorSale & Freddie Mac 1-Year ARM\ Freddie\_US\_30yr\_FRM\_PriorSale & Mac 30-Year fixed mortgage rate\ Freddie\_US\_30yr\_feespts\_PriorSale & Freddie 30-Year fixed mortgage rate and points\ \ N\_ShortSales & of sales that within same as the property the past 12 months.\ N\_PrePayments & Number of loans that have been prepaid\ N\_Outstanding\_Loans & How many are currently & Number loans were in full by term maturity\ N\_Originations & Number of same location and same time period of sale\ & Median between the original foreclosure month and SAF month of all SAF that occurred in given month.\ Median\_Time\_to\_NonInst\_SAF & Same as above but for non-institutional / non-corporate owner SAF.\ Median\_Time\_to\_SAF\_l12 as above except that rolls | 5/1 hybrid amortizing ARM fees anD points\
FreDdie\_REgiOnaL\_Nc\_1yrArM\_PrIorSale & Freddie mAc 1-YeAr ARM North Central\
FreddIe\_RegIoNAl\_NE\_1YRArM\_PriOrSale & FREdDIE MaC 1-YEaR ARm NORtH East\
freDdie\_RegIonal\_SE\_1yrArM\_PRiOrSale & FreddiE maC 1-Year ARM SoUth east\
Freddie\_REgiOnal\_SW\_1YraRM\_pRiorSAle & freddIe Mac 1-YEAr ARM SOuth West\
FReDDie\_RegIOnal\_W\_1yrarm\_PRiorsale & Freddie Mac 1-YeaR aRm west\
Freddie\_US\_15yR\_FRM\_PrIoRsaLE & freDdiE Mac 15-Year fiXeD mortGAge rate\
fReDDIE\_US\_15YR\_feespts\_Priorsale & Freddie mAc 15-YEar fixEd MorTGage raTe feeS aND poInts\
Freddie\_uS\_1yraRM\_Fees\_PrIorSalE & freddie mAc 1-Year ArM fees And PoiNts\
FREdDiE\_US\_1YrarM\_MARgIn\_PRIorsale & FredDiE MAc 1-YeaR ARM MARGIn\
FrEddIe\_US\_1YrARM\_priorSale & FredDie mac 1-YEAr ArM\
FreDdie\_Us\_30yr\_FrM\_priorsale & FrEddie maC 30-Year fixed mortgAge rAte\
FreddiE\_US\_30Yr\_FeeSpTs\_PriORSale & FRedDie mac 30-Year Fixed moRTgaGe RATE fEes and points\
\
N\_ShortsaLES & NUmber of sHort saLEs ThAT occurreD wIthIn thE SAme ziPcodE As The propeRty witHIn ThE past 12 moNtHs.\
N\_PrepaYmeNts & numbeR Of loAns thaT have beeN prepAId\
N\_Outstanding\_lOans & How many loANs ARE cURrenTly Outstanding\
n\_PIF & nUmbeR of lOAnS thAT were Paid iN fULl BY reaching their term mAtUrity\
N\_origiNations & Number Of loans oriGINAted at saMe loCAtIOn and same time pEriod Of sale\
MediAN\_Time\_to\_SaF & MedIan time bEtween the ORIginal foRecLosUre MonTH AnD the SAF month oF ALl SAf tHat occuRreD in the gIveN moNth.\
medIaN\_Time\_to\_NoNInst\_SAF & saMe As AbOve But foR Non-instiTuTioNaL / noN-corpORate owNer SAf.\
MedIaN\_TIMe\_tO\_SAF\_l12 & SaME aS ABove ExCePt thAt cOmPutatIon rOLls | 5/1 hybrid amortizing ARM fees andpoint s\Fre dd ie\_ Regi onal\_NC\_1yrA R M\_P riorSale & Freddie Mac 1-Ye ar ARMN or th Ce ntral\F re d d ie\ _R eg ion al \ _N E\_1y rAR M\_Prio rSale & Fr edd ie Mac 1-YearA RM North Eas t\Freddie\_Reg ion al\_SE \_ 1yr A RM\_P rio rSale & Fre d die Ma c 1-YearAR M South East\
F r e dd ie\_ Regional\_SW\_1yr A RM \ _PriorSale & F reddie M a c1 - Yea r A RM South W es t\
Fr e ddie\_R e gi o n a l\_ W \_1yrARM\_Pri orSale & Fr e ddi e Mac1- Yea r ARM W est\Fr e ddi e\_US\_15yr \_FR M\_PriorS ale &F reddieM ac 15-Y ear fi xed mo rtga g era te\
F r edd i e\ _US \ _15 yr\_fees pt s\ _Prio rSal e & Fred die Mac 15-Y ear fixed mor tga ge r a tefeesand p oint s\
Fred die\_U S\_1y rA RM\_Fees\_Prior Sale & Freddi e M ac 1- Ye ar AR M feesand po ints\
F reddie\ _ US\ _1 y r A RM \_Margin\_PriorSal e& Fr eddie Ma c 1-Ye a rAR M margin\
F red die\ _ U S\_1y rARM \ _P riorSale & Fre d di eMac 1-Y ea r ARM\
F red die \_US\ _ 30yr \_FRM\ _PriorSa le &F reddie Mac 30- Y ear fixed mor t ga g e r a te\Fre ddie\_US\_3 0yr\ _ fees pts\ _ Pr ior S ale & Fred di e M a c 30-Year fixed mor tg age ra te fe es and points \
\
N\_S h o r tSales & Num b er of short sales that occurredw ithin th e sam e zipcod e as thep r operty w ith inthe pa s t 1 2 months.\
N\ _ P rePa ym ents &Num ber ofloa nstha t h av e been pr epaid\
N \_ Ou ts ta ndi ng\_L o ans & Ho wman yloa ns ar e curre ntlyouts ta nd i ng\
N\_PIF &N u mber o floan s t ha t wer e pa i d i n fullby reachi ngt heir t er m matur ity\
N\_Origi na tions & Nu mb erof loa n s origina ted at same location an d same t ime peri od o f sale\
M edi an\_Ti me\ _ to\_SA F & Me dianti meb e tween t he or ig inal forec l o sur e mon th and the SA F month of all SAF tha t occurred in th e gi v e nmon t h. \
Me di a n\_ T i me\_to\_NonInst \_SAF & Sa me as above but for n on-inst itution al /n on-corp orate own er SAF.\Me dian \ _ Tim e\_to\_SAF \_l12 &Same as a b ove e x ce pt th atcomput at ion roll s | 5/1_hybrid amortizing_ARM fees and points\
Freddie\_Regional\_NC\_1yrARM\_PriorSale_& Freddie_Mac_1-Year ARM_North_Central\
Freddie\_Regional\_NE\_1yrARM\_PriorSale & Freddie_Mac 1-Year ARM_North East\
Freddie\_Regional\_SE\_1yrARM\_PriorSale & Freddie_Mac 1-Year ARM_South_East\
Freddie\_Regional\_SW\_1yrARM\_PriorSale & Freddie Mac 1-Year ARM South West\
Freddie\_Regional\_W\_1yrARM\_PriorSale & Freddie Mac 1-Year ARM West\
Freddie\_US\_15yr\_FRM\_PriorSale_&_Freddie Mac_15-Year_fixed_mortgage rate\
Freddie\_US\_15yr\_feespts\_PriorSale & Freddie Mac_15-Year fixed mortgage rate fees_and points\
Freddie\_US\_1yrARM\_Fees\_PriorSale_& Freddie Mac 1-Year ARM fees and points\
Freddie\_US\_1yrARM\_Margin\_PriorSale_&_Freddie Mac 1-Year_ARM margin\
Freddie\_US\_1yrARM\_PriorSale & Freddie Mac 1-Year ARM\
Freddie\_US\_30yr\_FRM\_PriorSale & Freddie_Mac 30-Year fixed mortgage rate\
Freddie\_US\_30yr\_feespts\_PriorSale &_Freddie Mac 30-Year_fixed_mortgage_rate fees and points\
\
N\_ShortSales_& Number of short sales that_occurred within the same zipcode as_the property within the past 12 months.\
N\_PrePayments_& Number of loans that have_been prepaid\
N\_Outstanding\_Loans & How many_loans are_currently outstanding\
N\_PIF & Number of_loans that were_paid in_full by reaching_their term maturity\
N\_Originations & Number of_loans originated at_same location and same time period_of_sale\
Median\_Time\_to\_SAF & Median_time_between_the original_foreclosure month and_the_SAF month_of_all SAF that occurred in the_given_month.\
Median\_Time\_to\_NonInst\_SAF & Same as above but for_non-institutional / non-corporate owner_SAF.\
Median\_Time\_to\_SAF\_l12_& Same as above_except that computation rolls |
be seen as asymmetric residuals, so that part of the galaxy has positive residual values and the other part has negative values. In such cases we fixed all the other fit parameters except for the galaxy center, and let GALFIT to refit the center of the Sérsic profile. After this the center was fixed again and the other Sérsic parameters were fit again. This procedure reduced the fit residuals significantly.
Some of the galaxies show also a peaked nucleus (see the 4th row in Fig. \[fit\_profiles\]) which is clearly a separate component in the galaxy center. In such cases, we manually place an additional *PSF*-component to fit the nucleus. The [*PSF*]{}-component is added to improve the overall fit of the galaxy rather than aiming for the detailed analysis of the nuclear star clusters. We fit the [*PSF*]{}-component using a fixed center position, and leave the total magnitude of the nucleus as a free parameter to fit.
Using the steps described before, our fits converged for all the galaxies in our sample. In the end, we examined the fits for systematics in the residuals, and corrected the ones where the center of the galaxy was wrong or the nucleus was missed. To ensure that the sky components are fit correctly, the radial surface brightness profiles and the cumulative radial profiles were inspected. Specifically we checked that GALFIT does not mix the sky component with the Sérsic profile, which would give unrealistically large effective radii and total magnitudes to the galaxies. The opposite (absorbing galaxy light to sky) is unlikely, since all our images had large areas of background sky not covered by the galaxy.
Comparison of the 1D and 2D methods
-----------------------------------
Fig. \[fig:fitcomparison\] shows a comparison of the Sérsic $n$-values, axis ratios ($b/a$), total r’-band magnitudes (m$_{r'}$), and effective radii (R$_{e}$) obtained with the two methods. Both methods give similar distributions for the Sérsic $n$ values, although the scatter is significant with RMS = 0.47. The effective radii and magnitudes are well in agreement in the two measurements (the lower panels of \[fig:fitcomparison\]), except for the four outliers. The outliers are explained due to the different fitting of their background levels in the 1D and 2D methods. Only for the axis ratios $b/a$ a systematic shift appears between the two methods | be seen as asymmetric residuals, so that part of the galax have positive residual values and the early part have negative value. In such case we fixed all the early fit argument except for the galaxy plaza, and let GALFIT to refit the center of the Sérsic visibility. After this the center was fixed again and the early Sérsic parameters were fit again. This operation reduce the fit residuals significantly.
Some of the galaxy show also a peaked core (see the 4th row in Fig. \[fit\_profiles\ ]) which is clearly a separate part in the galaxy center. In such cases, we manually place an additional * PSF*-component to fit the core. The [ * PSF*]{}-component is added to improve the overall fit of the galaxy rather than aiming for the detailed analysis of the nuclear star clusters. We fit the [ * PSF*]{}-component use a fixed center stead, and leave the total magnitude of the nucleus as a spare parameter to fit.
Using the steps trace before, our fits converged for all the galaxies in our sample. In the end, we examined the paroxysm for systematics in the residuals, and corrected the ones where the center of the galax was wrong or the lens nucleus was miss. To ensure that the sky components are fit correctly, the radial surface luminosity profiles and the cumulative radial profiles were inspected. Specifically we checked that GALFIT does not mix the sky component with the Sérsic profile, which would give unrealistically big effective radii and full magnitudes to the galaxies. The diametric (absorbing galaxy light to sky) is unlikely, since all our effigy had big areas of background sky not cover by the galaxy.
Comparison of the 1D and 2D method acting
-----------------------------------
Fig. \[fig: fitcomparison\ ] shows a comparison of the Sérsic $ n$-values, axis ratios ($ b / a$), total r’-band magnitudes (m$_{r'}$), and effective radii (R$_{e}$) obtained with the two methods. Both methods yield similar distributions for the Sérsic $ n$ value, although the scatter is significant with RMS = 0.47. The effective radii and magnitudes are well in agreement in the two measurements (the lower panels of \[fig: fitcomparison\ ]), except for the four outlier. The outliers are explained due to the different fitting of their background levels in the 1D and 2D method. merely for the bloc proportion $ b / a$ a systematic shift appears between the two methods | be seen as asymmetric resiauals, so that pcet of vhe galzxy has oositive residual values and tye otyer part has negative xalues. In such cawes xe fixed all the other flc parzletexs except for thg galaxy cender, and let GANFKT to refit the center of the Sérsic pwofile. Sfher this the cgnter ras rpxtd again and the other Sérsic parajeters xere fit again. Yhis procedure reduced the fit residuals signififantly.
Some if trw galaxies sfow also a peaked nuclgus (see the 4th row in Fig. \[fit\_prowiles\]) which is xlwarpi a separate compjnent in the galaxy wenter. On such cases, ee kanyally place an additimnal *PSF*-component jo fit the nbcleus. The [*PSF*]{}-componebt is agded to knpruve tie kveralp fmt of the gzlaxy rathee than aiming for tne eetailed analgsis os ehe nuclear star clusters. We fit the [*PSX*]{}-cojponent using a fixed cwnter position, and lewve the tjtal magnitude of the nucleus as a free parameter to fmt.
Jsiun gye steps described before, our fits converged fow akl the galaxies ln our sample. In tne emq, we examined the fnfs for systematics ij the rgsiduaos, and cowrecyed the ones where the centwr of the gajqxy was wrong or tke nucleus wcs misxed. Tp ensure that the sky cumpohents are flt correcfuy, the radial suffabe bsightness profiles and the cumulatite rabial prowilex were inspected. Speclxically we checked that GDLFIT does not mix the sky component with the Sérsic progine, fhich wobld giye unrealisticajly large effeetive rabii ana total mafnitudev to the gajaxies. The op[lsite (absorbmng galaxr litht ro sky) kr unlikely, sinve all oug nmages hae large areas of beckgtohnd sky not covzxee by the galaxy.
Vomoarysln os the 1D and 2D metfodr
-----------------------------------
Gig. \[fkg:fitcomparlsov\] shpws a comparison of dhe Aérsic $n$-values, axis rwtios ($b/a$), total r’-fand magnitudrs (m$_{r'}$), and effectivt radim (R$_{e}$) outainec wyth the two methods. Both methoss give slmijar distribueionw for the Sérxic $n$ values, although the scatter is sijnificant with RMS = 0.47. Thw effective radii aud magnitudes are rell in acreement in the two neasurements (the kower panels of \[fig:fifcompasison\]), except for the four outliers. The outliers are explained due to the diffeeent fmteing of thejr bsckgrmuud kevels in vhx 1D and 2D methods. Only for the axis ratios $b/a$ a rystemathc shift appears between the teo methods | be seen as asymmetric residuals, so that the has positive values and the In cases we fixed the other fit except for the galaxy center, and GALFIT to refit the center of the Sérsic profile. After this the center fixed again and the other Sérsic parameters were fit again. This procedure reduced fit significantly. of galaxies show also a peaked nucleus (see the 4th row in Fig. \[fit\_profiles\]) which is clearly separate component in the galaxy center. In such we manually place an *PSF*-component to fit the nucleus. [*PSF*]{}-component added to the fit the galaxy rather aiming for the detailed analysis of the nuclear star clusters. We fit the [*PSF*]{}-component using a fixed position, and total magnitude the as free parameter to the steps described before, our fits the galaxies in our sample. In the end, examined the for systematics in the residuals, and the ones where the center of the galaxy wrong or the nucleus was missed. To ensure that the sky components are fit correctly, surface brightness profiles and cumulative radial profiles inspected. we that does not the sky component with the Sérsic profile, which would give unrealistically effective radii and total magnitudes to the galaxies. The opposite light sky) is unlikely, all our images had areas background sky not covered galaxy. of 2D ----------------------------------- \[fig:fitcomparison\] shows a comparison the Sérsic $n$-values, axis ratios total r’-band magnitudes (m$_{r'}$), with the two methods. Both methods give similar for the Sérsic $n$ values, although the is significant with RMS = 0.47. The effective radii and magnitudes are in agreement two measurements (the lower panels of \[fig:fitcomparison\]), except the four outliers. The are explained due to the different fitting of their levels the 1D 2D methods. Only the axis ratios a systematic shift the two | be seen as asymmetric residuaLs, so that paRt of tHe gAlaXy Has pOsitIve residual valUEs anD the other part has negatiVe valUeS. in suCH cAses wE fixed aLL tHE OthEr FiT paRaMEtErs exCepT for the Galaxy centEr, aNd Let GALFIT to rEFiT the center Of tHe Sérsic profIle. after tHiS thE CenteR waS fixeD again ANd the oTher SérsiC pARameteRS were fiT AGaIn. ThIs procedure reduceD ThE Fit residuals siGnificAnTLy.
sOMe oF thE galaxies sHoW also A Peaked nUClEUS (See THe 4th row in Fig. \[fIt\_profiles\]) wHIch Is cleaRlY a sEParate CompoNeNT in The galaxy ceNter. in such casEs, we maNUally plACe an addItionaL *PSf*-coMponENt To Fit ThE NucLEuS. ThE [*pSF*]{}-ComponenT iS aDded tO impROVE The oVerAll fIt of tHe galaxy ratheR thAn aiMIng For thE detaIled AnAlysiS of the NucleAr Star clusters. We fIt thE [*PSF*]{}-compoNenT uSinG a Fixed CEnter pOsiTioN, and leaVe the toTAl mAgNITUdE of the nucleus as a frEe PARaMeter to fIt.
UsinG ThE sTEps descrIbEd bEforE, OUr fitS conVErGed for alL the gaLAxIeS in our sAmPle. In tHe End, We eXaminED the Fits foR systemaTics iN The residuals, anD Corrected the oNEs WHErE The cEntEr of the galaXy waS WronG or tHE nUclEUs was MisseD. TO EnSUre that the sky componEnTs are fIt corRectly, the radiAl surface bRIGHtness prOfilES aND the cumulative RadiaL profiles wERe inspecTed. SpEcificalLy we checkED That GALFiT dOes Not Mix THE sKy component wiTH The SÉrSic profIle, Which woUld GivE unReaLiStically lArge effeCtIvE rAdIi aNd totAL magnituDeS to ThE gaLaxieS. the oppOsite (AbsoRbInG GalAxy lighT To SKY) is uNlIkEly, sIncE aLl our ImagES haD large aReas of bacKgrOUnd sKy NoT covereD by the galaxy.
COmParison of tHe 1d anD 2D methODS
-----------------------------------
Fig. \[fig:fItcomparison\] shows a compaRIson of tHe SÉrsic $N$-valUes, axis raTioS ($b/a$), totAl r’-BAnd magNitudeS (m$_{r'}$), anD eFfeCTIve raDII (R$_{E}$) obTaIned with thE TWo mEthodS. BOth mEthods gIve similar distribuTIonS for the Sérsic $N$ vaLues, ALThOugH ThE ScaTtER is SIGnificant with RMs = 0.47. The effectIvE RaDii and magnITudEs Are well In agreeMent iN The two mEasuremenTs (the loweR pAnelS OF \[fiG:fitcomparIson\]), excePt for the fOUr outLIeRs. The OutLiers aRe ExpLaineD due to THe dIfferEnt fitTiNg of thEir baCkGround leVels in the 1D and 2D methods. OnLy for tHe axiS raTios $b/a$ a sySteMAtiC shift appEars Between the Two MetHods | be seen as asymmetric res iduals, so that pa rtof the gal axy has positi v e re sidual values and theother p a rt h a snegat ive val u es . Insu ch ca se s w e fix edall the other fit pa ra meters excep t f or the gal axy center, and le t GALF IT to refit th e cen ter of the Sé rsic prof il e . Afte r this t h e c ente r was fixed again an d the other Sér sic pa ra m et e r s w ere fit again .Thisp rocedur e r e d u ced the fit resid uals signif i can tly.
So meo f thegalax ie s sh ow also a p eake d nucleus (seet he 4thr ow in F ig. \[ fit \_p rofi l es \] ) w hi c h i s c lea r lya separa te c ompon enti n t he g ala xy c enter . In such cas es, wem anu allyplace anad ditio nal *P SF*-c om ponent to fit t he n ucleus. T he[* PSF *] {}-co m ponent is ad ded toimprove the o v e r al l fit of the galax yr a th er thanaiming fo rt he detai le d a naly s i s ofthen uc lear sta r clus t er s. We fit t he [*P SF *]{ }-c ompon e nt u sing a fixed c enter position, andl eave the tota l m a g ni t udeofthe nucleus asa fre e pa r am ete r to f it.
Us i ng the steps described b efore, ourfits converge d for allt h e galaxie s in ou r sample. In th e end , we exami n ed the f its f or syste matics in t he resid ual s,and co r r ec ted the onesw h ereth e cente r o f the g ala xywas wr on g or thenucleuswa smi ss ed. To e n sure tha tthe s kycompo n ents a re fi t co rr ec t ly, the ra d ia l surf ac ebrig htn es s pro file s an d the c umulative ra d ialpr of iles we re inspected. S pecificall ywechecke d that GAL FIT does not mix the sk y compon ent with the Sérsic p rof ile, w hic h would giveunrea li sti c a lly l a r ge ef fe ctive radi i and tota lmagn itudesto the galaxies. T h e o pposite (abso rbi ng g a l ax y l i gh t to s k y)i s unlikely, sinc e all ourim a ge s had larg e ar ea s of ba ckgroun d sky not cov ered by t he galaxy .
Com p a ris on of the1D and 2 D methods ----- - -- ----- --- ------ -- --- ----- ---
F i g.\[fig :fitco mp arison \] sh ow s a comp arison of the Sérsic $n $-valu es, a xis ratios ( $b/ a $), total r’ -ban d magnitud es(m$ _{r'} $), and e ffec t iv e r a dii ( R$_{ e }$) obtai n ed wi t h t he two meth o d s . B oth m eth o ds giv e si milar distributio n s for the Sérs ic $ n $ va lue s , al th ough the scatt eris s ignifica nt with RMS = 0.47. T he effec tive r adii a nd magn i t ud e s arewell in agreemen t i nt he twome as u rement s (t he lower panel s of\ [ fig:fitcompariso n\]), e xcept for thefo ur outl i ers. The outli ers are exp lained due to t he diff er ent fi tti ng of theirb ackground leve ls in t he 1Dand 2D me thod s . Only for t heaxis rati o s $ b /a $a sy stem aticsh iftappears b e tween th e t w o metho ds | be_seen as_asymmetric residuals, so that_part of_the_galaxy has_positive_residual values and_the other part_has negative values. In_such cases we_fixed_all the other fit parameters except for the galaxy center, and let GALFIT to_refit_the center_of_the_Sérsic profile. After this the_center was fixed again and_the other_Sérsic parameters were fit again. This procedure reduced_the_fit residuals significantly.
Some_of the galaxies show also a peaked nucleus (see_the 4th row in Fig. \[fit\_profiles\])_which is clearly_a_separate_component in the galaxy_center. In such cases, we manually_place an additional *PSF*-component to fit_the nucleus. The [*PSF*]{}-component is added to_improve the overall fit of the_galaxy rather than aiming for_the detailed_analysis of the nuclear star_clusters. We fit_the [*PSF*]{}-component_using a fixed_center position, and leave the total_magnitude of the_nucleus as a free parameter to_fit.
Using_the steps described_before,_our_fits converged_for all the_galaxies_in our_sample._In the end, we examined the_fits_for systematics in the residuals, and corrected_the ones where the_center_of the galaxy was_wrong or the nucleus was_missed. To ensure that the sky_components are_fit correctly,_the radial surface brightness profiles and the cumulative radial profiles were_inspected. Specifically we checked that GALFIT_does not mix the_sky component_with_the Sérsic profile,_which_would give_unrealistically large effective radii and total magnitudes_to the_galaxies. The opposite (absorbing galaxy light_to sky) is unlikely,_since_all our images had large areas_of background sky not covered by_the galaxy.
Comparison of the 1D_and_2D_methods
-----------------------------------
Fig. \[fig:fitcomparison\] shows a comparison_of the Sérsic $n$-values, axis ratios_($b/a$), total r’-band_magnitudes (m$_{r'}$), and effective radii (R$_{e}$) obtained_with_the two methods. Both methods give_similar_distributions for the Sérsic $n$ values,_although_the_scatter is significant with RMS_= 0.47. The effective radii and_magnitudes are well in agreement in the two measurements_(the lower panels_of \[fig:fitcomparison\]), except for the_four_outliers._The outliers are explained due to the different fitting of_their background_levels in the_1D and 2D methods. Only for the axis ratios $b/a$_a systematic shift appears between the two_methods |
$H=0.8$ (surface 3, blue line). Note that we include one decade roll-off region, from $q=10^3 \ {\rm m}^{-1}$ to $q=10^4 \ {\rm m}^{-1}$. We assume the effective Young’s modulus $E^*=1.33 \ {\rm GPa}$ and the radius of the cylinder $R=1 \ {\rm cm}$. We vary the applied pressure $f_{\rm N}$ from $10^3 \ {\rm N/m}$ to $1.5\times 10^6 \ {\rm N/m}$. This gives a variation in $\alpha = h_{\rm rms}/\delta$ with more than 3 decades.
Fig. \[1x.2p.Hertz.and.Roughness.eps\] shows the contact pressure $p(x)$ as a function of the coordinate $x$ for the rigid cylinder squeezed against an elastic half space with the force $f_{\rm N} = 10 \ {\rm kN/m}$. The dashed line is for perfectly smooth surfaces (Hertz contact), and the solid line for the substrate surface 1 with the surface roughness power spectrum given in Fig. \[1logq.2logC.eps\] (red line). Note that the surface roughness reduces the maximal contact pressure and makes the pressure profile Gaussian-like, but it becomes a perfect Gaussian only for much larger values of the parameter $\alpha$ (corresponding to smaller applied force $f_{\rm N}$ or larger rms-roughness amplitude $h_{\rm rms}$).
Fig. \[1logAlpha.2ratioArea.eps\] shows the ratio $w/s$ between the full-width-at-half-maximum (FWHM) and the standard deviation $s$ as a function of the logarithm of the dimensionless number $\alpha = h_{\rm rms}/\delta = (\pi/4) (h_{\rm rms} E^*/ f_{\rm N})$. In the calculation we have used the three power spectra shown in Fig. \[1logq.2logC.eps\] and varied the loading force $f_{\rm N}$ (between $10^3 \ {\rm N/m}$ and $1.5\times 10^6 \ {\rm N/m}$). Note that as a function of $\alpha$ the ratio $w/s$ is nearly the same for all three surfaces. This suggest that | $ H=0.8 $ (surface 3, blue line). Note that we admit one ten scroll - off region, from $ q=10 ^ 3 \ { \rm m}^{-1}$ to $ q=10 ^ 4 \ { \rm m}^{-1}$. We wear the effective Young ’s modulus $ E^*=1.33 \ { \rm GPa}$ and the spoke of the cylinder $ R=1 \ { \rm cm}$. We vary the applied pressure $ f_{\rm N}$ from $ 10 ^ 3 \ { \rm N / m}$ to $ 1.5\times 10 ^ 6 \ { \rm N / m}$. This give a variation in $ \alpha = h_{\rm rms}/\delta$ with more than 3 decades.
Fig. \[1x.2p. Hertz.and. Roughness.eps\ ] read the contact pressure $ p(x)$ as a routine of the coordinate $ x$ for the rigid cylinder squeezed against an elastic one-half space with the force $ f_{\rm N } = 10 \ { \rm kN / m}$. The dashed line is for perfectly smooth surfaces (Hertz liaison), and the solid line for the substrate surface 1 with the airfoil roughness power spectrum give in Fig. \[1logq.2logC.eps\ ] (crimson line). notice that the surface roughness reduces the maximal contact imperativeness and makes the pressure profile Gaussian - like, but it becomes a perfect Gaussian only for a lot bigger values of the parameter $ \alpha$ (corresponding to smaller applied force $ f_{\rm N}$ or larger rms - roughness amplitude $ h_{\rm rms}$).
Fig. \[1logAlpha.2ratioArea.eps\ ] shows the proportion $ tungsten / s$ between the full - width - at - one-half - maximal (FWHM) and the standard deviation $ s$ as a function of the logarithm of the dimensionless number $ \alpha = h_{\rm rms}/\delta = (\pi/4) (h_{\rm rms } E^*/ f_{\rm N})$. In the calculation we have used the three office spectrum shown in Fig. \[1logq.2logC.eps\ ] and varied the cargo force $ f_{\rm N}$ (between $ 10 ^ 3 \ { \rm N / m}$ and $ 1.5\times 10 ^ 6 \ { \rm N / m}$). Note that as a function of $ \alpha$ the ratio $ w / s$ is nearly the like for all three surfaces. This suggest that | $H=0.8$ (durface 3, blue line). Note uhat we include ouw decave roll-kff regiun, from $q=10^3 \ {\rm m}^{-1}$ to $q=10^4 \ {\rm m}^{-1}$. Wx aswume uke effective Young’s oodulus $E^*=1.33 \ {\rm GPa}$ and rhe radius of the gvlindsv $R=1 \ {\xm cm}$. We vary thg applied prassure $f_{\rm N}$ fsoo $10^3 \ {\rm N/m}$ to $1.5\times 10^6 \ {\rm N/m}$. This gives a varistlon in $\alpha = r_{\rm gmf}/\delfa$ with more than 3 decades.
Fig. \[1x.2p.Hedtz.and.Rmughness.eps\] snows the contact pressure $o(x)$ ad a function of thf coordinatg $x$ sir the rigid cylinder squeezed agajnst an elastic half space with the yorce $f_{\rm N} = 10 \ {\rl kN/m}$. The dawhed jine is for ivrfectlf smootn surfaces (Herbz coitacr), and the solid line hor the substrate sutface 1 witv che surface roughness piwer vpecdrum tivdn jn Fjg. \[1loge.2lojC.eps\] (red ljne). Note thqt the surface rougnnqws reduces ths maxiiaj contact pressure and makes the pressuge pdofile Gaussian-like, but it becomes a perfect Gaussian only for much larger values of the parameter $\alpva$ (cocrdspinqkbg to smaller applied force $f_{\rm N}$ or larger rms-wkubhkess amplitude $h_{\vm rms}$).
Fig. \[1logAlpha.2tahipWrea.eps\] shows the rcfik $w/s$ between the fkll-widtr-at-haof-maximum (FWHK) and the standard deviatiob $s$ as a funbtiob of the logarithm of the dimznsionkess mumber $\alpha = h_{\rm rms}/\deuta = (\pi/4) (h_{\rm rms} E^*/ f_{\rm N})$. Jv the calculatiov wv haee used ufe three power spqctra shoxn in Fig. \[1loeq.2lobC.eps\] wnd varied the loading force $f_{\rm N}$ (betwgen $10^3 \ {\sm N/m}$ and $1.5\himes 10^6 \ {\rm N/m}$). Note that as a fuirtion of $\alphs$ dhe ratio $w/f$ is kearly the same for all three surfaczs. Thir suggest nhat | $H=0.8$ (surface 3, blue line). Note that one roll-off region, $q=10^3 \ {\rm m}^{-1}$. assume the effective modulus $E^*=1.33 \ GPa}$ and the radius of the $R=1 \ {\rm cm}$. We vary the applied pressure $f_{\rm N}$ from $10^3 {\rm N/m}$ to $1.5\times 10^6 \ {\rm N/m}$. This gives a variation in = rms}/\delta$ more 3 decades. Fig. \[1x.2p.Hertz.and.Roughness.eps\] shows the contact pressure $p(x)$ as a function of the coordinate $x$ the rigid cylinder squeezed against an elastic half with the force $f_{\rm = 10 \ {\rm kN/m}$. dashed is for smooth (Hertz and the solid for the substrate surface 1 with the surface roughness power spectrum given in Fig. \[1logq.2logC.eps\] (red line). that the reduces the contact and the pressure profile it becomes a perfect Gaussian only values of the parameter $\alpha$ (corresponding to smaller force $f_{\rm or larger rms-roughness amplitude $h_{\rm rms}$). \[1logAlpha.2ratioArea.eps\] shows the ratio $w/s$ between the full-width-at-half-maximum and the standard deviation $s$ as a function of the logarithm of the dimensionless number h_{\rm rms}/\delta = (\pi/4) rms} E^*/ f_{\rm In calculation have the three spectra shown in Fig. \[1logq.2logC.eps\] and varied the loading force $f_{\rm (between $10^3 \ {\rm N/m}$ and $1.5\times 10^6 \ {\rm that a function of the ratio $w/s$ is the for all three surfaces. that | $H=0.8$ (surface 3, blue line). Note that wE include onE decaDe rOll-OfF regIon, fRom $q=10^3 \ {\rm m}^{-1}$ to $q=10^4 \ {\rm m}^{-1}$. WE AssuMe the effective Young’s moDulus $e^*=1.33 \ {\rM gPa}$ aND tHe radIus of thE CyLINdeR $R=1 \ {\Rm Cm}$. WE vARy The apPliEd pressUre $f_{\rm N}$ froM $10^3 \ {\rm n/m}$ To $1.5\times 10^6 \ {\rm N/m}$. THIs Gives a variAtiOn in $\alpha = h_{\rm Rms}/\Delta$ wItH moRE than 3 DecAdes.
FIg. \[1x.2p.HeRTz.and.ROughness.ePs\] SHows thE Contact PREsSure $P(x)$ as a function of thE CoORdinate $x$ for the Rigid cYlINdER SquEezEd against aN eLastiC Half spaCE wITH The FOrce $f_{\rm N} = 10 \ {\rm kN/m}$. the dashed liNE is For perFeCtlY Smooth SurfaCeS (herTz contact), anD the Solid line For the SUbstratE Surface 1 With thE suRfaCe roUGhNeSs pOwER spECtRum GIveN in Fig. \[1loGq.2LoGC.eps\] (Red lINE). nOte tHat The sUrfacE roughness redUceS the MAxiMal coNtact PresSuRe and Makes tHe preSsUre profile GaussIan-lIke, but it bEcoMeS a pErFect GAUssian OnlY foR much laRger valUEs oF tHE PArAmeter $\alpha$ (correspOnDINg To smalleR appliED fOrCE $f_{\rm N}$ or lArGer Rms-rOUGhnesS ampLItUde $h_{\rm rmS}$).
Fig. \[1loGalPhA.2ratioAReA.eps\] shOwS thE raTio $w/s$ BEtweEn the fUll-width-At-halF-Maximum (FWHM) and THe standard devIAtION $s$ AS a fuNctIon of the logAritHM of tHe diMEnSioNLess nUmber $\AlPHa = H_{\Rm rms}/\delta = (\pi/4) (h_{\rm rms} E^*/ F_{\rM N})$. In thE calcUlation we have Used the thrEE POwer specTra sHOwN In Fig. \[1logq.2logC.ePs\] and Varied the lOAding forCe $f_{\rm n}$ (between $10^3 \ {\Rm N/m}$ and $1.5\tiMES 10^6 \ {\rm N/m}$). NotE thAt aS a fUncTIOn Of $\alpha$ the ratIO $W/s$ is NeArly the SamE for all ThrEe sUrfAceS. THis suggesT that | $H=0.8$ (surface 3, blueline). Not e tha t w e i nc lude one decade roll-o f f re gion, from $q=10^3 \ { \rm m }^ { -1}$ to $q=1 0^4 \ { \ rm m }^{ -1 }$ . W ea ss ume t heeffecti ve Young’s mo du lus $E^*=1.3 3 \ {\rm GPa} $ a nd the radiu s o f thecy lin d er $R =1\ {\r m cm}$ . We va ry the ap pl i ed pre s sure $f _ { \r m N} $ from $10^3 \ {\ r mN /m}$ to $1.5\t imes 1 0^ 6 \ { \rm N/ m}$. Thisgi ves a variati o ni n $\a l pha = h_{\rmrms}/\delta $ wi th mor etha n 3 dec ades.
F ig. \[1x.2p.He rtz. and.Rough ness.e p s\] sho w s the c ontact pr ess ure$ p( x) $ a sa fu n ct ion ofthe coor di na te $x $ fo r t h e ri gid cyl inder squeezed aga ins t an ela stichalfspac ewiththe fo rce $ f_ {\rm N} = 10 \{\rm kN/m}$.The d ash ed line is for pe rfe ctly sm ooth su r fac es ( H er tz contact), and t he s ol id linefor th e s ub s trate su rf ace 1 w i t h the sur f ac e roughn ess po w er s pectrum g iven i nFig . \ [1log q .2lo gC.eps \] (redline) . Note that the surface rough n es s re d uces th e maximal c onta c t pr essu r eand makes thepr e ss u re profile Gaussian -l ike, b ut it becomes a pe rfect Gaus s i a n only f or m u ch larger valuesof th e paramete r $\alpha $ (co rrespond ing to sm a l ler appl ied fo rce $f _ { \r m N}$ or larg e r rms -r oughnes s a mplitud e $ h_{ \rm rm s} $).
Fig. \[1logA lp ha .2 ra tio Area. e ps\] sho ws th erat io $w / s$ bet weenthefu ll - wid th-at-h a lf - m axim um ( FWHM ) a nd thestan d ard deviat ion $s$ a s a func ti on of the logarithm of t he dimensi on les s numb e r $\alpha = h_{\rm rms}/\delta = (\pi/4) (h _{\rm rms } E^*/ f_ {\r m N})$ . I n the c alcula tionwe ha v e used t he th re e power sp e c tra show nin F ig. \[1 logq.2logC.eps\] a n d v aried the loa din g fo r c e$f_ { \r m N} $( bet w e en $10^3 \ {\rm N/m}$ and $ 1 .5 \times 10^ 6 \{\ rm N/m} $). Not e tha t as a f unction o f $\alpha $ther a tio $w/s$ isnearly t he same f o r all th ree s urf aces.Th issugge st tha t | $H=0.8$_(surface 3,_blue line). Note that_we include_one_decade roll-off_region,_from $q=10^3 \_{\rm m}^{-1}$ to_$q=10^4 \ {\rm m}^{-1}$._We assume the_effective_Young’s modulus $E^*=1.33 \ {\rm GPa}$ and the radius of the cylinder $R=1 \_{\rm_cm}$. We_vary_the_applied pressure $f_{\rm N}$ from_$10^3 \ {\rm N/m}$ to_$1.5\times 10^6_\ {\rm N/m}$. This gives a variation in_$\alpha_= h_{\rm rms}/\delta$_with more than 3 decades.
Fig. \[1x.2p.Hertz.and.Roughness.eps\] shows the contact_pressure $p(x)$ as a function of_the coordinate $x$_for_the_rigid cylinder squeezed against_an elastic half space with the_force $f_{\rm N} = 10 \_{\rm kN/m}$. The dashed line is for_perfectly smooth surfaces (Hertz contact), and_the solid line for the_substrate surface_1 with the surface roughness_power spectrum given_in Fig._\[1logq.2logC.eps\] (red line)._Note that the surface roughness reduces_the maximal contact_pressure and makes the pressure profile_Gaussian-like,_but it becomes_a_perfect_Gaussian only_for much larger_values_of the_parameter_$\alpha$ (corresponding to smaller applied force_$f_{\rm_N}$ or larger rms-roughness amplitude $h_{\rm rms}$).
Fig._\[1logAlpha.2ratioArea.eps\] shows the ratio_$w/s$_between the full-width-at-half-maximum (FWHM)_and the standard deviation $s$_as a function of the logarithm_of the_dimensionless number_$\alpha = h_{\rm rms}/\delta = (\pi/4) (h_{\rm rms} E^*/ f_{\rm N})$._In the calculation we have used_the three power spectra_shown in_Fig._\[1logq.2logC.eps\] and varied_the_loading force_$f_{\rm N}$ (between $10^3 \ {\rm N/m}$_and $1.5\times_10^6 \ {\rm N/m}$). Note that_as a function of_$\alpha$_the ratio $w/s$ is nearly the_same for all three surfaces. This_suggest that |
\pi i)\right)
{\rm sh}\left(\frac{1}{2\xi}(\beta-3\pi i)\right)
},\nonumber\\
&&
\overline{R}(\beta)_{0,0}^{1,-1}=
\overline{R}(\beta)_{-1,1}^{0,0}=\frac{
i e^{\frac{\beta-2\pi i}{2\xi}}
{\rm sh}\left(\frac{\pi i}{\xi}\right)
{\rm sh}\left(\frac{\beta}{2\xi}\right)
}{
{\rm sh}\left(\frac{1}{2\xi}(\beta-2\pi i)\right)
{\rm sh}\left(\frac{1}{2\xi}(\beta-3\pi i)\right)
},
\nonumber
\\
&&\overline{R}(\beta)_{0,0}^{-1,1}=
\overline{R}(\beta)_{1,-1}^{0,0}=
\frac{-i e^{\frac{-\beta+2\pi i}{2\xi}}
{\rm sh}\left(\frac{\pi i}{\xi}\right)
{\rm sh}\left(\frac{\beta}{2\xi}\right)
}{
{\rm sh}\left(\frac{1}{2\xi}(\beta-2\pi i)\right)
{\rm sh}\left(\frac{1}{2\xi}(\beta-3\pi i)\right)
},\nonumber\\
&&\overline{R}(\beta)_{0,0}^{0,0}=-
\frac{{\rm sh}\left(\frac{\beta}{2\xi}\right)}{{\rm sh}
\left(\frac{1}{2\xi}(\beta-2\pi i)\right)}
-
\frac{
{\rm sh}\left(\frac{\pi i}{\xi}\right)
{\rm sh}\left(\frac{3\pi i}{2\xi}\right)
}{
{\rm sh}\left(\frac{1}{2\xi}(\beta-2\pi i)\right)
{\rm sh}\left(\frac{1}{2\xi}(\beta-3\pi i)\right)
}.\end{aligned}$$ Our $R$-matrix $R(\beta)$ gives the Boltzmann weight of the Izergin-Korepin model [@IK]. The $R$-matrix $R(\beta | \pi i)\right)
{ \rm sh}\left(\frac{1}{2\xi}(\beta-3\pi i)\right)
}, \nonumber\\
& &
\overline{R}(\beta)_{0,0}^{1,-1}=
\overline{R}(\beta)_{-1,1}^{0,0}=\frac {
i e^{\frac{\beta-2\pi i}{2\xi } }
{ \rm sh}\left(\frac{\pi i}{\xi}\right)
{ \rm sh}\left(\frac{\beta}{2\xi}\right)
} {
{ \rm sh}\left(\frac{1}{2\xi}(\beta-2\pi i)\right)
{ \rm sh}\left(\frac{1}{2\xi}(\beta-3\pi i)\right)
},
\nonumber
\\
& & \overline{R}(\beta)_{0,0}^{-1,1}=
\overline{R}(\beta)_{1,-1}^{0,0}=
\frac{-i e^{\frac{-\beta+2\pi i}{2\xi } }
{ \rm sh}\left(\frac{\pi i}{\xi}\right)
{ \rm sh}\left(\frac{\beta}{2\xi}\right)
} {
{ \rm sh}\left(\frac{1}{2\xi}(\beta-2\pi i)\right)
{ \rm sh}\left(\frac{1}{2\xi}(\beta-3\pi i)\right)
}, \nonumber\\
& & \overline{R}(\beta)_{0,0}^{0,0}=-
\frac{{\rm sh}\left(\frac{\beta}{2\xi}\right)}{{\rm sh }
\left(\frac{1}{2\xi}(\beta-2\pi i)\right) }
-
\frac {
{ \rm sh}\left(\frac{\pi i}{\xi}\right)
{ \rm sh}\left(\frac{3\pi i}{2\xi}\right)
} {
{ \rm sh}\left(\frac{1}{2\xi}(\beta-2\pi i)\right)
{ \rm sh}\left(\frac{1}{2\xi}(\beta-3\pi i)\right)
} .\end{aligned}$$ Our $ R$-matrix $ R(\beta)$ gives the Boltzmann weight of the Izergin - Korepin model [ @IK ]. The $ R$-matrix $ R(\beta | \pi l)\right)
{\rm sh}\left(\frac{1}{2\xi}(\betx-3\pi i)\right)
},\nonumywr\\
&&
\overnine{R}(\bsta)_{0,0}^{1,-1}=
\overlkne{R}(\beta)_{-1,1}^{0,0}=\frac{
i e^{\frac{\beta-2\pi i}{2\xm}}
{\rm wh}\lefu(\yrac{\pi i}{\xi}\right)
{\rm sh}\ueft(\frac{\bvta}{2\xi}\righr)
}{
{\rm wy}\left(\frac{1}{2\xm}(\geta-2\pi l)\xight)
{\dl sh}\nxft(\frac{1}{2\xi}(\beta-3\pi i)\right)
},
\nongmber
\\
&&\overline{R}(\tega)_{0,0}^{-1,1}=
\lverline{R}(\beta)_{1,-1}^{0,0}=
\frac{-i e^{\frac{-\beta+2\pi i}{2\xi}}
{\ri sh}\lefy(\fgac{\pi i}{\xi}\right)
{\tm sh}\kqft(\fdac{\beta}{2\xi}\right)
}{
{\rm sh}\left(\frac{1}{2\xi}(\beta-2\pj i)\righu)
{\rm sh}\left(\frac{1}{2\xi}(\beya-3\pi i)\right)
},\nonumber\\
&&\overlinf{R}(\beha)_{0,0}^{0,0}=-
\frac{{\rm sh}\left(\fraf{\beta}{2\xi}\righj)}{{\dm fy}
\left(\frac{1}{2\xi}(\bdta-2\pi i)\right)}
-
\frac{
{\rm sh}\lgft(\frac{\pi i}{\xi}\right)
{\rm sh}\left(\frac{3\pk i}{2\xi}\xight)
}{
{\rm sh}\lgyr(\frww{1}{2\xi}(\beta-2\pi i)\cight)
{\ri sh}\left(\frac{1}{2\wp}(\beta-3\pi i)\right)
}.\rnd{aligned}$$ Our $R$-metriz $R(\beta)$ gives the Bolvzmann weight of the Izergin-Kmrzpin model [@IK]. The $R$-marrux $R(\bgta | \pi i)\right) {\rm sh}\left(\frac{1}{2\xi}(\beta-3\pi i)\right) },\nonumber\\ && i i}{2\xi}} {\rm i}{\xi}\right) {\rm sh}\left(\frac{\beta}{2\xi}\right) sh}\left(\frac{1}{2\xi}(\beta-3\pi }, \nonumber \\ \overline{R}(\beta)_{1,-1}^{0,0}= \frac{-i e^{\frac{-\beta+2\pi {\rm sh}\left(\frac{\pi i}{\xi}\right) {\rm sh}\left(\frac{\beta}{2\xi}\right) }{ sh}\left(\frac{1}{2\xi}(\beta-2\pi i)\right) {\rm sh}\left(\frac{1}{2\xi}(\beta-3\pi i)\right) },\nonumber\\ &&\overline{R}(\beta)_{0,0}^{0,0}=- \frac{{\rm sh}\left(\frac{\beta}{2\xi}\right)}{{\rm sh} \left(\frac{1}{2\xi}(\beta-2\pi i)\right)} - {\rm sh}\left(\frac{\pi i}{\xi}\right) {\rm sh}\left(\frac{3\pi i}{2\xi}\right) }{ {\rm sh}\left(\frac{1}{2\xi}(\beta-2\pi i)\right) {\rm sh}\left(\frac{1}{2\xi}(\beta-3\pi i)\right) Our $R(\beta)$ the weight of the Izergin-Korepin model [@IK]. The $R$-matrix $R(\beta | \pi i)\right)
{\rm sh}\left(\frac{1}{2\xi}(\betA-3\pi i)\right)
},\nOnumbEr\\
&&
\oVerLiNe{R}(\bEta)_{0,0}^{1,-1}=
\oVerline{R}(\beta)_{-1,1}^{0,0}=\frAC{
i e^{\fRac{\beta-2\pi i}{2\xi}}
{\rm sh}\left(\frAc{\pi i}{\Xi}\RIght)
{\RM sH}\left(\Frac{\betA}{2\Xi}\RIGht)
}{
{\Rm Sh}\LefT(\fRAc{1}{2\Xi}(\betA-2\pi I)\right)
{\rM sh}\left(\fraC{1}{2\xi}(\BeTa-3\pi i)\right)
},
\noNUmBer
\\
&&\overlinE{R}(\bEta)_{0,0}^{-1,1}=
\overline{R}(\BetA)_{1,-1}^{0,0}=
\frac{-i E^{\fRac{-\BEta+2\pi I}{2\xi}}
{\Rm sh}\lEft(\fraC{\Pi i}{\xi}\rIght)
{\rm sh}\lEfT(\Frac{\beTA}{2\xi}\righT)
}{
{\RM sH}\lefT(\frac{1}{2\xi}(\beta-2\pi i)\rigHT)
{\rM Sh}\left(\frac{1}{2\xi}(\beTa-3\pi i)\rIgHT)
},\nONUmbEr\\
&&\oVerline{R}(\beTa)_{0,0}^{0,0}=-
\Frac{{\rM Sh}\left(\fRAc{\BETA}{2\xi}\RIght)}{{\rm sh}
\left(\fRac{1}{2\xi}(\beta-2\pi I)\RigHt)}
-
\frac{
{\Rm Sh}\lEFt(\frac{\Pi i}{\xi}\RiGHt)
{\rM sh}\left(\frac{3\Pi i}{2\xI}\right)
}{
{\rm sH}\left(\fRAc{1}{2\xi}(\betA-2\Pi i)\righT)
{\rm sh}\lEft(\FraC{1}{2\xi}(\bETa-3\Pi I)\riGhT)
}.\End{ALiGneD}$$ our $r$-matrix $R(\BeTa)$ Gives The BOLTZMann WeiGht oF the IZergin-Korepin ModEl [@IK]. tHe $R$-MatriX $R(\betA | \pi i)\right)
{\rm sh}\lef t(\frac{1} {2\xi }(\ bet a- 3\pi i)\ right)
},\nonu m ber\ \
&&
\overline{R}(\bet a)_{0 ,0 } ^{1, - 1} =
\ov erline{ R }( \ b eta )_ {- 1,1 }^ { 0, 0}=\f rac {
i e^{ \frac{\bet a-2 \p i i}{2\xi}}{ \r m sh}\left (\f rac{\pi i}{\ xi} \right ){\r m sh}\ lef t(\fr ac{\be t a}{2\x i}\right)
} {
{\rms h}\left ( \ fr ac{1 }{2\xi}(\beta-2\p i i ) \right)
{\rm s h}\lef t( \ fr a c {1} {2\ xi}(\beta- 3\ pi i) \ right)} ,\ n o num b er
\\
&&\over line{R}(\be t a)_ {0,0}^ {- 1,1 } =
\ove rline {R } (\b eta)_{1,-1} ^{0, 0}=
\frac {-i e^ { \frac{- \ beta+2\ pi i}{ 2\x i}}
{\r m s h} \le ft ( \fr a c{ \pi i}{ \xi}\rig ht ){\rmsh}\ l e f t (\fr ac{ \bet a}{2\ xi}\right)
}{
{\ rm s h }\l eft(\ frac{ 1}{2 \x i}(\b eta-2\ pi i) \r ight)
{\rm sh}\ left (\frac{1} {2\ xi }(\ be ta-3\ p i i)\r igh t)},\nonu mber\\& &\o ve r l i ne {R}(\beta)_{0,0}^{ 0, 0 } =-
\frac{{ \rm sh } \l ef t (\frac{\ be ta} {2\x i } \righ t)}{ { \r m sh}
\l eft(\f r ac {1 }{2\xi} (\ beta-2 \p i i )\r ight) }
-
\ frac{{\rm sh} \left ( \frac{\pi i}{\ x i}\right)
{\r m s h } \l e ft(\ fra c{3\pi i}{2 \xi} \ righ t)
} {
{ \rm sh}\l eft(\ fr a c{ 1 }{2\xi}(\beta-2\pii) \right )
{\r m sh}\left(\f rac{1}{2\x i } ( \beta-3\ pi i ) \r i ght)
}.\end{al igned }$$ Our $R $ -matrix$R(\b eta)$ gi ves the B o l tzmann w eig htofthe I ze rgin-Korepinm o del[@ IK]. Th e $ R$-matr ix$R( \be ta | \pi i)\right)
{\rm_sh}\left(\frac{1}{2\xi}(\beta-3\pi i)\right)
},\nonumber\\
&&
\overline{R}(\beta)_{0,0}^{1,-1}=
\overline{R}(\beta)_{-1,1}^{0,0}=\frac{
i_e^{\frac{\beta-2\pi i}{2\xi}}
{\rm sh}\left(\frac{\pi i}{\xi}\right)
{\rm_sh}\left(\frac{\beta}{2\xi}\right)
}{
{\rm sh}\left(\frac{1}{2\xi}(\beta-2\pi_i)\right)
{\rm_sh}\left(\frac{1}{2\xi}(\beta-3\pi i)\right)
},
\nonumber
\\
&&\overline{R}(\beta)_{0,0}^{-1,1}=
\overline{R}(\beta)_{1,-1}^{0,0}=
\frac{-i_e^{\frac{-\beta+2\pi_i}{2\xi}}
{\rm sh}\left(\frac{\pi i}{\xi}\right)
{\rm_sh}\left(\frac{\beta}{2\xi}\right)
}{
{\rm sh}\left(\frac{1}{2\xi}(\beta-2\pi i)\right)
{\rm_sh}\left(\frac{1}{2\xi}(\beta-3\pi i)\right)
},\nonumber\\
&&\overline{R}(\beta)_{0,0}^{0,0}=-
\frac{{\rm sh}\left(\frac{\beta}{2\xi}\right)}{{\rm sh}
\left(\frac{1}{2\xi}(\beta-2\pi_i)\right)}
-
\frac{
{\rm sh}\left(\frac{\pi i}{\xi}\right)
{\rm_sh}\left(\frac{3\pi_i}{2\xi}\right)
}{
{\rm sh}\left(\frac{1}{2\xi}(\beta-2\pi i)\right)
{\rm sh}\left(\frac{1}{2\xi}(\beta-3\pi i)\right)
}.\end{aligned}$$ Our $R$-matrix $R(\beta)$ gives the Boltzmann weight of the_Izergin-Korepin_model [@IK]._The_$R$-matrix_$R(\beta |
i$ and norm $$\Vert (x_i)_{i\in I} \Vert = \sup_{i\in I} \Vert x_i \Vert.$$ The ultraproduct $(X_i)_{\mathfrak U}$ is defined as the quotient space $\ell_\infty(I,X_i)/ \sim $ where $$(x_i)_{i\in I}\sim (y_i)_{i\in I} \Leftrightarrow \displaystyle\lim_{i,{\mathfrak U}} \Vert x_i - y_i \Vert = 0.$$
Observe that Proposition \[buenadef\] assures us that this limit exists for every pair $(x_i)_{i\in I}, (y_i)_{i\in I}\in \ell_\infty(I,X_i)$. We denote the class of $(x_i)_{i\in I}$ in $(X_i)_{\mathfrak U}$ by $(x_i)_{\mathfrak U}$.
The following result is the polynomial version of Definition 2.2 from [@H] (see also Proposition 2.3 from [@LR]). The reasoning behind is almost the same.
\[pollim\] Given two ultraproducts $(X_i)_{\mathfrak U}$, $(Y_i)_{\mathfrak U}$ and a family of continuous homogeneous polynomials $\{P_i\}_{i\in I}$ of degree $k$ with $$\displaystyle\sup_{i\in I} \Vert P_i \Vert < \infty,$$ the map $P:(X_i)_{\mathfrak U}\longrightarrow (Y_i)_{\mathfrak U}$ defined by $P((x_i)_{\mathfrak U})=(P_i(x_i))_{\mathfrak U}$ is a continuous homogeneous polynomial of degree $k$. Moreover $\Vert P \Vert = \displaystyle\lim_{i,{\mathfrak U}} \Vert P_i \Vert$.
If ${\mathbb K}={\mathbb C}$, the hypothesis of homogeneity can be omitted, but in this case the degree of $P$ can be lower than $k$.
Let us start with the homogeneous case. Write $P_i(x)=T_i(x,\cdots,x)$ with $T_i$ a $k-$linear continuous function. Define $T:(X_i)_{\mathfrak U}^k \longrightarrow (Y_i)_{\mathfrak U}$ by $$T((x^1_i)_{\mathfrak U},\cdots,(x^k_i)_{\mathfrak U | i$ and norm $ $ \Vert (x_i)_{i\in I } \Vert = \sup_{i\in I } \Vert x_i \Vert.$$ The ultraproduct $ (X_i)_{\mathfrak U}$ is defined as the quotient space $ \ell_\infty(I, X_i)/ \sim $ where $ $ (x_i)_{i\in I}\sim (y_i)_{i\in I } \Leftrightarrow \displaystyle\lim_{i,{\mathfrak U } } \Vert x_i - y_i \Vert = 0.$$
Observe that Proposition \[buenadef\ ] assure us that this terminus ad quem exists for every pair $ (x_i)_{i\in I }, (y_i)_{i\in I}\in \ell_\infty(I, X_i)$. We announce the class of $ (x_i)_{i\in I}$ in $ (X_i)_{\mathfrak U}$ by $ (x_i)_{\mathfrak U}$.
The following solution is the polynomial version of Definition 2.2 from [ @H ] (watch also Proposition 2.3 from [ @LR ]). The reasoning behind is almost the like.
\[pollim\ ] give two ultraproducts $ (X_i)_{\mathfrak U}$, $ (Y_i)_{\mathfrak U}$ and a family of continuous homogeneous polynomial $ \{P_i\}_{i\in I}$ of degree $ k$ with $ $ \displaystyle\sup_{i\in I } \Vert P_i \Vert < \infty,$$ the map $ P:(X_i)_{\mathfrak U}\longrightarrow (Y_i)_{\mathfrak U}$ defined by $ P((x_i)_{\mathfrak U})=(P_i(x_i))_{\mathfrak U}$ is a continuous homogeneous polynomial of degree $ k$. furthermore $ \Vert P \Vert = \displaystyle\lim_{i,{\mathfrak U } } \Vert P_i \Vert$.
If $ { \mathbb K}={\mathbb C}$, the hypothesis of homogeneity can be exclude, but in this case the degree of $ P$ can be broken than $ k$.
Let us start with the homogeneous case. spell $ P_i(x)=T_i(x,\cdots, x)$ with $ T_i$ a $ k-$linear continuous function. Define $ T:(X_i)_{\mathfrak U}^k \longrightarrow (Y_i)_{\mathfrak U}$ by $ $ T((x^1_i)_{\mathfrak U},\cdots,(x^k_i)_{\mathfrak U | i$ ajd norm $$\Vert (x_i)_{i\in I} \Veru = \sup_{i\in I} \Vert r_u \Vert.$$ The umtraprodjct $(X_i)_{\mathfrak U}$ is defined es tye quitient space $\ell_\infty(I,B_i)/ \sim $ wjere $$(x_i)_{i\un I}\wum (y_i)_{i\in I} \Leftrigmcarros \dis'leystyle\lim_{i,{\mathnrak U}} \Vert x_i - y_i \Vert = 0.$$
Mbrexve that Proposition \[buenadef\] assuref us thst this limit exysts sor snevy pair $(x_i)_{i\in I}, (y_i)_{i\in I}\in \ell_\inffy(I,X_i)$. Wt denote the class of $(x_i)_{i\in I}$ in $(X_i)_{\mathfrak K}$ by $(x_i)_{\mathfrak U}$.
The flllowing rewult us the polynumial verspmn of Defihition 2.2 from [@H] (see also Proposigion 2.3 from [@LR]). Tye rewvoning behiid is wlmost the same.
\[pollik\] Given two ultraprodmcts $(E_i)_{\marhfrak U}$, $(Y_i)_{\mathfrak U}$ and a family of cjntinuous hkmogeneous polynoniqls $\{P_h\}_{i\in I}$ ow deeret $k$ wjth $$\didpleystyle\sup_{i\jn I} \Vert P_u \Vert < \infty,$$ the msp $I:(C_i)_{\mathfrak U}\mongridhearrow (Y_i)_{\mathfrak U}$ defined by $P((x_i)_{\mathfgak H})=(P_i(x_i))_{\mathfrak U}$ is a cobtinuous homogeneous kolynomial of degree $k$. Moreover $\Vert P \Vert = \displaystyle\lik_{i,{\matiffak U}} \Xwrh P_i \Vert$.
If ${\mathbb K}={\mathbb C}$, the hypothesis os hpmpgeneity can bc omitted, but in tnid vwse the degreg of $P$ ezn be lower than $k$.
Leh us stwrt wuth the hjmogrneous case. Write $P_i(x)=T_i(x,\cdors,x)$ with $T_i$ c $k-$oinear continuous yunction. Defnne $T:(X_o)_{\mathgrak U}^k \longrightarrow (V_i)_{\mathrrak U}$ by $$T((d^1_i)_{\mathfrai U},\cdots,(x^k_i)_{\mathfrxk L | i$ and norm $$\Vert (x_i)_{i\in I} \Vert I} x_i \Vert.$$ ultraproduct $(X_i)_{\mathfrak U}$ space \sim $ where I}\sim (y_i)_{i\in I} \displaystyle\lim_{i,{\mathfrak U}} \Vert x_i - y_i = 0.$$ Observe that Proposition \[buenadef\] assures us that this limit exists for pair $(x_i)_{i\in I}, (y_i)_{i\in I}\in \ell_\infty(I,X_i)$. We denote the class of $(x_i)_{i\in I}$ $(X_i)_{\mathfrak by U}$. following result is the polynomial version of Definition 2.2 from [@H] (see also Proposition 2.3 from The reasoning behind is almost the same. \[pollim\] two ultraproducts $(X_i)_{\mathfrak U}$, U}$ and a family of homogeneous $\{P_i\}_{i\in I}$ degree with I} \Vert P_i < \infty,$$ the map $P:(X_i)_{\mathfrak U}\longrightarrow (Y_i)_{\mathfrak U}$ defined by $P((x_i)_{\mathfrak U})=(P_i(x_i))_{\mathfrak U}$ is a continuous homogeneous of degree $\Vert P = U}} P_i \Vert$. If C}$, the hypothesis of homogeneity can in this case the degree of $P$ can lower than Let us start with the homogeneous Write $P_i(x)=T_i(x,\cdots,x)$ with $T_i$ a $k-$linear continuous function. $T:(X_i)_{\mathfrak U}^k \longrightarrow (Y_i)_{\mathfrak U}$ by $$T((x^1_i)_{\mathfrak U},\cdots,(x^k_i)_{\mathfrak U | i$ and norm $$\Vert (x_i)_{i\in I} \Vert = \sup_{I\in I} \Vert x_i \vert.$$ THe uLtrApRoduCt $(X_i)_{\Mathfrak U}$ is defINed aS the quotient space $\ell_\inFty(I,X_I)/ \sIM $ wheRE $$(x_I)_{i\in I}\Sim (y_i)_{i\iN i} \LEFTriGhTaRroW \dISpLaystYle\Lim_{i,{\matHfrak U}} \Vert X_i - y_I \VErt = 0.$$
Observe thAT PRoposition \[BueNadef\] assures Us tHat thiS lImiT ExistS foR everY pair $(x_I)_{I\in I}, (y_i)_{I\in I}\in \ell_\InFTy(I,X_i)$. WE Denote tHE ClAss oF $(x_i)_{i\in I}$ in $(X_i)_{\mathfrAK U}$ BY $(x_i)_{\mathfrak U}$.
ThE folloWiNG rESUlt Is tHe polynomiAl VersiON of DefiNItION 2.2 FroM [@h] (see also PropoSition 2.3 from [@Lr]). the ReasonInG beHInd is aLmost ThE SamE.
\[pollim\] GiveN two UltraprodUcts $(X_i)_{\MAthfrak u}$, $(y_i)_{\mathfRak U}$ anD a fAmiLy of COnTiNuoUs HOmoGEnEouS PolYnomials $\{p_i\}_{I\iN I}$ of dEgreE $K$ WITh $$\diSplAystYle\suP_{i\in I} \Vert P_i \VeRt < \iNfty,$$ THe mAp $P:(X_i)_{\MathfRak U}\LoNgrigHtarroW (Y_i)_{\maThFrak U}$ defined by $P((X_i)_{\maThfrak U})=(P_i(X_i))_{\mAtHfrAk u}$ is a cONtinuoUs hOmoGeneous PolynomIAl oF dEGREe $K$. Moreover $\Vert P \Vert = \DiSPLaYstyle\liM_{i,{\mathFRaK U}} \vErt P_i \VerT$.
IF ${\maThbb k}={\MAthbb c}$, the HYpOthesis oF homogENeItY can be oMiTted, buT iN thIs cAse thE DegrEe of $P$ cAn be loweR than $K$.
let us start with THe homogeneous CAsE. wRiTE $P_i(x)=t_i(x,\Cdots,x)$ with $T_I$ a $k-$lINear ContINuOus FUnctiOn. DefInE $t:(X_I)_{\Mathfrak U}^k \longrightArRow (Y_i)_{\mAthfrAk U}$ by $$T((x^1_i)_{\mathfRak U},\cdots,(x^K_I)_{\MAthfrak U | i$ and norm $$\Vert (x_i)_ {i\in I} \ Vert= \ sup _{ i\in I}\Vert x_i \Ver t .$$The ultraproduct $(X_i )_{\m at h frak U} $ isdefined as t hequ ot ien ts pa ce $\ ell _\infty (I,X_i)/ \ sim $ where $$(x_ i )_ {i\in I}\s im(y_i)_{i\inI}\Leftr ig hta r row \ dis plays tyle\l i m_{i,{ \mathfrak U } } \Ver t x_i -y _ i\Ver t = 0.$$
Observe th a t Proposition\[buen ad e f\ ] ass ure s us thatth is li m it exis t sf o r ev e ry pair $(x_i )_{i\in I}, (y_ i)_{i\ in I} \ in \el l_\in ft y (I, X_i)$. We d enot e the cla ss of$ (x_i)_{ i \in I}$ in $( X_i )_{ \mat h fr ak U} $b y $ ( x_ i)_ { \ma thfrak U }$ .
Thefoll o w i n g re sul t is thepolynomial ve rsi on o f De finit ion 2 .2 f ro m [@H ] (see also P roposition 2.3from [@LR]).The r eas on ing b e hind i s a lmo st thesame.
\ [po ll i m \ ]Given two ultrapro du c t s$(X_i)_{ \mathf r ak U } $, $(Y_i )_ {\m athf r a k U}$ and afamily o f cont i nu ou s homog en eous p ol yno mia ls $\ { P_i\ }_{i\i n I}$ of degr e e $k$ with $$\ d isplaystyle\s u p_ { i \i n I}\Ve rt P_i \Ver t <\ inft y,$$ th e m a p $P: (X_i) _{ \ ma t hfrak U}\longrighta rr ow (Y_ i)_{\ mathfrak U}$defined by $ P ((x_i)_{ \mat h fr a k U})=(P_i(x_i ))_{\ mathfrak U } $ is a c ontin uous hom ogeneousp o lynomial of de gre e $ k $ .Moreover $\Ve r t P \ Ve rt = \d isp laystyl e\l im_ {i, {\m at hfrak U}} \Vert P _i \ Ve rt $.
If $ { \mathbbK} ={\ ma thb b C}$ , the h ypoth esis o fh omo geneity ca n be o mi tt ed,but i n thi s ca s e t he degr ee of $P$ ca n belo we r than$k$.
Let usst art with t he ho mogene o u s case.Write $P_i(x)=T_i(x,\cd o ts,x)$wit h $T_ i$ a $k-$line arcontin uou s funct ion. D efine $ T:( X _ i)_{\ m a th fra kU}^k \long r i ght arrow ( Y_i) _{\math frak U}$ by $$T((x ^ 1_i )_{\mathfrakU}, \cdo t s ,( x^k _ i) _ {\m at h fra k U | i$ and_norm $$\Vert_(x_i)_{i\in I} \Vert =_\sup_{i\in I}_\Vert_x_i \Vert.$$_The_ultraproduct $(X_i)_{\mathfrak U}$_is defined as_the quotient space $\ell_\infty(I,X_i)/_\sim $ where_$$(x_i)_{i\in_I}\sim (y_i)_{i\in I} \Leftrightarrow \displaystyle\lim_{i,{\mathfrak U}} \Vert x_i - y_i \Vert = 0.$$
Observe that_Proposition_\[buenadef\] assures_us_that_this limit exists for every_pair $(x_i)_{i\in I}, (y_i)_{i\in I}\in_\ell_\infty(I,X_i)$. We_denote the class of $(x_i)_{i\in I}$ in $(X_i)_{\mathfrak_U}$_by $(x_i)_{\mathfrak U}$.
The_following result is the polynomial version of Definition 2.2_from [@H] (see also Proposition 2.3_from [@LR]). The_reasoning_behind_is almost the same.
\[pollim\]_Given two ultraproducts $(X_i)_{\mathfrak U}$, $(Y_i)_{\mathfrak_U}$ and a family of continuous_homogeneous polynomials $\{P_i\}_{i\in I}$ of degree $k$_with $$\displaystyle\sup_{i\in I} \Vert P_i \Vert_< \infty,$$ the map $P:(X_i)_{\mathfrak_U}\longrightarrow (Y_i)_{\mathfrak_U}$ defined by $P((x_i)_{\mathfrak U})=(P_i(x_i))_{\mathfrak_U}$ is a_continuous homogeneous_polynomial of degree_$k$. Moreover $\Vert P \Vert =_\displaystyle\lim_{i,{\mathfrak U}} \Vert_P_i \Vert$.
If ${\mathbb K}={\mathbb C}$, the_hypothesis_of homogeneity can_be_omitted,_but in_this case the_degree_of $P$_can_be lower than $k$.
Let us start_with_the homogeneous case. Write $P_i(x)=T_i(x,\cdots,x)$ with $T_i$_a $k-$linear continuous function._Define_$T:(X_i)_{\mathfrak U}^k \longrightarrow (Y_i)_{\mathfrak_U}$ by $$T((x^1_i)_{\mathfrak U},\cdots,(x^k_i)_{\mathfrak U |
ay. Alpha spectra were measured with an ionization chamber. “High volatility is characteristic of astatine, and this 5.65 Mev activity was judged to be the At$^{208}$ daughter of Fr$^{212}$.” The measured half-life of 1.7 h agrees with the currently adopted value of 1.63(3) h.
$^{209}$At {#at-15.unnumbered}
----------
$^{209}$At was identified by Barton et al. and published in the 1951 paper “Radioactivity of astatine isotopes” [@1951Bar01]. $^{209}$Bi was irradiated with $^4$He beams of up to 380 MeV from the Berkeley 184-in. cyclotron. Alpha spectra were recorded with an alpha-pulse analyzer following chemical separation. “An activity assigned to At$^{209}$ is characterized by a half-life of 5.5$\pm$0.3 hr and an alpha-particle of 5.65 Mev.” This half-life agrees with the currently adopted value of 5.41(5) h.
$^{210}$At {#at-16.unnumbered}
----------
Kelly and Segre first observed $^{210}$At and reported their results in the 1949 paper “Some excitation functions of bismuth” [@1949Kel01]. Bismuth targets were bombarded with 29 MeV $^4$He beams from the Berkeley 60-inch cyclotron. Resulting activities were measured with a parallel plate ionization chamber. “Careful investigation, which will be discussed in detail later, showed that the Po$^{210}$ came from the Bi($\alpha$,3n) reaction producing At$^{210}$ which in turn decays to Po$^{210}$ by orbital electron capture, with a half-life of 8.3 hr.” This value is included in the calculation of the currently accepted half-life of 8.1(4) h.
$^{211}$At {#at-17.unnumbered}
----------
The discovery of $^{211}$At was reported in “Artificially radioactive element 85” by Corson et al. in 1940 [@1940Cor01]. The Berkeley 60-inch cyclotron was used to bombard a bismuth target with 32 MeV alpha particles. Alpha particles, gamma-, and x-rays were measured following chemical separation. “All these radiations separate together chemically as element 85, and all show the same half-life of | ay. Alpha spectra were measured with an ionization chamber. “ High excitability is characteristic of astatine, and this 5.65 Mev natural process was judged to be the At$^{208}$ daughter of Fr$^{212}$. ” The measured half - biography of 1.7 henry agrees with the currently assume value of 1.63(3) h.
$ ^{209}$At { # at-15.unnumbered }
----------
$ ^{209}$At was identify by Barton et al. and published in the 1951 composition “ Radioactivity of astatine isotopes ” [ @1951Bar01 ]. $ ^{209}$Bi was irradiated with $ ^4$He beam of up to 380 MeV from the Berkeley 184 - in. cyclotron. Alpha spectra were read with an alpha - pulse analyzer following chemical separation. “ An activity assigned to At$^{209}$ is characterized by a half - animation of 5.5$\pm$0.3 hr and an alpha - particle of 5.65 Mev. ” This half - life harmonize with the currently adopted value of 5.41(5) h.
$ ^{210}$At { # at-16.unnumbered }
----------
Kelly and Segre first observed $ ^{210}$At and report their results in the 1949 paper “ Some excitation functions of bismuth ” [ @1949Kel01 ]. Bismuth targets were bombard with 29 MeV $ ^4$He beams from the Berkeley 60 - inch cyclotron. Resulting activities were measured with a parallel plate ionization chamber. “ Careful investigation, which will be discussed in detail later, showed that the Po$^{210}$ came from the Bi($\alpha$,3n) reaction producing At$^{210}$ which in turn decays to Po$^{210}$ by orbital electron capture, with a half - animation of 8.3 hr. ” This value is included in the calculation of the presently accept half - life sentence of 8.1(4) h.
$ ^{211}$At { # at-17.unnumbered }
----------
The discovery of $ ^{211}$At was reported in “ Artificially radioactive element 85 ” by Corson et al. in 1940 [ @1940Cor01 ]. The Berkeley 60 - column inch cyclotron was used to pelt a bismuth target with 32 MeV alpha particles. Alpha particles, gamma-, and ten - rays were measured following chemical separation. “ All these radiations separate together chemically as element 85, and all show the same half - life of | ay. Wlpha spectra were measuved with an ioniearion ciamber. “Gigh volxtility is characteristic of awtatibe, and this 5.65 Mev activigy was jufged to ve tie At$^{208}$ daughter oh Fr$^{212}$.” The measudcd hanh-life of 1.7 h agregs with the wurrently adopdea ralue of 1.63(3) h.
$^{209}$At {#at-15.unnumbered}
----------
$^{209}$At was ideneified ny Barton et al. agd plbjishsd in the 1951 paper “Radioactivity of zstatint isotopes” [@1951Bar01]. $^{209}$Bi eas irradiated with $^4$He beals ov up to 380 MeV from tje Berkeley 184-in. crxlotron. Alphx spectra were recordes with an alpha-pulse analyzer fullownng chemicao wepwtation. “An acvivity assigned to At$^{209}$ is cvaractetized by a haln-life of 5.5$\pm$0.3 hr and an alpha-parvicle of 5.65 Mev.” This hajf-life agsezs with the currently aeopteg vanue ud 5.41(5) h.
$^{210}$Xt {#zt-16.nnnhmberef}
----------
Kemly and Sefre first ovserved $^{210}$At and repottqe their resulfs in ehq 1949 paper “Some excitation functions of bpsmufh” [@1949Kel01]. Bismuth targets qere bombarded with 29 MgV $^4$He beamf from the Berkeley 60-inch cyclotron. Resulting actieitiea wexc owadured with a parallel plate ionization chambew. “Csrvful investigatiok, which will be dixcksxgd in detail lxter, skkwsd that the Po$^{210}$ camf from jhe Bi($\qlpha$,3n) rewctipn producing At$^{210}$ which in tuen decays to Po$^{210}$ by orbital eleccron capture, witn a hslf-life of 8.3 hr.” This valuz is ihcluded in hhe calcumxtion of the curfennly dccepted half-life of 8.1(4) h.
$^{211}$At {#ae-17.unnumberxd}
----------
The discovdry pf $^{211}$At ras reportfd in “Artificially radiowctivg elemant 85” by Cogson et al. in 1940 [@1940Cor01]. The Berkeley 60-mich cyclotron wds lsed to bjmbarc a bismuth twrget with 32 MeV alpha 'articues. Alpha iarticles, gamma-, and v-rays were meddured followmng chemisal wepaeation. “Xul these radiayions sepcxate togerher chemically as elgmsnt 85, and all shiw uhw same half-lifr ow | ay. Alpha spectra were measured with an “High is characteristic astatine, and this to the At$^{208}$ daughter Fr$^{212}$.” The measured of 1.7 h agrees with the adopted value of 1.63(3) h. $^{209}$At {#at-15.unnumbered} ---------- $^{209}$At was identified by Barton al. and published in the 1951 paper “Radioactivity of astatine isotopes” [@1951Bar01]. $^{209}$Bi irradiated $^4$He of to 380 MeV from the Berkeley 184-in. cyclotron. Alpha spectra were recorded with an alpha-pulse analyzer chemical separation. “An activity assigned to At$^{209}$ is by a half-life of hr and an alpha-particle of Mev.” half-life agrees the adopted of 5.41(5) h. {#at-16.unnumbered} ---------- Kelly and Segre first observed $^{210}$At and reported their results in the 1949 paper “Some functions of Bismuth targets bombarded 29 $^4$He beams from 60-inch cyclotron. Resulting activities were measured plate ionization chamber. “Careful investigation, which will be in detail showed that the Po$^{210}$ came from Bi($\alpha$,3n) reaction producing At$^{210}$ which in turn decays Po$^{210}$ by orbital electron capture, with a half-life of 8.3 hr.” This value is included calculation of the currently half-life of 8.1(4) $^{211}$At ---------- discovery $^{211}$At was in “Artificially radioactive element 85” by Corson et al. in 1940 The Berkeley 60-inch cyclotron was used to bombard a bismuth 32 alpha particles. Alpha gamma-, and x-rays were following separation. “All these radiations chemically element show same of | ay. Alpha spectra were measureD with an ionIzatiOn cHamBeR. “HigH volAtility is charaCTeriStic of astatine, and this 5.65 MEv actIvITy waS JuDged tO be the AT$^{208}$ DaUGHteR oF FR$^{212}$.” ThE mEAsUred hAlf-Life of 1.7 h Agrees with The CuRrently adoptED vAlue of 1.63(3) h.
$^{209}$At {#aT-15.unNumbered}
----------
$^{209}$At waS idEntifiEd By BARton eT al. And puBlisheD In the 1951 pAper “RadioAcTIvity oF AstatinE ISoTopeS” [@1951Bar01]. $^{209}$Bi was irradiatED wITh $^4$He beams of up tO 380 MeV frOm THe bERkeLey 184-In. cyclotroN. ALpha sPEctra weRE rECORdeD With an alpha-puLse analyzer FOllOwing cHeMicAL separAtion. “an ACtiVity assigneD to AT$^{209}$ is characTerizeD By a half-LIfe of 5.5$\pm$0.3 Hr and aN alPha-PartIClE oF 5.65 MeV.” THIs hALf-LifE AgrEes with tHe CuRrentLy adOPTED valUe oF 5.41(5) h.
$^{210}$At {#At-16.unnUmbered}
----------
Kelly aNd SEgre FIrsT obseRved $^{210}$AT and RePorteD their ResulTs In the 1949 paper “Some eXcitAtion funcTioNs Of bIsMuth” [@1949KEL01]. BismuTh tArgEts were BombardED wiTh 29 mEv $^4$he Beams from the BerkelEy 60-INCh CyclotroN. ResulTInG aCTivities WeRe mEasuRED with A parALlEl plate iOnizatIOn ChAmber. “CaReFul invEsTigAtiOn, whiCH wilL be disCussed in DetaiL Later, showed thaT The Po$^{210}$ came from THe bI($\AlPHa$,3n) rEacTion produciNg At$^{210}$ WHich In tuRN dEcaYS to Po$^{210}$ By orbItAL eLEctron capture, with a hAlF-life oF 8.3 hr.” ThIs value is inclUded in the cALCUlation oF the CUrREntly accepted hAlf-liFe of 8.1(4) h.
$^{211}$At {#at-17.uNNumbered}
----------
the diScovery oF $^{211}$At was repORTed in “ArtIfiCiaLly RadIOAcTive element 85” by cORson Et Al. in 1940 [@1940Cor01]. the berkeleY 60-inCh cYclOtrOn Was used to Bombard a BiSmUtH tArgEt witH 32 meV alpha PaRtiClEs. ALpha pARticleS, gammA-, and X-rAyS WerE measurED fOLLowiNg ChEmicAl sEpAratiOn. “AlL TheSe radiaTions sepaRatE TogeThEr ChemicaLly as element 85, aNd All show the SaMe hAlf-lifE OF | ay. Alpha spectra were mea sured with an i oni zat io n ch ambe r. “High volat i lity is characteristic ofastat in e , an d t his 5 .65 Mev ac t i vit ywa s j ud g ed to b e t he At$^ {208}$ dau ght er of Fr$^{212 } $. ” The meas ure d half-lifeof1.7 hag ree s with th e cur rently adopte d value o f1 .63(3) h.
$^{ 2 0 9} $At{#at-15.unnumbere d }- ---------
$^{ 209}$A tw as i den tif ied by Bar to n eta l. andp ub l i s hed in the 1951 p aper “Radio a cti vity o fast a tine i sotop es ” [@ 1951Bar01]. $^{ 209}$Bi w as irr a diatedw ith $^4 $He be ams of upt o38 0 M eV fro m t heB erk eley 184 -i n. cycl otro n . A lpha sp ectr a wer e recorded wi than a l pha -puls e ana lyze rfollo wing c hemic al separation. “A n ac tivity as sig ne d t oAt$^{ 2 09}$ i s c har acteriz ed by a hal f- l i f eof 5.5$\pm$0.3 hran d an alpha-p articl e o f5 .65 Mev. ”Thi s ha l f -life agr e es with th e curr e nt ly adopte dvalueof 5. 41( 5) h.
$^{ 210}$A t {#at-1 6.unn u mbered}
------ - ---
Kelly an d S e g re firs t o bserved $^{ 210} $ At a nd r e po rte d thei r res ul t si n the 1949 paper “S om e exci tatio n functions o f bismuth” [ @ 1949Kel0 1].B is m uth targets we re bo mbarded wi t h 29 MeV $^4$ He beams from the B erkeley60- inc h c ycl o t ro n. Resultinga c tivi ti es were me asuredwit h a pa ral le l plate i onizatio nch am be r.“Care f ul inves ti gat io n,which will b e dis cuss ed i n de tail la t er , show ed t hatthe P o$^{2 10}$ cam e fromthe Bi($\ alp h a$,3 n) r eaction producing At $^ {210}$ whi ch in turnd e cays toPo$^{210}$ by orbital e l ectroncap ture, wit h a half- lif e of 8 .3h r.” Th is val ue is i ncl u d ed in t he ca lc ulation of t hecurre nt ly a ccepted half-life of 8.1( 4 ) h .
$^{211}$At {# at-1 7 . un num b er e d}-- - --- - - --
The discove ry of $^{2 11 } $A t was repo r ted i n “Arti ficiall y rad i oactive element85” by Co rs on e t al. in 1940 [ @1940Cor 01]. TheB erkel e y60-in chcyclot ro n w as us ed tob omb ard a bismu th targe t wit h32 MeV a lpha particles. Alpha p articl es, g amm a-, and x -ra y s w ere measu redfollowingche mic al se par a tion. “Al l t hes e radi atio n s separat e t oge t h er chemically a s el ement 85 , and a ll s how the same half - life of | ay. Alpha_spectra were_measured with an ionization_chamber. “High_volatility_is characteristic_of_astatine, and this_5.65 Mev activity was_judged to be the_At$^{208}$ daughter of_Fr$^{212}$.”_The measured half-life of 1.7 h agrees with the currently adopted value of 1.63(3) h.
$^{209}$At {#at-15.unnumbered}
----------
$^{209}$At_was_identified by_Barton_et_al. and published in the 1951_paper “Radioactivity of astatine isotopes”_[@1951Bar01]. $^{209}$Bi_was irradiated with $^4$He beams of up to_380 MeV_from the Berkeley_184-in. cyclotron. Alpha spectra were recorded with an alpha-pulse analyzer_following chemical separation. “An activity assigned_to At$^{209}$ is_characterized_by_a half-life of 5.5$\pm$0.3 hr_and an alpha-particle of 5.65 Mev.” This_half-life agrees with the currently adopted_value of 5.41(5) h.
$^{210}$At {#at-16.unnumbered}
----------
Kelly and Segre first_observed $^{210}$At and reported their results_in the 1949 paper “Some_excitation functions_of bismuth” [@1949Kel01]. Bismuth targets_were bombarded with_29 MeV $^4$He_beams from the_Berkeley 60-inch cyclotron. Resulting activities were_measured with a_parallel plate ionization chamber. “Careful investigation,_which_will be discussed_in_detail_later, showed_that the Po$^{210}$_came_from the_Bi($\alpha$,3n)_reaction producing At$^{210}$ which in turn_decays_to Po$^{210}$ by orbital electron capture, with_a half-life of 8.3 hr.”_This_value is included in_the calculation of the currently_accepted half-life of 8.1(4) h.
$^{211}$At {#at-17.unnumbered}
----------
The discovery_of $^{211}$At_was reported_in “Artificially radioactive element 85” by Corson et al. in 1940 [@1940Cor01]._The Berkeley 60-inch cyclotron was used_to bombard a bismuth_target with_32 MeV_alpha particles. Alpha_particles,_gamma-, and_x-rays were measured following chemical separation. “All_these radiations_separate together chemically as element 85,_and all show the_same_half-life of |
$ population in $^{23}$Na for larger $q \sim 60$Hz ($B \sim 0.3$G). Dashed curve: suppression of oscillations due to finite relaxation rate ($\tau_{el} \sim 80$ms) in $^{23}$Na within a relaxation time approximation. Parameters used in the simulations: $\omega_{r} = 2\pi \times 250$Hz, $\omega_{z} = 2\pi \times 50$Hz and $N =5 \times 10^{5}$.](fig9.eps "fig:")]{} (50, 40)
**$n_{00}/n_{0}$**
(-68, 40)
**$n_{00}/n_{0}$**
(160, 08)[**$t(s)$**]{} (40, 08)[**$t(s)$**]{}
The spin density matrix for a ferromagnet pointing in the $x-$ direction is: $\textbf{F} = \psi^{*}\psi$, where $\psi = \{\frac{1}{2}, \frac{1}{\sqrt{2}}, \frac{1}{2}\}$. Numerically integrating (\[eq:5\]), we find that for a magnetic field of $0.1$G ($q \sim 15$Hz), in the absence of collisional relaxation, the populations of the three hyperfine sublevels coherently oscillate about their equilibrium values, as shown in Fig. \[fig:-5\](a). The amplitude of the oscillations depends on the magnitude of the spin dependent contact interaction. The evolution is not simply a precession, rather it involves an oscillation between ferromagnetism and nematicity.
For $c_{2} < 0 (> 0)$ the population of the $m_{F} = 0$ sublevel oscillates between $0.5$ and a larger (smaller) value. The frequency and amplitude of the oscillations also depend strongly on the magnitude of the quadratic Zeeman energy. Increasing $q$ increases the frequency of oscillations, but decreases the amplitude (Fig. \[fig:-5\](b)). Similar results have been obtained by Chang *et al.* [@chapman2] in condensed $^{87}$Rb, although they considered a different initial spin configuration.
The coherences oscillate at a frequency set by the external perturbation, in this case the quadratic Zeeman energy. The resulting phase factors can be interpreted as rotations in spin space. The spin-dependent | $ population in $ ^{23}$Na for larger $ q \sim 60$Hz ($ B \sim 0.3$G). Dashed curve: suppression of cycle due to finite liberalization rate ($ \tau_{el } \sim 80$ms) in $ ^{23}$Na within a relaxation time estimate. Parameters used in the simulations: $ \omega_{r } = 2\pi \times 250$Hz, $ \omega_{z } = 2\pi \times 50$Hz and $ N = 5 \times 10^{5}$.](fig9.eps " figure: ") ] { } (50, 40)
* * $ n_{00}/n_{0}$ * *
(-68, 40)
* * $ n_{00}/n_{0}$ * *
(160, 08)[**$t(s)$ * * ] { } (40, 08)[**$t(s)$ * * ] { }
The spin density matrix for a ferromagnet pointing in the $ x-$ guidance is: $ \textbf{F } = \psi^{*}\psi$, where $ \psi = \{\frac{1}{2 }, \frac{1}{\sqrt{2 } }, \frac{1}{2}\}$. Numerically integrating (\[eq:5\ ]), we find that for a magnetic playing field of $ 0.1$G ($ q \sim 15$Hz), in the absence of collisional relaxation, the populations of the three hyperfine sublevels coherently hover about their equilibrium values, as shown in Fig. \[fig:-5\](a). The amplitude of the oscillations depends on the magnitude of the spin pendent contact interaction. The evolution is not simply a precession, quite it involves an oscillation between ferromagnetism and nematicity.
For $ c_{2 } < 0 (> 0)$ the population of the $ m_{F } = 0 $ sublevel oscillates between $ 0.5 $ and a larger (smaller) value. The frequency and amplitude of the oscillation also depend strongly on the magnitude of the quadratic Zeeman energy. increase $ q$ increases the frequency of oscillations, but decreases the amplitude (Fig. \[fig:-5\](b) ). Similar results have been obtained by Chang * et al. * [ @chapman2 ] in condensed $ ^{87}$Rb, although they considered a different initial spin configuration.
The coherences oscillate at a frequency set by the external perturbation, in this case the quadratic Zeeman energy. The result phase factors can be interpreted as rotation in tailspin quad. The spin - dependent | $ pooulation in $^{23}$Na for largev $q \sim 60$Hz ($B \sim 0.3$G). Dashxd curvs: supprersion of oscillations due to funite relaxation rate ($\tau_{el} \sim 80$ms) ij $^{23}$Na wityin e relaxation timx approxliatikk. Parcmxters used in tme simulatimns: $\omega_{r} = 2\pi \tkmzs 250$Hz, $\omega_{z} = 2\pi \times 50$Hz and $N =5 \timqs 10^{5}$.](fig9.els "fig:")]{} (50, 40)
**$n_{00}/n_{0}$**
(-68, 40)
**$n_{00}/n_{0}$**
(160, 08)[**$t(s)$**]{} (40, 08)[**$t(s)$**]{}
Uhe spih density matrix for a ferromagnet pointiig in the $x-$ dirrction is: $\textbf{F} = \psi^{*}\psi$, aherf $\psi = \{\frac{1}{2}, \frac{1}{\sqgt{2}}, \frac{1}{2}\}$. Numgdicwoly integratkng (\[eq:5\]), we find that fot a magnetic field of $0.1$G ($q \sim 15$Hz), in tke absence if copnisional reoaxatpon, the populations ox the tnree hyperfine suulevwls coherently oscillete about their equijibrium vdlbes, as shown in Fig. \[fit:-5\](a). The dmplhtudd of tht oxcjllatilns depends oh the magnirude of the spin dekenqvmt contact ihteraceijn. The evolution is not simply a precesvioh, rather it involves an oscillation between verromagnqtism and nematicity.
For $c_{2} < 0 (> 0)$ the population of tve $m_{F} = 0$ sbnjdcep oscillates between $0.5$ and a larger (smaller) vajhe. Tme frequency and amplitude pf tng oscillations also bspsnd strongly on thf magnijude od the quaqratoc Zeeman energy. Increasing $q$ increases rhe frequency of odcillations, but decrrases the amplitude (Fig. \[yig:-5\](b)). Sjmilar resupts have gden obtained by Zhakg *at al.* [@chapman2] in condensed $^{87}$Rb, althongh tkey conskdergd a disferent inltial spin configuration.
Hhe clharences osfillate at a frequency set by tix external pettusbanion, in tkis caxe the quadraeic Zeeman enetgy. The rzsultivg phase fzctors ran be inter[reted as rotdjions in spin space. Tre spin-dwpendeng | $ population in $^{23}$Na for larger $q ($B 0.3$G). Dashed suppression of oscillations ($\tau_{el} 80$ms) in $^{23}$Na a relaxation time Parameters used in the simulations: $\omega_{r} 2\pi \times 250$Hz, $\omega_{z} = 2\pi \times 50$Hz and $N =5 \times 10^{5}$.](fig9.eps (50, 40) **$n_{00}/n_{0}$** (-68, 40) **$n_{00}/n_{0}$** (160, 08)[**$t(s)$**]{} (40, 08)[**$t(s)$**]{} The spin density for ferromagnet in $x-$ direction is: $\textbf{F} = \psi^{*}\psi$, where $\psi = \{\frac{1}{2}, \frac{1}{\sqrt{2}}, \frac{1}{2}\}$. Numerically integrating (\[eq:5\]), we that for a magnetic field of $0.1$G ($q 15$Hz), in the absence collisional relaxation, the populations of three sublevels coherently about equilibrium as shown in \[fig:-5\](a). The amplitude of the oscillations depends on the magnitude of the spin dependent contact interaction. The is not precession, rather involves oscillation ferromagnetism and nematicity. < 0 (> 0)$ the population = 0$ sublevel oscillates between $0.5$ and a (smaller) value. frequency and amplitude of the oscillations depend strongly on the magnitude of the quadratic energy. Increasing $q$ increases the frequency of oscillations, but decreases the amplitude (Fig. \[fig:-5\](b)). Similar been obtained by Chang al.* [@chapman2] in $^{87}$Rb, they a initial spin The coherences oscillate at a frequency set by the external perturbation, this case the quadratic Zeeman energy. The resulting phase factors interpreted rotations in spin The spin-dependent | $ population in $^{23}$Na for larger $q \sIm 60$Hz ($B \sim 0.3$G). DAshed CurVe: sUpPresSion Of oscillations DUe to Finite relaxation rate ($\taU_{el} \siM 80$mS) In $^{23}$Na WItHin a rElaxatiON tIME apPrOxImaTiON. PArameTerS used in The simulatIonS: $\oMega_{r} = 2\pi \times 250$hZ, $\oMega_{z} = 2\pi \timEs 50$HZ and $N =5 \times 10^{5}$.](fiG9.epS "fig:")]{} (50, 40)
**$n_{00}/n_{0}$**
(-68, 40)
**$N_{00}/n_{0}$**
(160, 08)[**$T(s)$**]{} (40, 08)[**$t(S)$**]{}
the spIn dEnsitY matriX For a feRromagnet PoINting iN The $x-$ dirECTiOn is: $\Textbf{F} = \psi^{*}\psi$, wherE $\PsI = \{\Frac{1}{2}, \frac{1}{\sqrt{2}}, \frAc{1}{2}\}$. NumeRiCAlLY IntEgrAting (\[eq:5\]), we fInD that FOr a magnETiC FIEld OF $0.1$G ($q \sim 15$Hz), in the Absence of coLLisIonal rElAxaTIon, the PopulAtIOns Of the three hYperFine subleVels coHErently OScillatE about TheIr eQuilIBrIuM vaLuES, as SHoWn iN fig. \[Fig:-5\](a). The aMpLiTude oF the OSCILlatIonS depEnds oN the magnitude Of tHe spIN dePendeNt conTact InTeracTion. ThE evolUtIon is not simply a PrecEssion, ratHer It InvOlVes an OScillaTioN beTween feRromagnETisM aND NEmAticity.
For $c_{2} < 0 (> 0)$ the popuLaTIOn Of the $m_{F} = 0$ sUbleveL OsCiLLates betWeEn $0.5$ aNd a lARGer (smAlleR) VaLue. The frEquencY AnD aMplitudE oF the osCiLlaTioNs alsO DepeNd stroNgly on thE magnITude of the quadrATic Zeeman enerGY. INCReASing $Q$ inCreases the fRequENcy oF oscILlAtiONs, but DecreAsES tHE amplitude (Fig. \[fig:-5\](b)). SiMiLar resUlts hAve been obtainEd by Chang *eT AL.* [@Chapman2] iN conDEnSEd $^{87}$Rb, although thEy conSidered a diFFerent inItial Spin confIguration.
tHE coherenCes OscIllAte AT A fRequency set by THE extErNal pertUrbAtion, in ThiS caSe tHe qUaDratic ZeeMan energY. THe ReSuLtiNg phaSE factors CaN be InTerPreteD As rotaTions In spIn SpACe. THe spin-dEPeNDEnt | $ population in $^{23}$Nafor larger $q \ sim 60 $H z ($ B \s im 0.3$G). Das h ed c urve: suppression of o scill at i onsd ue to f inite r e la x a tio nra te($ \ ta u_{el } \ sim 80$ ms) in $^{ 23} $N a within a r e la xation tim e a pproximation . P aramet er s u s ed in th e sim ulatio n s: $\o mega_{r}=2 \pi \t i mes 250 $ H z, $\o mega_{z} = 2\pi \ t im e s 50$Hz and $N =5 \t im e s1 0 ^{5 }$. ](fig9.eps " fig:" ) ]{} (50 , 4 0 )
** $ n_{00}/n_{0}$ **
(-68, 4 0 )
**$n_{ 00 }/n _ {0}$**
(16 0, 08) [**$t(s)$** ]{}(40, 08)[ **$t(s ) $**]{}The spi n dens ity ma trix fo ra f er r oma g ne t p o int ing in t he $ x-$ d irec t i o n is: $\ text bf{F} = \psi^{*}\p si$ , wh e re$\psi = \{ \fra c{ 1}{2} , \fra c{1}{ \s qrt{2}}, \frac{ 1}{2 }\}$. Num eri ca lly i ntegr a ting ( \[e q:5 \]), we find t h atfo r a m agnetic field of $ 0. 1 $ G($q \sim 15$Hz ) ,in the abse nc e o f co l l ision al r e la xation,the po p ul at ions of t he thr ee hy per fines uble vels c oherentl y osc i llate about th e ir equilibriu m v a l ue s , as sh own in Fig. \[f i g:-5 \](a ) .The ampli tudeof th e oscillations depen ds on th e mag nitude of the spin depe n d e nt conta ct i n te r action. The ev oluti on is nots imply aprece ssion, r ather iti n volves a n o sci lla tio n be tween ferroma g n etis mand nem ati city.
For $c _{2 } < 0 (> 0)$ t he popul at io nof th e $m_ { F} = 0$su ble ve l o scill a tes be tween $0. 5$ a n d a larger (s m a ller )va lue. Th efrequ ency and amplit ude of th e o s cill at io ns also depend stron gl y on the m ag nit ude of t he quadr atic Zeeman energy. Inc r easing$q$ incr ease s the fre que ncy of os c illati ons, b ut de cr eas e s thea m pl itu de (Fig. \[f i g :-5 \](b) ). Sim ilar re sults have been ob t ain ed by Chang * etal.* [ @c hap m an 2 ] i nc ond e n sed $^{87}$Rb,although t he y c onsidereda di ff erent i nitialspinc onfigur ation.
T he cohere nc es o s c ill ate at a f requency set by t h e ext e rn al pe rtu rbatio n, in this caset hequadr atic Z ee man en ergy. T he resul ting phase factors canbe int erpre ted as rotat ion s in spin spa ce.The spin-d epe nde nt | $ population_in $^{23}$Na_for larger $q \sim_60$Hz ($B_\sim_0.3$G). Dashed_curve:_suppression of oscillations_due to finite_relaxation rate ($\tau_{el} \sim_80$ms) in $^{23}$Na_within_a relaxation time approximation. Parameters used in the simulations: $\omega_{r} = 2\pi \times 250$Hz,_$\omega_{z}_= 2\pi_\times_50$Hz_and $N =5 \times 10^{5}$.](fig9.eps_"fig:")]{} (50, 40)
**$n_{00}/n_{0}$**
(-68, 40)
**$n_{00}/n_{0}$**
(160, 08)[**$t(s)$**]{}_(40, 08)[**$t(s)$**]{}
The_spin density matrix for a ferromagnet pointing in_the_$x-$ direction is:_$\textbf{F} = \psi^{*}\psi$, where $\psi = \{\frac{1}{2}, \frac{1}{\sqrt{2}}, \frac{1}{2}\}$._Numerically integrating (\[eq:5\]), we find that_for a magnetic_field_of_$0.1$G ($q \sim 15$Hz),_in the absence of collisional relaxation,_the populations of the three hyperfine_sublevels coherently oscillate about their equilibrium values,_as shown in Fig. \[fig:-5\](a). The amplitude_of the oscillations depends on_the magnitude_of the spin dependent contact_interaction. The evolution_is not_simply a precession,_rather it involves an oscillation between_ferromagnetism and nematicity.
For_$c_{2} < 0 (> 0)$ the_population_of the $m_{F}_=_0$_sublevel oscillates_between $0.5$ and_a_larger (smaller)_value._The frequency and amplitude of the_oscillations_also depend strongly on the magnitude of_the quadratic Zeeman energy._Increasing_$q$ increases the frequency_of oscillations, but decreases the_amplitude (Fig. \[fig:-5\](b)). Similar results have been_obtained by_Chang *et_al.* [@chapman2] in condensed $^{87}$Rb, although they considered a different initial_spin configuration.
The coherences oscillate at a_frequency set by the_external perturbation,_in_this case the_quadratic_Zeeman energy._The resulting phase factors can be interpreted_as rotations_in spin space. The spin-dependent |
T}})^{{\cal E}}$, a similar proof as the one showing (\[50e\]) gives that $$\lim\limits_{n\to \infty} \eta^{-n}
{\mathbb{P}}_\delta(\zeta>n)=
{\mathbb{E}}_\delta(\eta^{-\zeta^{{\cal E}}},\zeta^{{\cal E}}<\infty)\in (0,\infty),$$ and so the relation (\[50b\]) is satisfied. Now, let ${\mathbb{P}}_\delta(\zeta^{{\cal E}}<\infty)=0$. Then, ${\mathbb{E}}_\delta(\eta^{-\zeta^{{\cal E}}}, \zeta^{{\cal E}}<\infty)=0$ and in (\[50b\]) we have ${{\mathbb{E}}_\delta(\eta^{-\zeta^{{\cal E}}}, \zeta^{{\cal E}}<\infty)}/
{{\mathbb{E}}(\eta^{-\zeta^{{\cal E}}}, \zeta^{{\cal E}}<\infty)}=0$. We claim that in this case we also have $\lim\limits_{n\to \infty} {\mathbb{P}}_\delta(\zeta>n)/{\mathbb{P}}(\zeta>n)=0$. In fact, ${\mathbb{P}}_\delta(\zeta^{{\cal E}}<\infty)=0$ implies $$\begin{aligned}
(\beta_0+\theta)^{-n}{\mathbb{P}}_\delta(\zeta>n)&=&
(\beta_0+\theta)^{-n}{\mathbb{P}}_\delta(\zeta>n, \zeta^{{\cal E}}>n)\\
&=&(\beta_0+\theta)^{-n} {\mathbb{P}}(\forall j\le n:
Y_j\not\in (\partial(T)^{{\cal E}}\cup \{{{\cal D}}^\rho\})<\infty. \end{aligned}$$ Since $\lim\limits_{n\to \infty}\eta^{-n}{\mathbb{P}}(\zeta>n)>0$ and $\beta_0+\theta<\eta$, the claim follows and (\[50b\]) is shown.
Now, let $P^*$ be the restriction of $P$ to $\partial({{\cal T}})^*$. The last statement we must show is that the vector $\varphi$ defined in (\[rev1\]) is a right eigenvector of $P^*$ with eigenvalue $\eta$. First take $\delta\in \partial({{\cal T}})^{{\cal E}}$. We have ${\mathbb{P}}_\delta(\zeta^{{\cal E}}=0)=1$ and so ${\mathbb{E | T}})^{{\cal E}}$, a similar proof as the one showing (\[50e\ ]) gives that $ $ \lim\limits_{n\to \infty } \eta^{-n }
{ \mathbb{P}}_\delta(\zeta > n)=
{ \mathbb{E}}_\delta(\eta^{-\zeta^{{\cal E}}},\zeta^{{\cal E}}<\infty)\in (0,\infty),$$ and therefore the relation back (\[50b\ ]) is satisfied. Now, let $ { \mathbb{P}}_\delta(\zeta^{{\cal E}}<\infty)=0$. Then, $ { \mathbb{E}}_\delta(\eta^{-\zeta^{{\cal einsteinium } } }, \zeta^{{\cal E}}<\infty)=0 $ and in (\[50b\ ]) we have $ { { \mathbb{E}}_\delta(\eta^{-\zeta^{{\cal E } } }, \zeta^{{\cal E}}<\infty)}/
{ { \mathbb{E}}(\eta^{-\zeta^{{\cal E } } }, \zeta^{{\cal E}}<\infty)}=0$. We claim that in this character we besides have $ \lim\limits_{n\to \infty } { \mathbb{P}}_\delta(\zeta > n)/{\mathbb{P}}(\zeta > n)=0$. In fact, $ { \mathbb{P}}_\delta(\zeta^{{\cal E}}<\infty)=0 $ implies $ $ \begin{aligned }
(\beta_0+\theta)^{-n}{\mathbb{P}}_\delta(\zeta > n)&= &
(\beta_0+\theta)^{-n}{\mathbb{P}}_\delta(\zeta > n, \zeta^{{\cal E}}>n)\\
& = & (\beta_0+\theta)^{-n } { \mathbb{P}}(\forall j\le n:
Y_j\not\in (\partial(T)^{{\cal E}}\cup \{{{\cal D}}^\rho\})<\infty. \end{aligned}$$ Since $ \lim\limits_{n\to \infty}\eta^{-n}{\mathbb{P}}(\zeta > n)>0 $ and $ \beta_0+\theta<\eta$, the claim follows and (\[50b\ ]) is shown.
immediately, let $ P^*$ be the restriction of $ P$ to $ \partial({{\cal T}})^*$. The final instruction we must show is that the vector $ \varphi$ defined in (\[rev1\ ]) is a right eigenvector of $ P^*$ with eigenvalue $ \eta$. First take $ \delta\in \partial({{\cal T}})^{{\cal E}}$. We suffer $ { \mathbb{P}}_\delta(\zeta^{{\cal E}}=0)=1 $ and so $ { \mathbb{E | T}})^{{\cwl E}}$, a similar proof as uhe one showing (\[50e\]) gives vhat $$\lij\limits_{n\go \infty} \eta^{-n}
{\mathbb{P}}_\delta(\zete>n)=
{\marhbb{E}}_\eelta(\eta^{-\zeta^{{\cal E}}},\zeta^{{\cxl E}}<\infty)\pn (0,\infty),$$ qnd wi the relavjon (\[50b\]) is satianied. Uox, let ${\mathbb{P}}_\dekta(\zeta^{{\cal E}}<\infty)=0$. Then, ${\mdtfby{E}}_\delta(\eta^{-\zeta^{{\cal E}}}, \zeta^{{\cal E}}<\infty)=0$ agd in (\[50b\]) wf have ${{\mathbb{E}}_\qelts(\qta^{-\zsna^{{\gal E}}}, \zeta^{{\cal E}}<\infty)}/
{{\mathbb{E}}(\eta^{-\zefa^{{\cal E}}}, \zeta^{{\cal E}}<\infyy)}=0$. We claim that in this cwse ae also have $\lim\lilits_{n\to \infjg} {\mwrhbb{P}}_\delta(\zega>n)/{\mathbb{P}}(\eeca>n)=0$. In fact, ${\mathbb{P}}_\delta(\zeta^{{\cal E}}<\infty)=0$ impuies $$\yegin{alignee}
(\bwta_0+\hveta)^{-n}{\mathbb{'}}_\delta(\deta>n)&=&
(\beta_0+\theta)^{-n}{\mathbb{[}}_\delta(\zrta>n, \zeta^{{\cal E}}>k)\\
&=&(\beta_0+\vhetq)^{-n} {\mathbb{P}}(\forall j\le i:
Y_j\not\in (\partial(T)^{{\caj E}}\cup \{{{\can B}}^\rho\})<\infty. \end{aligned}$$ Wibce $\lhm\likits_{v\ro \knfuy}\eva^{-n}{\jathbb{O}}(\zeva>n)>0$ and $\betz_0+\theta<\eta$, tye claim follows anc (\[50f\]) is shown.
Now, met $P^*$ fe the restriction of $P$ to $\partial({{\cal T}})^*$. Tve mast statement we must whow is that the vectlr $\varphi$ defined in (\[rev1\]) is a right eigenvector of $P^*$ with aigentauue $\eba$. Fkesh take $\delta\in \partial({{\cal T}})^{{\cal E}}$. We have ${\mathfg{P}}_\cekta(\zeta^{{\cal E}}=0)=1$ akd so ${\mathbb{E | T}})^{{\cal E}}$, a similar proof as the (\[50e\]) that $$\lim\limits_{n\to \eta^{-n} {\mathbb{P}}_\delta(\zeta>n)= {\mathbb{E}}_\delta(\eta^{-\zeta^{{\cal the (\[50b\]) is satisfied. let ${\mathbb{P}}_\delta(\zeta^{{\cal E}}<\infty)=0$. ${\mathbb{E}}_\delta(\eta^{-\zeta^{{\cal E}}}, \zeta^{{\cal E}}<\infty)=0$ and in we have ${{\mathbb{E}}_\delta(\eta^{-\zeta^{{\cal E}}}, \zeta^{{\cal E}}<\infty)}/ {{\mathbb{E}}(\eta^{-\zeta^{{\cal E}}}, \zeta^{{\cal E}}<\infty)}=0$. We claim that this case we also have $\lim\limits_{n\to \infty} {\mathbb{P}}_\delta(\zeta>n)/{\mathbb{P}}(\zeta>n)=0$. In fact, ${\mathbb{P}}_\delta(\zeta^{{\cal E}}<\infty)=0$ implies (\beta_0+\theta)^{-n}{\mathbb{P}}_\delta(\zeta>n)&=& \zeta^{{\cal &=&(\beta_0+\theta)^{-n} j\le n: Y_j\not\in (\partial(T)^{{\cal E}}\cup \{{{\cal D}}^\rho\})<\infty. \end{aligned}$$ Since $\lim\limits_{n\to \infty}\eta^{-n}{\mathbb{P}}(\zeta>n)>0$ and $\beta_0+\theta<\eta$, the claim follows (\[50b\]) is shown. Now, let $P^*$ be the of $P$ to $\partial({{\cal The last statement we must is the vector defined (\[rev1\]) a right eigenvector $P^*$ with eigenvalue $\eta$. First take $\delta\in \partial({{\cal T}})^{{\cal E}}$. We have ${\mathbb{P}}_\delta(\zeta^{{\cal E}}=0)=1$ and so ${\mathbb{E | T}})^{{\cal E}}$, a similar proof as the onE showing (\[50e\]) gIves tHat $$\Lim\LiMits_{N\to \iNfty} \eta^{-n}
{\mathbb{p}}_\DeltA(\zeta>n)=
{\mathbb{E}}_\delta(\eta^{-\zEta^{{\caL E}}},\ZEta^{{\cAL E}}<\Infty)\In (0,\infty),$$ ANd SO The ReLaTioN (\[50b\]) IS sAtisfIed. now, let ${\mAthbb{P}}_\deltA(\zeTa^{{\Cal E}}<\infty)=0$. TheN, ${\MaThbb{E}}_\delta(\Eta^{-\Zeta^{{\cal E}}}, \zeta^{{\Cal e}}<\infty)=0$ AnD in (\[50B\]) We havE ${{\maThbb{E}}_\Delta(\eTA^{-\zeta^{{\cAl E}}}, \zeta^{{\caL E}}<\INfty)}/
{{\maTHbb{E}}(\eta^{-\ZETa^{{\Cal E}}}, \Zeta^{{\cal E}}<\infty)}=0$. We clAIm THat in this case wE also hAvE $\LiM\LImiTs_{n\To \infty} {\matHbB{P}}_\delTA(\zeta>n)/{\mAThBB{p}}(\ZetA>N)=0$. In fact, ${\mathbb{p}}_\delta(\zeta^{{\cAL E}}<\iNfty)=0$ imPlIes $$\BEgin{alIgned}
(\BeTA_0+\thEta)^{-n}{\mathbb{P}}_\DeltA(\zeta>n)&=&
(\betA_0+\theta)^{-N}{\Mathbb{P}}_\DElta(\zetA>n, \zeta^{{\Cal e}}>n)\\
&=&(\bEta_0+\tHEtA)^{-n} {\MatHbB{p}}(\foRAlL j\lE N:
Y_j\Not\in (\parTiAl(t)^{{\cal E}}\Cup \{{{\cAL d}}^\RHo\})<\inFty. \End{aLigneD}$$ Since $\lim\limiTs_{n\To \inFTy}\eTa^{-n}{\maThbb{P}}(\Zeta>N)>0$ aNd $\betA_0+\theta<\Eta$, thE cLaim follows and (\[50b\]) Is shOwn.
Now, let $p^*$ be ThE reStRictiON of $P$ to $\ParTiaL({{\cal T}})^*$. ThE last stATemEnT WE MuSt show is that the vecToR $\VArPhi$ definEd in (\[reV1\]) Is A rIGht eigenVeCtoR of $P^*$ WITh eigEnvaLUe $\Eta$. First Take $\deLTa\In \Partial({{\CaL T}})^{{\cal E}}$. we HavE ${\maThbb{P}}_\DElta(\Zeta^{{\caL E}}=0)=1$ and so ${\mAthbb{e | T}})^{{\cal E}}$, a simil ar proof a s the on e s ho wing (\[ 50e\]) gives t h at $ $\lim\limits_{n\to \in fty}\e t a^{- n }{\mat hbb{P}} _ \d e l ta( \z et a>n )= {\ mathb b{E }}_\del ta(\eta^{- \ze ta ^{{\cal E}}} , \z eta^{{\cal E} }<\infty)\in (0 ,\inft y) ,$$ and s o t he re lation (\[50b \]) is sa ti s fied.N ow, let $ {\ math bb{P}}_\delta(\ze t a^ { {\cal E}}<\inf ty)=0$ .T he n , ${ \ma thbb{E}}_\ de lta(\ e ta^{-\z e ta ^ { { \ca l E}}}, \zeta^ {{\cal E}}< \ inf ty)=0$ a ndi n (\[5 0b\]) w e ha ve ${{\math bb{E }}_\delta (\eta^ { -\zeta^ { {\cal E }}}, \ zet a^{ {\ca l E }} <\i nf t y)} /
{ {\m a thb b{E}}(\e ta ^{ -\zet a^{{ \ c a l E}} },\zet a^{{\ cal E}}<\inft y)} =0$. Weclaim that inth is ca se wealsoha ve $\lim\limits _{n\ to \infty } { \m ath bb {P}}_ \ delta( \ze ta> n)/{\ma thbb{P} } (\z et a > n )= 0$. In fact, ${\ma th b b {P }}_\delt a(\zet a ^{ {\ c al E}}<\ in fty )=0$ i mplie s $$ \ be gin{alig ned}
( \ be ta _0+\the ta )^{-n} {\ mat hbb {P}}_ \ delt a(\zet a>n)&=&(\bet a _0+\theta)^{-n } {\mathbb{P}}_ \ de l t a( \ zeta >n, \zeta^{{\c al E } }>n) \\
& = &( \be t a_0+\ theta )^ { -n } {\mathbb{P}}(\fora ll j\len:
Y _j\not\in (\p artial(T)^ { { \ cal E}}\ cup\ {{ { \cal D}}^\rho\ })<\i nfty. \end { aligned} $$ Si nce $\li m\limits_ { n \to \inf ty} \et a^{ -n} { \ ma thbb{P}}(\zet a > n)>0 $and $\b eta _0+\the ta< \et a$, th eclaim fol lows and ( \[ 50 b\ ])is sh o wn.
Now ,let $ P^* $ bet he res trict ionof $ P $ t o $\par t ia l ( {{\c al T }})^ *$. T he la st s t ate ment we must sho w i s tha tth e vecto r $\varphi$ d ef ined in (\ [r ev1 \]) is a right e igenvector of $P^*$ wit h eigenv alu e $\e ta$. First ta ke$\delt a\i n \part ial({{ \calT} })^ { { \calE } }$ . W ehave ${\ma t h bb{ P}}_\ de lta( \zeta^{ {\cal E}}=0)=1$ an d so ${\mathbb{E | T}})^{{\cal_E}}$, a_similar proof as the_one showing_(\[50e\])_gives that_$$\lim\limits_{n\to_\infty} \eta^{-n}
{\mathbb{P}}_\delta(\zeta>n)=
{\mathbb{E}}_\delta(\eta^{-\zeta^{{\cal E}}},\zeta^{{\cal_E}}<\infty)\in (0,\infty),$$ and_so the relation (\[50b\])_is satisfied. Now,_let_${\mathbb{P}}_\delta(\zeta^{{\cal E}}<\infty)=0$. Then, ${\mathbb{E}}_\delta(\eta^{-\zeta^{{\cal E}}}, \zeta^{{\cal E}}<\infty)=0$ and in (\[50b\]) we have ${{\mathbb{E}}_\delta(\eta^{-\zeta^{{\cal E}}},_\zeta^{{\cal_E}}<\infty)}/
{{\mathbb{E}}(\eta^{-\zeta^{{\cal E}}},_\zeta^{{\cal_E}}<\infty)}=0$._We claim that in this_case we also have $\lim\limits_{n\to_\infty} {\mathbb{P}}_\delta(\zeta>n)/{\mathbb{P}}(\zeta>n)=0$._In fact, ${\mathbb{P}}_\delta(\zeta^{{\cal E}}<\infty)=0$ implies $$\begin{aligned}
(\beta_0+\theta)^{-n}{\mathbb{P}}_\delta(\zeta>n)&=&
(\beta_0+\theta)^{-n}{\mathbb{P}}_\delta(\zeta>n, \zeta^{{\cal E}}>n)\\
&=&(\beta_0+\theta)^{-n}_{\mathbb{P}}(\forall_j\le n:
Y_j\not\in_(\partial(T)^{{\cal E}}\cup \{{{\cal D}}^\rho\})<\infty. \end{aligned}$$ Since $\lim\limits_{n\to \infty}\eta^{-n}{\mathbb{P}}(\zeta>n)>0$ and_$\beta_0+\theta<\eta$, the claim follows and (\[50b\])_is shown.
Now, let_$P^*$_be_the restriction of $P$_to $\partial({{\cal T}})^*$. The last statement_we must show is that the_vector $\varphi$ defined in (\[rev1\]) is a_right eigenvector of $P^*$ with eigenvalue_$\eta$. First take $\delta\in \partial({{\cal_T}})^{{\cal E}}$._We have ${\mathbb{P}}_\delta(\zeta^{{\cal E}}=0)=1$ and_so ${\mathbb{E |
: {#second-brittle-regime.unnumbered}
The strength gain in the second regime, albeit slower, can again be approximated by an exponential function for about one order of magnitude of strength in the observed time span (Fig.7 and coefficients in Tab.4). The exponents of the evolution law for the punch, compression and tension tests are similar, implying that in this regime the ratio between compressive and tensile strength remains roughly constant (at 5.4 in this case). The shear strength exhibits a significant amount of scatter due to frictional forces in the test setup and thus remains inconclusive. Of course a further slowdown in strength gain is anticipated for all tests until the concrete is fully hydrated and strength converges to a final value. This trend is indicated by the four-point bending test data (see Fig.7a), which is qualitatively related to the tensile strength.
It is important to observe that a very large portion of the well-known difference in compressive and tensile strength of hardened concrete emerges already at the early transition from a pressure-insensitive thixotropic yield stress fluid to a brittle cohesive frictional solid. It is not a putative difference in rate (or exponent) of compressive and tensile strength growth that gives rise to the pressure sensitivity of the final material, but rather their different times of transition from the fluid to the solid regime. The difference in “brittle” compression and tension may always be existent, except that in fresh concrete the low plastic yield strength undercuts these failure modes. The solid regime thus emerges only after the plastic strength, growing at a higher rate, surpasses brittle strength – a transition which occurs at different times for compression and tension because of the inherent difference in brittle compressive and tensile strength.
\[tab1\]
------------------------ ----------------------------- ---------------------- ------- -----
$a$ \[kPa\] $b$ \[min$^{-1}$\] $R^2$ $N$
(test (b)) Punch 4.0 \[-1.08,+1.48\] 0.012 \[$\pm$0.001\] 0.967 27
(test (c)) Shear 13.9 \[-9.33,+28.29\] 0.005 \[$\pm$0.0044\] 0.381 13
(test (d)) Compression 3.7 \[-0.91,+1.22\] | : { # second-brittle-regime.unnumbered }
The strength gain in the second regime, albeit slower, can again be approximate by an exponential affair for about one order of magnitude of force in the observed time couple (Fig.7 and coefficients in Tab.4). The exponents of the development jurisprudence for the punch, compression and latent hostility tests are similar, implying that in this regime the proportion between compressive and tensile strength remain roughly constant (at 5.4 in this event). The shear strength exhibits a significant sum of scatter due to frictional forces in the test setup and therefore remains inconclusive. Of naturally a further slowdown in strength gain is anticipated for all tests until the concrete is in full hydrated and strength converges to a final value. This trend is indicated by the four - distributor point bending test data (see Fig.7a), which is qualitatively related to the tensile strength.
It is important to observe that a very large part of the well - known difference in compressive and ductile forte of temper concrete emerges already at the early transition from a pressure - insensitive thixotropic yield stress fluid to a brittle cohesive frictional solid. It is not a putative difference in pace (or exponent) of compressive and tensile strength growth that gives rise to the pressure sensitivity of the final material, but rather their different times of transition from the fluid to the upstanding regime. The deviation in “ brittle ” compression and tension may always be existent, except that in fresh concrete the gloomy plastic yield strength undercuts these failure modes. The solid regime thus emerge only after the credit card strength, growing at a higher rate, surpasses brittle military capability – a transition which happen at different clock time for compression and tension because of the inherent difference in brittle compressive and tensile military capability.
\[tab1\ ]
------------------------ ----------------------------- ---------------------- ------- -----
$ a$ \[kPa\ ] $ b$ \[min$^{-1}$\ ] $ R^2 $ $ N$
(test (b-complex vitamin) ) Punch 4.0 \[-1.08,+1.48\ ] 0.012 \[$\pm$0.001\ ] 0.967 27
(test (c) ) Shear 13.9 \[-9.33,+28.29\ ] 0.005 \[$\pm$0.0044\ ] 0.381 13
(test (d) ) Compression 3.7 \[-0.91,+1.22\ ] | : {#sefond-brittle-regime.unnumbeved}
The strength yqin in the sscond reeime, albeit slower, can again bw appeoximated by an exponevtial funbtion for abont one order of magnitudc of abrengch in the observgd time span (Fig.7 and coeffhckeuts in Tab.4). The exponents of the evolttion lsw for the punch, comkrefsioh and tension tests are similar, imllying uhat in this regimr the ratio between compredsivf and tensile strejgth remainw rotthly constang (at 5.4 in this case). The shear strength exhibits a signkficaut amount od wcahjer due to fcictiogal forces ik the tast setip and thus rekaiis ibconclusive. Of course a further slowdowg in stretgch gain is anticipatee dor anl tasts yntkl uhe ckncretf ia fully hysrated and wtrength converges uo w final value. Fhis twegd is indicated by the four-point bendinc tsst data (see Fig.7a), which is qualitatively relwted to tre tensile strength.
It is important to observe thad a vxrh lcvne puetlon of the well-known difference in compressivq amd tensile strennth of hardened comcgeyg emerges alrexdy at ths early transition from a preswure-insenfitife thixotropic yield stress fluid to a yrirtle cohesive friccional solid. It os noy a putative difference in date (or explnent) of dumpressive and tdnsple vtrength growth that gives rise to vhe pxessure rensotivitr of the flnal material, but rather theit diffarent timed of transition from the fluid vp the solid teghme. The difyerencc in “brittle” coipression and jension mcy alwxys be exiatent, eecept that ig fresh concraje the low plestic yiejd srrentth unddfcuts these faolure modvs. The solid regime thus emgrfes only after cke plastic strengyh, erorijg ae a higher rade, sjrpxxses crittle strcnggh – s transition which owcura at different timrs for comkression wnd tension brcause of the inhegent vifferxnce im btittle compressive and tensile atrength.
\[twb1\]
------------------------ ----------------------------- ---------------------- ------- -----
$a$ \[kPa\] $b$ \[min$^{-1}$\] $R^2$ $N$
(test (b)) Punch 4.0 \[-1.08,+1.48\] 0.012 \[$\pm$0.001\] 0.967 27
(test (c)) Shear 13.9 \[-9.33,+28.29\] 0.005 \[$\pm$0.0044\] 0.381 13
(test (d)) Cokpression 3.7 \[-0.91,+1.22\] | : {#second-brittle-regime.unnumbered} The strength gain in the albeit can again approximated by an order magnitude of strength the observed time (Fig.7 and coefficients in Tab.4). The of the evolution law for the punch, compression and tension tests are similar, that in this regime the ratio between compressive and tensile strength remains roughly (at in case). shear strength exhibits a significant amount of scatter due to frictional forces in the test setup thus remains inconclusive. Of course a further slowdown strength gain is anticipated all tests until the concrete fully and strength to final This trend is by the four-point bending test data (see Fig.7a), which is qualitatively related to the tensile strength. It important to a very portion the difference in compressive strength of hardened concrete emerges already transition from a pressure-insensitive thixotropic yield stress fluid a brittle frictional solid. It is not a difference in rate (or exponent) of compressive and strength growth that gives rise to the pressure sensitivity of the final material, but rather times of transition from fluid to the regime. difference “brittle” and tension always be existent, except that in fresh concrete the low plastic strength undercuts these failure modes. The solid regime thus emerges the strength, growing at higher rate, surpasses brittle – transition which occurs at for and the difference brittle compressive and tensile \[tab1\] ------------------------ ----------------------------- ---------------------- ------- $a$ \[kPa\] $b$ \[min$^{-1}$\] 4.0 \[-1.08,+1.48\] 0.012 \[$\pm$0.001\] 0.967 27 (test (c)) 13.9 \[-9.33,+28.29\] 0.005 \[$\pm$0.0044\] 0.381 13 (test Compression 3.7 \[-0.91,+1.22\] | : {#second-brittle-regime.unnumbEred}
The strEngth GaiN in ThE secOnd rEgime, albeit sloWEr, caN again be approximated by An expOnENtiaL FuNctioN for aboUT oNE OrdEr Of MagNiTUdE of stRenGth in thE observed tIme SpAn (Fig.7 and coefFIcIents in Tab.4). the Exponents of tHe eVolutiOn Law FOr the PunCh, comPressiON and teNsion testS aRE similAR, implyiNG ThAt in This regime the ratiO BeTWeen compressivE and teNsILe STRenGth Remains rouGhLy conSTant (at 5.4 iN ThIS CAse). tHe shear strengTh exhibits a SIgnIficanT aMouNT of scaTter dUe TO frIctional forCes iN the test sEtup anD Thus remAIns incoNclusiVe. OF coUrse A FuRtHer SlOWdoWN iN stREngTh gain is AnTiCipatEd foR ALL TestS unTil tHe conCrete is fully hYdrAted ANd sTrengTh conVergEs To a fiNal valUe. ThiS tRend is indicated By thE four-poinT beNdIng TeSt datA (See Fig.7A), whIch Is qualiTativelY RelAtED TO tHe tensile strength.
IT iS IMpOrtant to ObservE ThAt A Very largE pOrtIon oF THe welL-knoWN dIfferencE in comPReSsIve and tEnSile stReNgtH of HardeNEd coNcrete Emerges aLreadY At the early tranSItion from a preSSuRE-InSEnsiTivE thixotropiC yieLD strEss fLUiD to A BrittLe cohEsIVe FRictional solid. It is nOt A putatIve diFference in ratE (or exponenT) OF CompressIve aND tENsile strength gRowth That gives rISe to the pRessuRe sensitIvity of thE FInal mateRiaL, buT raTheR THeIr different tiMES of tRaNsition FroM the fluId tO thE soLid ReGime. The diFference In “BrItTlE” coMpresSIon and teNsIon MaY alWays bE ExisteNt, excEpt tHaT iN FreSh concrETe THE low PlAsTic yIelD sTrengTh unDErcUts thesE failure mOdeS. the sOlId Regime tHus emerges onlY aFter the plaStIc sTrengtH, GRowing at A higher rate, surpasses briTTle streNgtH – a traNsitIon which oCcuRs at diFfeREnt timEs for cOmpreSsIon AND tensION bEcaUsE of the inheRENt dIfferEnCe in Brittle Compressive and tensILe sTrength.
\[tab1\]
------------------------ ----------------------------- ---------------------- ------- -----
$a$ \[kPA\] $b$ \[mIn$^{-1}$\] $R^2$ $N$
(TESt (B)) PuNCh 4.0 \[-1.08,+1.48\] 0.012 \[$\PM$0.001\] 0.967 27
(teSt (C)) sheAR 13.9 \[-9.33,+28.29\] 0.005 \[$\Pm$0.0044\] 0.381 13
(test (d)) CompressIon 3.7 \[-0.91,+1.22\] | : {#second-brittle-regime. unnumbered }
Th e s tre ng th g ainin the secondr egim e, albeit slower, canagain b e app r ox imate d by an ex p o nen ti al fu nc t io n for ab out one order ofmag ni tude of stre n gt h in the o bse rved time sp an(Fig.7 a ndc oeffi cie nts i n Tab. 4 ). The exponent so f thee volutio n la w fo r the punch, comp r es s ion and tensio n test sa re s imi lar , implying t hat i n this r e gi m e the ratio between compressiv e an d tens il e s t rength rema in s ro ughly const ant(at 5.4 i n this case).T he shea r stre ngt h e xhib i ts a si gn i fic a nt am o unt of scat te rdue t o fr i c t i onal fo rces in t he test setup an d th u s r emain s inc oncl us ive.Of cou rse a f urther slowdown instrengthgai nisan ticip a ted fo r a lltests u ntil th e co nc r e t eis fully hydratedan d st rength c onverg e sto a finalva lue . Th i s tren d is in dicatedby the fo ur -pointbe ndingte stdat a (se e Fig .7a),which is qual i tatively relat e d to the tens i le s tr e ngth .
It is impor tant to o bser v etha t a ve ry la rg e p o rtion of the well-k no wn dif feren ce in compres sive and t e n s ile stre ngth of hardened concr ete e merges alr e ady at t he ea rly tran sition fr o m a press ure -in sen sit i v ethixotropic y i e ld s tr ess flu idto a br itt lecoh esi ve friction al solid .It i snot a pu t ative di ff ere nc e i n rat e (or e xpone nt)of c o mpr essivea nd t ensi le s tren gth g rowth tha t gi ves ris e to thepre s sure s en sitivit y of the fina lmaterial,bu t r athert h eir diff erent times of transiti o n fromthe flui d to the soli d r egime. Th e diffe rencein “b ri ttl e ” comp r e ss ion a nd tension m ayalway sbe e xistent , except that in f r esh concrete the lo w pl a s ti c y i el d st re n gth u ndercuts thesefailure mo de s .The solidr egi me thus e mergesonlya fter th e plastic strength ,grow i n g a t a higher rate, s urpassesb rittl e s treng th– a tr an sit ion w hich o c cur s atdiffer en t time s for c ompressi on and tension becauseof the inhe ren t differe nce inbrittle c ompr essive and te nsi le st ren g th.
\[ta b 1\ ]
---- ---- - --------- - -- --- - -- ----------- - - - --- ----- --- - ----- ---- ------------- --- - --- -----
$a$ \[kPa\ ] $b$ \[mi n$^{-1}$ \] $ R^2$ $N$
(test( b )) Punch 4.0 \[-1 .08 ,+ 1 .48\] 0 .0 12 \[ $\pm $0 .001\] 0. 9 6 7 27
(test ( c)) S h e ar 13. 9\[-9.33 , +28. 29\] 0.005 \[$\pm$0 .0044\ ] 0.381 1 3
(t est ( d)) Compre s sion 3. 7 \[- 0.91,+1 .2 2\] | : {#second-brittle-regime.unnumbered}
The_strength gain_in the second regime,_albeit slower,_can_again be_approximated_by an exponential_function for about_one order of magnitude_of strength in_the_observed time span (Fig.7 and coefficients in Tab.4). The exponents of the evolution law_for_the punch,_compression_and_tension tests are similar, implying_that in this regime the_ratio between_compressive and tensile strength remains roughly constant (at_5.4_in this case)._The shear strength exhibits a significant amount of scatter_due to frictional forces in the_test setup and_thus_remains_inconclusive. Of course a_further slowdown in strength gain is_anticipated for all tests until the_concrete is fully hydrated and strength converges_to a final value. This trend_is indicated by the four-point_bending test_data (see Fig.7a), which is_qualitatively related to_the tensile_strength.
It is important_to observe that a very large_portion of the_well-known difference in compressive and tensile_strength_of hardened concrete_emerges_already_at the_early transition from_a_pressure-insensitive thixotropic_yield_stress fluid to a brittle cohesive_frictional_solid. It is not a putative difference_in rate (or exponent)_of_compressive and tensile strength_growth that gives rise to_the pressure sensitivity of the final_material, but_rather their_different times of transition from the fluid to the solid regime._The difference in “brittle” compression and_tension may always be_existent, except_that_in fresh concrete_the_low plastic_yield strength undercuts these failure modes. The_solid regime_thus emerges only after the plastic_strength, growing at a_higher_rate, surpasses brittle strength – a_transition which occurs at different times_for compression and tension because_of_the_inherent difference in brittle compressive_and tensile strength.
\[tab1\]
------------------------ -----------------------------_---------------------- ------- -----
_ __ __ ___ $a$ \[kPa\]_ _ _ _$b$ \[min$^{-1}$\] __$R^2$_ $N$
(test (b)) Punch _ _ _4.0 \[-1.08,+1.48\] 0.012 \[$\pm$0.001\] _ 0.967_ 27
(test (c)) Shear _ 13.9 \[-9.33,+28.29\] 0.005 _ \[$\pm$0.0044\] _ 0.381 __13
(test (d)) Compression _3.7_\[-0.91,+1.22\]__ _ _ _ |
ordering operation can be used without affecting the validity of the proof.\
\
Suppose that the cardinality of the attributes after reordering are $\lambda_i$ where $i \in [0,d-1]$, where d is the dimensionality of the dataset. $$\begin{aligned}
\sum_{x \in S}\psi_j(x)&=\exp \Bigl(\frac{2{\pi}i{j_0}0}{\lambda_0} \Bigr) \times \exp \Bigl(\frac{2{\pi}i{j_1}{x_1}}{\lambda_1} \Bigr) \times \cdots \times \exp \Bigl(\frac{2{\pi}i{j_{d-1}}x_{d-1}} {\lambda_{d-1}}\Bigr) \nonumber \\
&+\exp\Bigl(\frac{2{\pi}i{j_0}1}{\lambda_0}\Bigr) \times \exp\Bigl(\frac{2{\pi}i{j_1}{x_1}}{\lambda_1}\Bigr) \times \cdots \times \exp\Bigl(\frac{2{\pi}i{j_{d-1}}{x_{d-1}}}{\lambda_{d-1}}\Bigr) \nonumber \\
&\vdots \nonumber \\
&+\exp\Bigl(\frac{2{\pi}i{j_0}(\lambda_0-1)}{\lambda_0}\Bigr) \times \exp\Bigl(\frac{2{\pi}i{j_1}{x_1}}{\lambda_1}\Bigr) \times \cdots \times \exp\Bigl(\frac{2{\pi}i{j_{d-1}}{x_{d-1}}}{\lambda_{d-1}}\Bigr) \nonumber \\
&= \exp \Bigl(\frac{2{\pi}i{j_1}{x_1}}{\lambda_1} \Bigr) \times \cdots \times \exp \Bigl(\frac{2{\pi}i{j_{d-1}}x_{d-1}} {\lambda_{d-1}}\Bigr)\sum_{k=0}^{\lambda_0-1}\Bigl(\exp \Bigl(\frac{2{\pi}i{j_0}k}{\lambda_0} \Bigr)\Bigr) \nonumber \\
%&= \exp \Bigl(\frac{2{\pi}i{j_1 | ordering operation can be used without affecting the cogency of the proof.\
\
presuppose that the cardinality of the attributes after reordering are $ \lambda_i$ where $ i \in [ 0,d-1]$, where five hundred is the dimensionality of the dataset. $ $ \begin{aligned }
\sum_{x \in S}\psi_j(x)&=\exp \Bigl(\frac{2{\pi}i{j_0}0}{\lambda_0 } \Bigr) \times \exp \Bigl(\frac{2{\pi}i{j_1}{x_1}}{\lambda_1 } \Bigr) \times \cdots \times \exp \Bigl(\frac{2{\pi}i{j_{d-1}}x_{d-1 } } { \lambda_{d-1}}\Bigr) \nonumber \\
& + \exp\Bigl(\frac{2{\pi}i{j_0}1}{\lambda_0}\Bigr) \times \exp\Bigl(\frac{2{\pi}i{j_1}{x_1}}{\lambda_1}\Bigr) \times \cdots \times \exp\Bigl(\frac{2{\pi}i{j_{d-1}}{x_{d-1}}}{\lambda_{d-1}}\Bigr) \nonumber \\
& \vdots \nonumber \\
& + \exp\Bigl(\frac{2{\pi}i{j_0}(\lambda_0 - 1)}{\lambda_0}\Bigr) \times \exp\Bigl(\frac{2{\pi}i{j_1}{x_1}}{\lambda_1}\Bigr) \times \cdots \times \exp\Bigl(\frac{2{\pi}i{j_{d-1}}{x_{d-1}}}{\lambda_{d-1}}\Bigr) \nonumber \\
& = \exp \Bigl(\frac{2{\pi}i{j_1}{x_1}}{\lambda_1 } \Bigr) \times \cdots \times \exp \Bigl(\frac{2{\pi}i{j_{d-1}}x_{d-1 } } { \lambda_{d-1}}\Bigr)\sum_{k=0}^{\lambda_0 - 1}\Bigl(\exp \Bigl(\frac{2{\pi}i{j_0}k}{\lambda_0 } \Bigr)\Bigr) \nonumber \\
% & = \exp \Bigl(\frac{2{\pi}i{j_1 | ordfring operation can be ured without affgcring tie valisity of ghe proof.\
\
Suppose that the cacdinqlity of the attributes aftdr reordeging are $\oambva_i$ where $i \in [0,d-1]$, where d is ths dimznwionality of tme dataset. $$\tegin{aligned}
\sgm_{b \nn S}\psi_j(x)&=\exp \Bigl(\frac{2{\pi}i{j_0}0}{\lambda_0} \Bigr) \times \rxo \Bigl(\frac{2{\pi}i{j_1}{v_1}}{\lamnqa_1} \Bjgr) \times \cdots \times \exp \Bigl(\frac{2{\li}i{j_{d-1}}x_{d-1}} {\lambda_{d-1}}\Bigr) \monumber \\
&+\exp\Bigl(\frac{2{\pi}i{j_0}1}{\lalbda_0}\Higr) \times \exp\Bigl(\vrac{2{\pi}i{j_1}{x_1}}{\lanbda_1}\Fugr) \times \cduts \times \tx'\Bigl(\frac{2{\pi}j{j_{d-1}}{x_{d-1}}}{\lambda_{d-1}}\Bigr) \nonumber \\
&\vdots \nonukber \\
&+\exp\Bitl(\draf{2{\ki}i{j_0}(\lambda_0-1)}{\lanbda_0}\Bpgr) \times \exp\Npgl(\frac{2{\[i}i{j_1}{x_1}}{\lakbda_1}\Bigr) \times \cdmts \times \exp\Bigl(\frac{2{\pi}i{o_{d-1}}{x_{d-1}}}{\lambda_{d-1}}\Bigr) \nonuiber \\
&= \exp \Bngl(\frac{2{\pi}i{j_1}{x_1}}{\lambda_1} \Bitr) \timev \cdmts \gumer \esp \Bjgl(\fraf{2{\pi}m{j_{d-1}}x_{d-1}} {\lambdz_{d-1}}\Bigr)\sum_{k=0}^{\lqmbda_0-1}\Bigl(\exp \Bigl(\frsc{2{\[p}o{j_0}k}{\lambda_0} \Bifr)\Bigr) \njnumber \\
%&= \exp \Bigl(\frac{2{\pi}i{j_1 | ordering operation can be used without affecting of proof.\ \ that the cardinality are where $i \in where d is dimensionality of the dataset. $$\begin{aligned} \sum_{x S}\psi_j(x)&=\exp \Bigl(\frac{2{\pi}i{j_0}0}{\lambda_0} \Bigr) \times \exp \Bigl(\frac{2{\pi}i{j_1}{x_1}}{\lambda_1} \Bigr) \times \cdots \times \exp \Bigl(\frac{2{\pi}i{j_{d-1}}x_{d-1}} {\lambda_{d-1}}\Bigr) \\ &+\exp\Bigl(\frac{2{\pi}i{j_0}1}{\lambda_0}\Bigr) \times \exp\Bigl(\frac{2{\pi}i{j_1}{x_1}}{\lambda_1}\Bigr) \times \cdots \times \exp\Bigl(\frac{2{\pi}i{j_{d-1}}{x_{d-1}}}{\lambda_{d-1}}\Bigr) \nonumber \\ &\vdots \nonumber \\ \times \times \times \nonumber \\ &= \exp \Bigl(\frac{2{\pi}i{j_1}{x_1}}{\lambda_1} \Bigr) \times \cdots \times \exp \Bigl(\frac{2{\pi}i{j_{d-1}}x_{d-1}} {\lambda_{d-1}}\Bigr)\sum_{k=0}^{\lambda_0-1}\Bigl(\exp \Bigl(\frac{2{\pi}i{j_0}k}{\lambda_0} \Bigr)\Bigr) \nonumber \\ \exp \Bigl(\frac{2{\pi}i{j_1 | ordering operation can be useD without afFectiNg tHe vAlIditY of tHe proof.\
\
Suppose THat tHe cardinality of the attrIbuteS aFTer rEOrDerinG are $\lamBDa_I$ WHerE $i \In [0,D-1]$, whErE D iS the dImeNsionalIty of the daTasEt. $$\Begin{aligned}
\SUm_{X \in S}\psi_j(x)&=\eXp \BIgl(\frac{2{\pi}i{j_0}0}{\lAmbDa_0} \Bigr) \TiMes \EXp \BigL(\frAc{2{\pi}i{J_1}{x_1}}{\lambDA_1} \Bigr) \tImes \cdots \TiMEs \exp \BIGl(\frac{2{\pI}I{J_{d-1}}X_{d-1}} {\laMbda_{d-1}}\Bigr) \nonumber \\
&+\EXp\bIgl(\frac{2{\pi}i{j_0}1}{\lamBda_0}\BigR) \tIMeS \EXp\BIgl(\Frac{2{\pi}i{j_1}{x_1}}{\lAmBda_1}\BiGR) \times \cDOtS \TIMes \EXp\Bigl(\frac{2{\pi}i{J_{d-1}}{x_{d-1}}}{\lambda_{d-1}}\bIgr) \NonumbEr \\
&\VdoTS \nonumBer \\
&+\exP\BIGl(\fRac{2{\pi}i{j_0}(\lambDa_0-1)}{\laMbda_0}\Bigr) \tImes \exP\bigl(\fraC{2{\Pi}i{j_1}{x_1}}{\laMbda_1}\BiGr) \tImeS \cdoTS \tImEs \eXp\bIgl(\FRaC{2{\pi}I{J_{d-1}}{x_{D-1}}}{\lambda_{d-1}}\biGr) \NonumBer \\
&= \eXP \bIGl(\frAc{2{\pI}i{j_1}{x_1}}{\LambdA_1} \Bigr) \times \cdoTs \tImes \EXp \BIgl(\frAc{2{\pi}i{J_{d-1}}x_{d-1}} {\LaMbda_{d-1}}\bigr)\suM_{k=0}^{\lamBdA_0-1}\Bigl(\exp \Bigl(\fraC{2{\pi}i{J_0}k}{\lambda_0} \BIgr)\biGr) \nOnUmber \\
%&= \EXp \Bigl(\FraC{2{\pi}I{j_1 | ordering operation can beused witho ut af fec tin gthevali dity of the pr o of.\
\
Suppose that the ca rdina li t y of th e att ributes af t e r r eo rd eri ng ar e $\l amb da_i$ w here $i \i n [ 0, d-1]$, where dis the dim ens ionality ofthe datas et . $ $ \begi n{a ligne d}
\s u m_{x \ in S}\psi _j ( x)&=\e x p \Bigl ( \ fr ac{2 {\pi}i{j_0}0}{\la m bd a _0} \Bigr) \ti mes \e xp \B i g l(\ fra c{2{\pi}i{ j_ 1}{x_ 1 }}{\lam b da _ 1 } \B i gr) \times \c dots \times \ex p \Big l( \fr a c{2{\p i}i{j _{ d -1} }x_{d-1}} { \lam bda_{d-1} }\Bigr ) \nonum b er \\
& +\exp\ Big l(\ frac { 2{ \p i}i {j _ 0}1 } {\ lam b da_ 0}\Bigr) \ ti mes \ exp\ B i g l (\fr ac{ 2{\p i}i{j _1}{x_1}}{\la mbd a_1} \ Big r) \t imes\cdo ts \tim es \ex p\Big l( \frac{2{\pi}i{j _{d- 1}}{x_{d- 1}} }{ \la mb da_{d - 1}}\Bi gr) \n onumber \\
&\v d ots \ n o n um ber \\
&+\exp\Bigl (\ f r ac {2{\pi}i {j_0}( \ la mb d a_0-1)}{ \l amb da_0 } \ Bigr) \ti m es \exp\Bi gl(\fr a c{ 2{ \pi}i{j _1 }{x_1} }{ \la mbd a_1}\ B igr) \time s \cdots \tim e s \exp\Bigl(\f r ac{2{\pi}i{j_ { d- 1 } }{ x _{d- 1}} }{\lambda_{ d-1} } \Big r) \ n on umb e r \\&= \e xp \B i gl(\frac{2{\pi}i{j_ 1} {x_1}} {\lam bda_1} \Bigr) \times \c d o t s \times \ex p \ B igl(\frac{2{\p i}i{j _{d-1}}x_{ d -1}} {\l ambda _{d-1}}\ Bigr)\sum _ { k=0}^{\l amb da_ 0-1 }\B i g l( \exp \Bigl(\f r a c{2{ \p i}i{j_0 }k} {\lambd a_0 } \ Big r)\ Bi gr) \nonu mber \\%& =\e xp \B igl(\ f rac{2{\p i} i{j _1 | ordering operation_can be_used without affecting the_validity of_the_proof.\
\
Suppose that_the_cardinality of the_attributes after reordering_are $\lambda_i$ where $i_\in [0,d-1]$, where_d_is the dimensionality of the dataset. $$\begin{aligned}
\sum_{x \in S}\psi_j(x)&=\exp \Bigl(\frac{2{\pi}i{j_0}0}{\lambda_0} \Bigr) \times \exp_\Bigl(\frac{2{\pi}i{j_1}{x_1}}{\lambda_1}_\Bigr) \times_\cdots_\times_\exp \Bigl(\frac{2{\pi}i{j_{d-1}}x_{d-1}} {\lambda_{d-1}}\Bigr) \nonumber \\
&+\exp\Bigl(\frac{2{\pi}i{j_0}1}{\lambda_0}\Bigr)_\times \exp\Bigl(\frac{2{\pi}i{j_1}{x_1}}{\lambda_1}\Bigr) \times \cdots \times_\exp\Bigl(\frac{2{\pi}i{j_{d-1}}{x_{d-1}}}{\lambda_{d-1}}\Bigr) \nonumber_\\
&\vdots \nonumber \\
&+\exp\Bigl(\frac{2{\pi}i{j_0}(\lambda_0-1)}{\lambda_0}\Bigr) \times \exp\Bigl(\frac{2{\pi}i{j_1}{x_1}}{\lambda_1}\Bigr) \times \cdots \times_\exp\Bigl(\frac{2{\pi}i{j_{d-1}}{x_{d-1}}}{\lambda_{d-1}}\Bigr)_\nonumber \\
&= \exp_\Bigl(\frac{2{\pi}i{j_1}{x_1}}{\lambda_1} \Bigr) \times \cdots \times \exp \Bigl(\frac{2{\pi}i{j_{d-1}}x_{d-1}} {\lambda_{d-1}}\Bigr)\sum_{k=0}^{\lambda_0-1}\Bigl(\exp \Bigl(\frac{2{\pi}i{j_0}k}{\lambda_0}_\Bigr)\Bigr) \nonumber \\
%&= \exp \Bigl(\frac{2{\pi}i{j_1 |
frac{3\ddot{R}}%
R+\frac{\ddot{A}}{2A}-\frac{\dot{A}^2}{4A^2}+\frac{\dot{R}\dot{A}}{RA}
\nonumber \label{8} \\
&&-\frac{3\dot{R}\dot{B}}{2RB}-\frac{\dot{A}\dot{B}}{4AB} \label{8}\end{aligned}$$
$$R_{\theta \theta }=-1+\frac 1A-\frac{r\acute{A}}{2A^2}+\frac{r\acute{B}}{2AB}%
+\frac{R^2r^2}B[-\frac{2\dot{R}^2}{R^2}-\frac{\ddot{R}}R+\frac{\dot{R}\dot{B}%
}{2RB}-\frac{\dot{R}\dot{A}}{2RA}] \label{9}$$
$$R_{\varphi \varphi }=\sin ^2\theta \;R_{\theta \theta
}\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;%
\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;$$
$$R_{rt}=-\frac{\dot{A}}{Ar}-\frac{\acute{B}\dot{R}}{BR}\;\;\;\;\;\;\;\;\;\;\;%
\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;%
\;\;\;\;\; \label{10}$$
The Einstein field equations with $\Lambda $ for empty space are reduced to
$$R_{\mu \nu }=-\Lambda \;g_{\mu \nu } \label{11}$$
With the use of (3) and (11) we may write
$$\frac{R_{rr}}{R^2A}+\frac{R_{tt}}B=0 \label{12}$$
Now inserting (7) and (8) in (12) results
$$-\frac 1{R^2rA}(\frac{\acute{A}}A+\frac{\acute{B}}B)-\frac 1B(-\frac{2\ddot{R%
}}R+\frac{2\dot{R}^2}{R^2}+\frac{\dot{R}\dot{A}}{RA}+\frac{\dot{R}\dot{B | frac{3\ddot{R}}%
R+\frac{\ddot{A}}{2A}-\frac{\dot{A}^2}{4A^2}+\frac{\dot{R}\dot{A}}{RA }
\nonumber \label{8 } \\
& & -\frac{3\dot{R}\dot{B}}{2RB}-\frac{\dot{A}\dot{B}}{4AB } \label{8}\end{aligned}$$
$ $ R_{\theta \theta } = -1+\frac 1A-\frac{r\acute{A}}{2A^2}+\frac{r\acute{B}}{2AB}%
+ \frac{R^2r^2}B[-\frac{2\dot{R}^2}{R^2}-\frac{\ddot{R}}R+\frac{\dot{R}\dot{B}%
} { 2RB}-\frac{\dot{R}\dot{A}}{2RA } ] \label{9}$$
$ $ R_{\varphi \varphi } = \sin ^2\theta \;R_{\theta \theta
} \;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;%
\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;$$
$ $ R_{rt}=-\frac{\dot{A}}{Ar}-\frac{\acute{B}\dot{R}}{BR}\;\;\;\;\;\;\;\;\;\;\;%
\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;%
\;\;\;\;\; \label{10}$$
The Einstein field equations with $ \Lambda $ for empty space are reduced to
$ $ R_{\mu \nu } = -\Lambda \;g_{\mu \nu } \label{11}$$
With the habit of (3) and (11) we may spell
$ $ \frac{R_{rr}}{R^2A}+\frac{R_{tt}}B=0 \label{12}$$
Now inserting (7) and (8) in (12) results
$ $ -\frac 1{R^2rA}(\frac{\acute{A}}A+\frac{\acute{B}}B)-\frac 1B(-\frac{2\ddot{R%
} } R+\frac{2\dot{R}^2}{R^2}+\frac{\dot{R}\dot{A}}{RA}+\frac{\dot{R}\dot{B | fraf{3\ddot{R}}%
R+\frac{\ddot{A}}{2A}-\frac{\dot{X}^2}{4A^2}+\frac{\dot{R}\dot{A}}{RC}
\nonumuer \lagel{8} \\
&&-\frac{3\aot{R}\dot{B}}{2RB}-\frac{\dot{A}\dot{B}}{4AB} \lauel{8}\ebd{alitned}$$
$$R_{\theta \theta }=-1+\frac 1X-\frac{r\acune{A}}{2A^2}+\frac{r\qcutt{B}}{2AB}%
+\frac{R^2r^2}B[-\frac{2\dov{D}^2}{R^2}-\frac{\ddot{R}}R+\fdwc{\doc{R}\vot{B}%
}{2RB}-\frac{\dot{R}\dpt{A}}{2RA}] \labal{9}$$
$$R_{\varphi \varpvi }=\snn ^2\theta \;R_{\theta \theta
}\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;%
\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;$$
$$R_{rt}=-\frac{\dot{A}}{Ar}-\frwc{\acute{N}\dlt{R}}{BR}\;\;\;\;\;\;\;\;\;\;\;%
\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;%
\;\;\;\;\; \label{10}$$
Thg Einxeein field equations with $\Lambda $ for smpty skace are reduced tp
$$R_{\mu \nu }=-\Lambda \;g_{\mu \nu } \lahel{11}$$
Wlth the use of (3) anf (11) we may wtjte
$$\seac{R_{rr}}{R^2A}+\frac{F_{tt}}B=0 \label{12}$$
Now insertihg (7) and (8) in (12) results
$$-\frac 1{R^2rA}(\fraz{\acutz{A}}A+\frac{\acutg{Y}}V)-\frww 1B(-\frac{2\ddot{C%
}}R+\frac{2\qot{R}^2}{R^2}+\frac{\dot{V}\cot{A}}{RA}+\xrac{\dot{T}\dot{B | frac{3\ddot{R}}% R+\frac{\ddot{A}}{2A}-\frac{\dot{A}^2}{4A^2}+\frac{\dot{R}\dot{A}}{RA} \nonumber \label{8} \\ &&-\frac{3\dot{R}\dot{B}}{2RB}-\frac{\dot{A}\dot{B}}{4AB} \label{8}\end{aligned}$$ }=-1+\frac +\frac{R^2r^2}B[-\frac{2\dot{R}^2}{R^2}-\frac{\ddot{R}}R+\frac{\dot{R}\dot{B}% }{2RB}-\frac{\dot{R}\dot{A}}{2RA}] $$R_{\varphi \varphi }=\sin $$R_{rt}=-\frac{\dot{A}}{Ar}-\frac{\acute{B}\dot{R}}{BR}\;\;\;\;\;\;\;\;\;\;\;% \;\;\;\;\; \label{10}$$ The field equations with $ for empty space are reduced $$R_{\mu \nu }=-\Lambda \;g_{\mu \nu } \label{11}$$ With the use of (3) and we may write $$\frac{R_{rr}}{R^2A}+\frac{R_{tt}}B=0 \label{12}$$ Now inserting (7) and (8) in (12) results 1{R^2rA}(\frac{\acute{A}}A+\frac{\acute{B}}B)-\frac }}R+\frac{2\dot{R}^2}{R^2}+\frac{\dot{R}\dot{A}}{RA}+\frac{\dot{R}\dot{B | frac{3\ddot{R}}%
R+\frac{\ddot{A}}{2A}-\frac{\dOt{A}^2}{4A^2}+\frac{\doT{R}\dot{a}}{RA}
\NonUmBer \lAbel{8} \\
&&-\Frac{3\dot{R}\dot{B}}{2RB}-\FRac{\dOt{A}\dot{B}}{4AB} \label{8}\end{alignEd}$$
$$R_{\thEtA \ThetA }=-1+\FrAc 1A-\frAc{r\acutE{a}}{2A^2}+\FRAc{r\AcUtE{B}}{2Ab}%
+\fRAc{r^2r^2}B[-\frAc{2\dOt{R}^2}{R^2}-\fraC{\ddot{R}}R+\fraC{\doT{R}\Dot{B}%
}{2RB}-\frac{\doT{r}\dOt{A}}{2RA}] \label{9}$$
$$r_{\vaRphi \varphi }=\siN ^2\thEta \;R_{\thEtA \thETa
}\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;%
\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;$$
$$R_{rt}=-\FraC{\dot{A}}{ar}-\frac{\ACute{B}\dOt{R}}{BR}\;\;\;\;\;\;\;\;\;\;\;%
\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;%
\;\;\;\;\; \labeL{10}$$
THE EinstEIn field EQUaTionS with $\Lambda $ for empTY sPAce are reduced tO
$$R_{\mu \nu }=-\laMBdA \;G_{\Mu \nU } \laBel{11}$$
With the UsE of (3) anD (11) We may wrITe
$$\FRAC{R_{rR}}{r^2A}+\frac{R_{tt}}B=0 \labEl{12}$$
Now insertINg (7) aNd (8) in (12) reSuLts
$$-\FRac 1{R^2rA}(\Frac{\aCuTE{A}}A+\Frac{\acute{B}}B)-\Frac 1b(-\frac{2\ddot{r%
}}R+\frac{2\DOt{R}^2}{R^2}+\fraC{\Dot{R}\dot{a}}{RA}+\fraC{\doT{R}\dOt{B | frac{3\ddot{R}}%
R+\frac{\ ddot{A}}{2 A}-\f rac {\d ot {A}^ 2}{4 A^2}+\frac{\do t {R}\ dot{A}}{RA}
\nonumber \la be l {8}\ \&&-\f rac{3\d o t{ R } \do t{ B} }{2 RB } -\ frac{ \do t{A}\do t{B}}{4AB} \ la bel{8}\end{a l ig ned}$$
$$ R_{ \theta \thet a } =-1+\f ra c 1 A -\fra c{r \acut e{A}}{ 2 A^2}+\ frac{r\ac ut e {B}}{2 A B}%
+\f r a c{ R^2r ^2}B[-\frac{2\dot { R} ^ 2}{R^2}-\frac{ \ddot{ R} } R+ \ f rac {\d ot{R}\dot{ B} %
}{2 R B}-\fra c {\ d o t {R} \ dot{A}}{2RA}] \label{9} $ $
$$R_{\ va rph i \varp hi }= \s i n ^ 2\theta \;R _{\t heta \the ta
}\; \ ;\;\;\; \ ;\;\;\; \;\;\; \;\ ;\; \;\; \ ;\ ;\ ;\; \; \ ;\; \ ;\ ;\; \ ;\; \;\;\;\; \; \; \;\;%
\;\ ; \ ; \ ;\;\ ;\; \;\; \;\;\ ;\;\;\;\;\;\; \;\ ;$$$$R _{rt} =-\fr ac{\ do t{A}} {Ar}-\ frac{ \a cute{B}\dot{R}} {BR} \;\;\;\;\ ;\; \; \;\ ;\ ;\;%\ ;\;\;\ ;\; \;\ ;\;\;\; \;\;\;\ ; \;\ ;\ ; \ ; \; \;\;\;\;\;\;\;\;\; \; \ ; \; \;\;\;\; \;\;\; %
\ ;\ ; \;\;\; \l abe l{10 } $ $
Th e Ei n st ein fiel d equa t io ns with $ \L ambda$for em pty s p aceare re duced to
$$R _ {\mu \nu }=-\L a mbda \;g_{\mu \n u }\lab el{ 11}$$
With the useof ( 3 )and (11)we ma yw ri t e
$$\frac{R_{rr}}{ R^ 2A}+\f rac{R _{tt}}B=0 \l abel{12}$$ N ow inser ting (7 ) and (8) in (1 2) re sults
$$- \ frac 1{R ^2rA} (\frac{\ acute{A}} A + \frac{\a cut e{B }}B )-\ f r ac 1B(-\frac{2\ d d ot{R %}}R+\fr ac{ 2\dot{R }^2 }{R ^2} +\f ra c{\dot{R} \dot{A}} {R A} +\ fr ac{ \dot{ R }\dot{B | frac{3\ddot{R}}%
R+\frac{\ddot{A}}{2A}-\frac{\dot{A}^2}{4A^2}+\frac{\dot{R}\dot{A}}{RA}
\nonumber_ \label{8}_\\
&&-\frac{3\dot{R}\dot{B}}{2RB}-\frac{\dot{A}\dot{B}}{4AB} \label{8}\end{aligned}$$
$$R_{\theta \theta_}=-1+\frac 1A-\frac{r\acute{A}}{2A^2}+\frac{r\acute{B}}{2AB}%
+\frac{R^2r^2}B[-\frac{2\dot{R}^2}{R^2}-\frac{\ddot{R}}R+\frac{\dot{R}\dot{B}%
}{2RB}-\frac{\dot{R}\dot{A}}{2RA}]__\label{9}$$
$$R_{\varphi \varphi_}=\sin_^2\theta \;R_{\theta \theta
}\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;%
\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;$$
$$R_{rt}=-\frac{\dot{A}}{Ar}-\frac{\acute{B}\dot{R}}{BR}\;\;\;\;\;\;\;\;\;\;\;%
\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;%
\;\;\;\;\;_ \label{10}$$
The Einstein_field equations with $\Lambda_$ for empty_space_are reduced to
$$R_{\mu \nu }=-\Lambda \;g_{\mu \nu } \label{11}$$
With the use of (3)_and_(11) we_may_write
$$\frac{R_{rr}}{R^2A}+\frac{R_{tt}}B=0_ \label{12}$$
Now inserting (7) and_(8) in (12) results
$$-\frac 1{R^2rA}(\frac{\acute{A}}A+\frac{\acute{B}}B)-\frac_1B(-\frac{2\ddot{R%
}}R+\frac{2\dot{R}^2}{R^2}+\frac{\dot{R}\dot{A}}{RA}+\frac{\dot{R}\dot{B |
alpha\neq 1$, $
\log \mathcal{N}_I({\varepsilon},\mathscr{C},P)\leq L{\varepsilon}^{-\alpha}$. Then $$\begin{aligned}
\sup_{C_0 \in \mathscr{C}} {\mathbb{E}}_{C_0} {\lvert\hat{C}_n\Delta C_0\rvert} \lesssim n^{-1/(\alpha+1)}.
\end{aligned}$$
By [@yang1999information], the rate $n^{-1/(\alpha+1)}$ cannot be improved in a minimax sense if furthermore a lower bound on the metric entropy on the same order as that of the upper bound is available.
As a straightforward corollary of the above Theorem \[thm:rate\_optimal\_lse\_reg\], let $\mathscr{C}_d$ be the collection of all convex bodies contained in the unit ball in ${\mathbb{R}}^d$ and $P$ the uniform distribution on the unit ball.
Fix $d\geq 4$. Then $$\begin{aligned}
\sup_{C_0 \in \mathscr{C}_d} {\mathbb{E}}_{C_0} {\lvert\hat{C}_n\Delta C_0\rvert} \lesssim n^{-2/(d+1)}.
\end{aligned}$$
The claim essentially follows from Theorem 8.25 and Corollary 8.26 of [@dudley1999uniform], asserting that we can take $\alpha=(d-1)/2$ in Theorem \[thm:rate\_optimal\_lse\_reg\].
The corollary shows that we can use a global least squares estimator rather than a sieved least squares estimator (cf. [@brunel2013adaptive]) to achieve the optimal rate of convergence.
It is possible to impose certain tail conditions on the density of $P$ to extend the above corollary to a maximum risk bound over all convex sets in ${\mathbb{R}}^d$. A proof in this vein is carried out in the context of $s$-concave density estimation in ${\mathbb{R}}^d$ to be detailed below.
Before the proof of Theorem \[thm:rate\_optimal\_lse\_reg\], we need following:
\[lem:lse\_additive\_error\] Consider the regression model (\[eqn:reg\_model\]) and the least squares estimator $\hat{C}_n$ in (\[def: | alpha\neq 1 $, $
\log \mathcal{N}_I({\varepsilon},\mathscr{C},P)\leq L{\varepsilon}^{-\alpha}$. Then $ $ \begin{aligned }
\sup_{C_0 \in \mathscr{C } } { \mathbb{E}}_{C_0 } { \lvert\hat{C}_n\Delta C_0\rvert } \lesssim n^{-1/(\alpha+1) }.
\end{aligned}$$
By [ @yang1999information ], the rate $ n^{-1/(\alpha+1)}$ cannot be improved in a minimax sense if furthermore a lower bound on the measured information on the same order as that of the upper bounce is available.
As a aboveboard corollary of the above Theorem \[thm: rate\_optimal\_lse\_reg\ ], let $ \mathscr{C}_d$ be the collection of all convex torso contained in the unit musket ball in $ { \mathbb{R}}^d$ and $ P$ the uniform distribution on the unit testis.
Fix $ d\geq 4$. Then $ $ \begin{aligned }
\sup_{C_0 \in \mathscr{C}_d } { \mathbb{E}}_{C_0 } { \lvert\hat{C}_n\Delta C_0\rvert } \lesssim n^{-2/(d+1) }.
\end{aligned}$$
The claim essentially follow from Theorem 8.25 and Corollary 8.26 of [ @dudley1999uniform ], asserting that we can take $ \alpha=(d-1)/2 $ in Theorem \[thm: rate\_optimal\_lse\_reg\ ].
The corollary shows that we can use a global least squares calculator rather than a sieved least squares calculator (cf. [ @brunel2013adaptive ]) to achieve the optimal rate of convergence.
It is possible to enforce certain tail conditions on the density of $ P$ to extend the above corollary to a maximal risk bound over all convex sets in $ { \mathbb{R}}^d$. A proof in this vein is carried out in the context of $ s$-concave density estimate in $ { \mathbb{R}}^d$ to be detailed below.
Before the proof of Theorem \[thm: rate\_optimal\_lse\_reg\ ], we need following:
\[lem: lse\_additive\_error\ ] Consider the regression model (\[eqn: reg\_model\ ]) and the least squares estimator $ \hat{C}_n$ in (\[def: | alpja\neq 1$, $
\log \mathcal{N}_I({\yarepsilon},\mathsct{C},P)\leq L{\tarepsimon}^{-\alpha}$. Then $$\begin{aligned}
\sup_{C_0 \ii \marhscr{X}} {\mathbb{E}}_{C_0} {\lvert\hat{C}_n\Ddlta C_0\rvegt} \lesssin n^{-1/(\aopha+1)}.
\end{emigned}$$
Bn [@yanf1999lnforketion], the rate $k^{-1/(\alpha+1)}$ cannmt be improved iv c minimax sense if furthermore a lowqr bounc ln the metric gntroky jn tgv wame order as that of the uppsr bounv is available.
Ax a straightforward corollwry lf the above Theorfm \[thm:rate\_okfimwo\_lse\_reg\], let $\oathscr{C}_d$ be the colledtion of all convex bodies contxined in the unut bapn in ${\mathbb{C}}^d$ and $P$ the unifovk distsibutiom on the unit nall.
Fmx $d\teq 4$. Then $$\begin{alignev}
\sup_{C_0 \in \mathscr{S}_d} {\mathbb{A}}_{C_0} {\lvert\hat{C}_n\Delta X_0\rcert} \nessvim v^{-2/(e+1)}.
\ehd{elifned}$$
Thf cmaim essenfially folliws from Theorem 8.25 amd Bprollary 8.26 of [@dudler1999ugiform], asserting that we can take $\alpha=(g-1)/2$ ih Theorem \[thm:rate\_optimao\_lse\_reg\].
The corollary dhows thae we can use a global least squares estimator ratver tiav a slevea lfast squares estimator (cf. [@brunel2013adaptive]) to asgitve the optimal rcte of convergenve.
Lt ys possible tu impose dertain tail condihions og the density jf $P$ to extend the above corollqry to a maxpmum risk bound over apl convex szts in ${\mathnb{R}}^d$. A proof in this venn is darried out in the ckvtext of $s$-concavd dvnsidy estimation in ${\mathbb{R}}^d$ eo be deteiled below.
Bdforg the pwoof of Thforem \[thm:rate\_optimal\_lse\_rfg\], we naed followlng:
\[lem:lse\_additive\_error\] Consider the regressiom kodvl (\[eqn:reg\_iodel\]) and the lease squares estikator $\hct{C}_n$ iv (\[def: | alpha\neq 1$, $ \log \mathcal{N}_I({\varepsilon},\mathscr{C},P)\leq L{\varepsilon}^{-\alpha}$. Then \in {\mathbb{E}}_{C_0} {\lvert\hat{C}_n\Delta \lesssim n^{-1/(\alpha+1)}. \end{aligned}$$ cannot improved in a sense if furthermore lower bound on the metric entropy the same order as that of the upper bound is available. As a corollary of the above Theorem \[thm:rate\_optimal\_lse\_reg\], let $\mathscr{C}_d$ be the collection of all bodies in unit in ${\mathbb{R}}^d$ and $P$ the uniform distribution on the unit ball. Fix $d\geq 4$. Then $$\begin{aligned} \in \mathscr{C}_d} {\mathbb{E}}_{C_0} {\lvert\hat{C}_n\Delta C_0\rvert} \lesssim n^{-2/(d+1)}. \end{aligned}$$ claim essentially follows from 8.25 and Corollary 8.26 of asserting we can $\alpha=(d-1)/2$ Theorem The corollary shows we can use a global least squares estimator rather than a sieved least squares estimator (cf. [@brunel2013adaptive]) achieve the of convergence. is to certain tail conditions density of $P$ to extend the a maximum risk bound over all convex sets ${\mathbb{R}}^d$. A in this vein is carried out the context of $s$-concave density estimation in ${\mathbb{R}}^d$ be detailed below. Before the proof of Theorem \[thm:rate\_optimal\_lse\_reg\], we need following: \[lem:lse\_additive\_error\] Consider the (\[eqn:reg\_model\]) and the least estimator $\hat{C}_n$ in | alpha\neq 1$, $
\log \mathcal{N}_I({\varepSilon},\mathsCr{C},P)\lEq L{\VarEpSiloN}^{-\alpHa}$. Then $$\begin{aliGNed}
\sUp_{C_0 \in \mathscr{C}} {\mathbb{E}}_{C_0} {\lVert\hAt{c}_N\DelTA C_0\Rvert} \Lesssim N^{-1/(\AlPHA+1)}.
\enD{aLiGneD}$$
BY [@YaNg1999infOrmAtion], thE rate $n^{-1/(\alphA+1)}$ caNnOt be improved IN a Minimax senSe iF furthermore A loWer bouNd On tHE metrIc eNtropY on the SAme ordEr as that oF tHE upper BOund is aVAIlAble.
as a straightforwarD CoROllary of the aboVe TheoReM \[ThM:RAte\_OptImal\_lse\_reg\], LeT $\mathSCr{C}_d$ be tHE cOLLEctIOn of all convex Bodies contaINed In the uNiT baLL in ${\matHbb{R}}^d$ AnD $p$ thE uniform disTribUtion on thE unit bALl.
Fix $d\gEQ 4$. Then $$\beGin{aliGneD}
\suP_{C_0 \in \MAtHsCr{C}_D} {\mAThbB{e}}_{C_0} {\LveRT\haT{C}_n\Delta c_0\rVeRt} \lesSsim N^{-2/(D+1)}.
\END{aliGneD}$$
The Claim Essentially foLloWs frOM ThEorem 8.25 And CoRollArY 8.26 of [@duDley1999unIform], AsSerting that we caN takE $\alpha=(d-1)/2$ in theOrEm \[tHm:Rate\_oPTimal\_lSe\_rEg\].
THe corolLary shoWS thAt WE CAn Use a global least squArES EsTimator rAther tHAn A sIEved leasT sQuaRes eSTImatoR (cf. [@bRUnEl2013adaptiVe]) to acHIeVe The optiMaL rate oF cOnvErgEnce.
IT Is poSsible To impose CertaIN tail conditionS On the density oF $p$ tO EXtENd thE abOve corollarY to a MAximUm riSK bOunD Over aLl conVeX SeTS in ${\mathbb{R}}^d$. A proof in ThIs vein Is carRied out in the cOntext of $s$-cONCAve densiTy esTImATion in ${\mathbb{R}}^d$ To be dEtailed belOW.
Before tHe proOf of TheoRem \[thm:ratE\_OPtimal\_lsE\_reG\], we NeeD foLLOwIng:
\[lem:lse\_addiTIVe\_erRoR\] ConsidEr tHe regreSsiOn mOdeL (\[eqN:rEg\_model\]) anD the leasT sQuArEs EstImatoR $\Hat{C}_n$ in (\[dEf: | alpha\neq 1$, $
\log \ mathcal{N} _I({\ var eps il on}, \mat hscr{C},P)\leq L{\v arepsilon}^{-\alpha}$. Then $ $ \beg i n{ align ed}
\s u p _{C _0 \ in\m a th scr{C }}{\mathb b{E}}_{C_0 } { \l vert\hat{C}_ n \D elta C_0\r ver t} \lesssimn^{ -1/(\a lp ha+ 1 )}.
\end{ aligne d }$$
B y [@yang1 99 9 inform a tion],t h erate $n^{-1/(\alpha+1 ) }$ cannot be impr oved i na m i n ima x s ense if fu rt hermo r e a low e rb o u ndo n the metricentropy ont hesame o rd era s that of t he upp er bound is ava ilable.
As a s t raightf o rward c orolla ryofthea bo ve Th eo r em\ [t hm: r ate \_optima l\ _l se\_r eg\] , l e t $\ mat hscr {C}_d $ be the coll ect iono f a ll co nvexbodi es cont ainedin th eunit ball in ${ \mat hbb{R}}^d $ a nd $P $the u n iformdis tri butionon theu nit b a l l .
Fix $d\geq 4$. Th en $ $\ begin{al igned} \ sup_{C_0 \ in\mat h s cr{C} _d}{ \m athbb{E} }_{C_0 } { \l vert\ha t{ C}_n\D el taC_0 \rver t } \l esssim n^{-2/( d+1)} .
\end{alig n ed}$$
The cl a im e ss e ntia lly follows fr om T h eore m 8. 2 5and Corol lary8. 2 6o f [@dudley1999unifo rm ], ass ertin g that we can take $\al p h a =(d-1)/2 $ in Th e orem \[thm:rat e\_op timal\_lse \ _reg\].
Thecorollar y shows t h a t we can us e a gl oba l le ast squares e s t imat or rather th an a si eve d l eas t s qu ares esti mator (c f. [ @b ru nel 2013a d aptive]) t o a ch iev e the optima l rat e of c on v erg ence.
I ti s pos si bl e to im po se ce rtai n ta il cond itions on th e den si ty of $P$ to extend th eabove coro ll ary to am a ximum ri sk bound over all conve x sets i n $ {\mat hbb{ R}}^d$. A pr oof in th i s vein is ca rried o uti n thec o nt ext o f $s$-conc a v e d ensit yesti mationin ${\mathbb{R}}^d $ to be detailedbel ow.B ef ore th e pr oo f of T heorem \[thm:ra te\_optima l\ _ ls e\_reg\],w e n ee d follo wing:
\[lem : lse\_ad ditive\_e rror\] Co ns ider t heregression model ( \[eqn:reg \ _mode l \] ) and th e leas tsqu aresestima t or$\hat {C}_n$ i n (\[d ef: | alpha\neq 1$,_$
_ \log \mathcal{N}_I({\varepsilon},\mathscr{C},P)\leq_L{\varepsilon}^{-\alpha}$. Then_$$\begin{aligned}
_ __\sup_{C_0 \in \mathscr{C}}_{\mathbb{E}}_{C_0} {\lvert\hat{C}_n\Delta C_0\rvert}_\lesssim n^{-1/(\alpha+1)}.
_ \end{aligned}$$
By [@yang1999information],_the_rate $n^{-1/(\alpha+1)}$ cannot be improved in a minimax sense if furthermore a lower bound_on_the metric_entropy_on_the same order as that_of the upper bound is_available.
As a_straightforward corollary of the above Theorem \[thm:rate\_optimal\_lse\_reg\], let_$\mathscr{C}_d$_be the collection_of all convex bodies contained in the unit ball_in ${\mathbb{R}}^d$ and $P$ the uniform_distribution on the_unit_ball.
Fix_$d\geq 4$. Then $$\begin{aligned}
_ \sup_{C_0 \in \mathscr{C}_d}_{\mathbb{E}}_{C_0} {\lvert\hat{C}_n\Delta C_0\rvert} \lesssim n^{-2/(d+1)}.
_ \end{aligned}$$
The claim essentially follows from_Theorem 8.25 and Corollary 8.26 of_[@dudley1999uniform], asserting that we can_take $\alpha=(d-1)/2$_in Theorem \[thm:rate\_optimal\_lse\_reg\].
The corollary shows_that we can_use a_global least squares_estimator rather than a sieved least_squares estimator (cf._[@brunel2013adaptive]) to achieve the optimal rate_of_convergence.
It is possible_to_impose_certain tail_conditions on the_density_of $P$_to_extend the above corollary to a_maximum_risk bound over all convex sets in_${\mathbb{R}}^d$. A proof in_this_vein is carried out_in the context of $s$-concave_density estimation in ${\mathbb{R}}^d$ to be_detailed below.
Before_the proof_of Theorem \[thm:rate\_optimal\_lse\_reg\], we need following:
\[lem:lse\_additive\_error\] Consider the regression model (\[eqn:reg\_model\])_and the least squares estimator $\hat{C}_n$_in (\[def: |
wedge j + R_{\bot}^{\alpha\beta}\wedge{\tau _{\alpha\beta}}_{\bot}\nonumber\\
&&\hskip1.5cm -\Gamma _{\bot}^{\alpha\beta}\bigl(\,\vartheta _{[\alpha}\wedge\Sigma ^{\rm matt}_{\beta ]}\bigr)_{\bot} \,\Bigr]\,,\label{energycurrder02}\end{aligned}$$ and finally, comparison of (\[energycurrder02\]) with (\[energycurrder01\]), making use of (\[spincurrconserv\]), yields $${\it l}_u\mathfrak{s} -{{\underline{d}\,q}\over T} = {1\over T}\,\bigl[\,E\wedge j + R_{\bot}^{\alpha\beta}\wedge{\tau _{\alpha\beta}}_{\bot} +\Gamma _{\bot}^{\alpha\beta}\bigl(\,D\,\tau _{\alpha\beta}\bigr)_{\bot}\,\bigr]\,.\label{secondlaw}$$ In the lhs of (\[secondlaw\]) we find the rate of change of the entropy 3-form combined in a familiar way with heat flux and temperature. The interpretation of the first term in the rhs is facilitated by the fact that, according to Ohm’s law $j=\sigma\,{}^\# E$, it is proportional to $E\wedge j ={1\over\sigma} j\wedge{}^\# j \geq 0$, so that it is responsible for entropy growth. The second term is analogous to the first one. If we suppose that all terms in the rhs of (\[secondlaw\]) are $\geq 0$, or, in any case, for vanishing macroscopic free spin current $\tau _{\alpha\beta}$, we can consider (\[secondlaw\]) to be a particular realization of the second law of thermodynamics.
On the other hand, the first law is no other than the conservation equation (\[emmattender\]) for matter energy, rewritten as (\[diff01\]) in terms of the internal energy current 3-form (\[intenergycurr01\]). This reformulation is necessary in order to bring to light the components of $\epsilon ^{\rm u}$ defined in (\[intenergycurr02\]), that is, heat flux and internal energy respectively, thus making possible to compare the first law with the second one (\[secondlaw\]) deduced above. (By the way, notice that the inversion of (\[intenergycurr01\]) to express $\epsilon ^{\rm matt}$ in terms of $\epsilon ^{\rm u}$ suggests to interpret $\epsilon ^{\rm matt}$ as | wedge j + R_{\bot}^{\alpha\beta}\wedge{\tau _ { \alpha\beta}}_{\bot}\nonumber\\
& & \hskip1.5 cm -\Gamma _ { \bot}^{\alpha\beta}\bigl(\,\vartheta _ { [ \alpha}\wedge\Sigma ^{\rm matt}_{\beta ] } \bigr)_{\bot } \,\Bigr]\,,\label{energycurrder02}\end{aligned}$$ and finally, comparison of (\[energycurrder02\ ]) with (\[energycurrder01\ ]), making use of (\[spincurrconserv\ ]), yields $ $ { \it l}_u\mathfrak{s } -{{\underline{d}\,q}\over metric ton } = { 1\over T}\,\bigl[\,E\wedge j + R_{\bot}^{\alpha\beta}\wedge{\tau _ { \alpha\beta}}_{\bot } + \Gamma _ { \bot}^{\alpha\beta}\bigl(\,D\,\tau _ { \alpha\beta}\bigr)_{\bot}\,\bigr]\,.\label{secondlaw}$$ In the luteinizing hormone of (\[secondlaw\ ]) we find the rate of variety of the entropy 3 - form combine in a familiar room with estrus flux and temperature. The interpretation of the first term in the releasing hormone is facilitated by the fact that, according to Ohm ’s jurisprudence $ j=\sigma\,{}^\ # E$, it is proportional to $ E\wedge j = { 1\over\sigma } j\wedge{}^\ # j \geq 0 $, so that it is responsible for entropy growth. The second term is analogous to the inaugural one. If we suppose that all terms in the rhs of (\[secondlaw\ ]) are $ \geq 0 $, or, in any character, for vanishing macroscopic free spin current $ \tau _ { \alpha\beta}$, we can consider (\[secondlaw\ ]) to be a especial realization of the second law of thermodynamics.
On the other hand, the first law is no early than the conservation equation (\[emmattender\ ]) for matter energy, rewritten as (\[diff01\ ]) in terms of the internal department of energy current 3 - kind (\[intenergycurr01\ ]). This reformulation is necessary in order to bring to light the components of $ \epsilon ^{\rm u}$ defined in (\[intenergycurr02\ ]), that is, heat flux and internal energy respectively, therefore create possible to compare the beginning jurisprudence with the second one (\[secondlaw\ ]) deduced above. (By the way, notice that the anastrophe of (\[intenergycurr01\ ]) to express $ \epsilon ^{\rm matt}$ in term of $ \epsilon ^{\rm u}$ suggests to interpret $ \epsilon ^{\rm matt}$ as | wedhe j + R_{\bot}^{\alpha\beta}\wedge{\uau _{\alpha\beta}}_{\bot}\nonumber\\
&&\iskip1.5cm -\Gamma _{\but}^{\alpha\beta}\bigl(\,\vartheta _{[\alphe}\wedte\Signa ^{\rm matt}_{\beta ]}\bigr)_{\bot} \,\Bigr]\,,\labep{energycyrrdtr02}\end{aligned}$$ and hjnally, gjmpadlson mh (\[energycurrder02\]) with (\[enercycurrder01\]), makitg ude of (\[spincurrconserv\]), yields $${\it l}_u\mwthfrak{x} -{{\knderline{d}\,q}\ovet T} = {1\pder F}\,\bigl[\,E\wedge j + R_{\bot}^{\alpha\beta}\wedge{\tzu _{\alphe\beta}}_{\bot} +\Gamma _{\not}^{\alpha\beta}\bigl(\,D\,\tau _{\alpha\heta}\higr)_{\bot}\,\bigr]\,.\label{sefondlaw}$$ In jge jys of (\[seconduaw\]) we find the rate or change of the entropy 3-form coobineb in a famioiqr ady with heav flux and temperablre. The interptetation of thc firvt rerm in the rhs is farilitated by the facj that, accmrbing to Ohm’s law $j=\signa\,{}^\# E$, it is [ropuetiunam vo $S\wedge j ={1\kver\sigma} n\wedge{}^\# j \gew 0$, so that it is rexpjbsible for enfropy drjwth. The second term is analogous to tht firat one. If we suppose thqt all terms in the rjs of (\[secjndlaw\]) are $\geq 0$, or, in any case, for vanishing macrmscopmc frte spin xugrent $\tau _{\alpha\beta}$, we can consider (\[secondlaw\]) fo bv a particular reclization of the sfcpgd law of thetmodynamica.
On the other hand, the fitst laq is no ouher yhan the conservation equatuon (\[emmattenber\]) for matter energy, rewritten cs (\[difg01\]) in yerms of the internal euergy durrent 3-forl (\[intenerfhcurr01\]). This reforoulsthon is ntzessary in order eo bring vo liyht the zomppnents of $\epsiloj ^{\rm m}$ defined in (\[inteneggycutr02\]), thad is, heat vlux and internal energy respecvmvely, thus malitg iossible co comiare the first jaw with the sgcond one (\[secovdlaw\]) dedubed above. (By the way, notice that dje inversion of (\[integergtcure01\]) to exofess $\epsilon ^{\rk matt}$ in terms of $\wpsilon ^{\rm u}$ suggexts fo interpret $\epwilin ^{\rm matt}$ as | wedge j + R_{\bot}^{\alpha\beta}\wedge{\tau _{\alpha\beta}}_{\bot}\nonumber\\ &&\hskip1.5cm -\Gamma ^{\rm ]}\bigr)_{\bot} \,\Bigr]\,,\label{energycurrder02}\end{aligned}$$ finally, comparison of of yields $${\it l}_u\mathfrak{s} T} = {1\over j + R_{\bot}^{\alpha\beta}\wedge{\tau _{\alpha\beta}}_{\bot} +\Gamma _{\bot}^{\alpha\beta}\bigl(\,D\,\tau In the lhs of (\[secondlaw\]) we find the rate of change of the 3-form combined in a familiar way with heat flux and temperature. The interpretation the term the is facilitated by the fact that, according to Ohm’s law $j=\sigma\,{}^\# E$, it is proportional to j ={1\over\sigma} j\wedge{}^\# j \geq 0$, so that is responsible for entropy The second term is analogous the one. If suppose all in the rhs (\[secondlaw\]) are $\geq 0$, or, in any case, for vanishing macroscopic free spin current $\tau _{\alpha\beta}$, we consider (\[secondlaw\]) a particular of second of thermodynamics. On hand, the first law is no conservation equation (\[emmattender\]) for matter energy, rewritten as in terms the internal energy current 3-form (\[intenergycurr01\]). reformulation is necessary in order to bring to the components of $\epsilon ^{\rm u}$ defined in (\[intenergycurr02\]), that is, heat flux and internal thus making possible to the first law the one deduced (By the notice that the inversion of (\[intenergycurr01\]) to express $\epsilon ^{\rm matt}$ terms of $\epsilon ^{\rm u}$ suggests to interpret $\epsilon ^{\rm | wedge j + R_{\bot}^{\alpha\beta}\wedge{\tAu _{\alpha\betA}}_{\bot}\nOnuMbeR\\
&&\hSkip1.5Cm -\GaMma _{\bot}^{\alpha\betA}\Bigl(\,\Vartheta _{[\alpha}\wedge\SigmA ^{\rm maTt}_{\BEta ]}\bIGr)_{\Bot} \,\BiGr]\,,\label{ENeRGYcuRrDeR02}\enD{aLIgNed}$$ anD fiNally, coMparison of (\[EneRgYcurrder02\]) with (\[ENeRgycurrder01\]), MakIng use of (\[spinCurRconseRv\]), YieLDs $${\it l}_U\maThfraK{s} -{{\undeRLine{d}\,q}\Over T} = {1\over t}\,\bIGl[\,E\wedGE j + R_{\bot}^{\aLPHa\Beta}\Wedge{\tau _{\alpha\beta}}_{\BOt} +\gAmma _{\bot}^{\alpha\beTa}\bigl(\,d\,\tAU _{\aLPHa\bEta}\Bigr)_{\bot}\,\bigR]\,.\lAbel{sECondlaw}$$ iN tHE LHs oF (\[Secondlaw\]) we fiNd the rate of CHanGe of thE eNtrOPy 3-form CombiNeD In a Familiar way With Heat flux aNd tempERature. THE interpRetatiOn oF thE firST tErM in ThE Rhs IS fAciLItaTed by the FaCt That, aCcorDING To OhM’s lAw $j=\sIgma\,{}^\# E$, It is proportioNal To $E\wEDge J ={1\over\Sigma} J\wedGe{}^\# J \geq 0$, sO that iT is reSpOnsible for entroPy grOwth. The seConD tErm Is AnaloGOus to tHe fIrsT one. If wE supposE ThaT aLL TErMs in the rhs of (\[secondLaW\]) ARe $\Geq 0$, or, in aNy case, FOr VaNIshing maCrOscOpic FREe spiN curREnT $\tau _{\alphA\beta}$, wE CaN cOnsider (\[SeCondlaW\]) tO be A paRticuLAr reAlizatIon of the SeconD Law of thermodynAMics.
On the otheR HaND, ThE FirsT laW is no other tHan tHE conServATiOn eQUatioN (\[emmaTtENdER\]) for matter energy, rewRiTten as (\[Diff01\]) iN terms of the inTernal enerGY CUrrent 3-foRm (\[inTEnERgycurr01\]). This refOrmulAtion is necESsary in oRder tO bring to Light the cOMPonents oF $\epSilOn ^{\rM u}$ dEFInEd in (\[intenergyCURr02\]), thAt Is, heat fLux And inteRnaL enErgY reSpEctively, tHus makinG pOsSiBlE to CompaRE the firsT lAw wItH thE secoND one (\[seCondlAw\]) deDuCeD AboVe. (By the WAy, NOTice ThAt The iNveRsIon of (\[InteNErgYcurr01\]) to Express $\epSilON ^{\rm mAtT}$ iN terms oF $\epsilon ^{\rm u}$ suGgEsts to inteRpRet $\EpsiloN ^{\RM matt}$ as | wedge j + R_{\bot}^{\alpha \beta}\wed ge{\t au_{\ al pha\ beta }}_{\bot}\nonu m ber\ \
&&\hskip1.5cm -\Gamm a _{\ bo t }^{\ a lp ha\be ta}\big l (\ , \ var th et a _ {[ \ al pha}\ wed ge\Sigm a ^{\rm ma tt} _{ \beta ]}\big r )_ {\bot} \,\ Big r]\,,\label{ ene rgycur rd er0 2 }\end {al igned }$$ an d final ly, compa ri s on of( \[energ y c ur rder 02\]) with (\[ene r gy c urrder01\]), m akingus e o f (\[ spi ncurrconse rv \]),y ields $ $ {\ i t l}_ u \mathfrak{s}-{{\underli n e{d }\,q}\ ov erT } = {1 \over T } \,\ bigl[\,E\we dgej + R_{\b ot}^{\ a lpha\be t a}\wedg e{\tau _{ \al pha\ b et a} }_{ \b o t}+ \G amm a _{ \bot}^{\ al ph a\bet a}\b i g l ( \,D\ ,\t au _ {\alp ha\beta}\bigr )_{ \bot } \,\ bigr] \,.\l abel {s econd law}$$ In t he lhs of (\[seco ndla w\]) we f ind t hera te of change of th e entro py 3-fo r m c om b i n ed in a familiar way w i t hheat flu x andt em pe r ature. T he in terp r e tatio n of th e firstterm i n t he rhs is f acilit at edbythe f a ct t hat, a ccording to O h m’s law $j=\si g ma\,{}^\# E$, it i sp ropo rti onal to $E\ wedg e j = {1\o v er \si g ma} j \wedg e{ } ^\ # j \geq 0$, so that i t is r espon sible for ent ropy growt h . The seco nd t e rm is analogous t o the first one . If we s uppos e that a ll termsi n the rhs of (\ [se con d l aw \]) are $\geq 0 $, o r, in any ca se, for va nis hin g m ac roscopicfree spi ncu rr en t $ \tau_ {\alpha\ be ta} $, we canc onside r (\[ seco nd la w \]) to bea p a r ticu la rreal iza ti on of the sec ond law of therm ody n amic s.
On theother hand, t he first law i s n o othe r than the conservation equation( \[emmat ten der\] ) fo r matterene rgy, r ewr i tten a s (\[d iff01 \] ) i n terms o fthe i nternal en e r gycurre nt 3-f orm (\[ intenergycurr01\]) . Th is reformulat ion isn e ce ssa r yi n o rd e r t o bring to lightthe compon en t sof $\epsil o n ^ {\ rm u}$defined in ( \ [intene rgycurr02 \]), that i s, h e a t f lux and in ternal e nergy res p ectiv e ly , thu s m akingpo ssi ble t o comp a rethe f irst l aw withthe s ec ond one(\[secondlaw\]) deduced above . (By th e way, no tic e th at the in vers ion of (\[ int ene rgycu rr0 1 \]) t o ex p re ss$ \epsi lon^ {\rm matt } $int e rm s of $\epsi l o n ^{ \rm u }$s uggest s to interpret $\epsi l on ^{\rm matt} $ as | wedge j_+ R_{\bot}^{\alpha\beta}\wedge{\tau__{\alpha\beta}}_{\bot}\nonumber\\
&&\hskip1.5cm -\Gamma _{\bot}^{\alpha\beta}\bigl(\,\vartheta _{[\alpha}\wedge\Sigma_^{\rm matt}_{\beta_]}\bigr)_{\bot}_\,\Bigr]\,,\label{energycurrder02}\end{aligned}$$ and_finally,_comparison of (\[energycurrder02\])_with (\[energycurrder01\]), making_use of (\[spincurrconserv\]), yields_$${\it l}_u\mathfrak{s} -{{\underline{d}\,q}\over_T}_= {1\over T}\,\bigl[\,E\wedge j + R_{\bot}^{\alpha\beta}\wedge{\tau _{\alpha\beta}}_{\bot} +\Gamma _{\bot}^{\alpha\beta}\bigl(\,D\,\tau _{\alpha\beta}\bigr)_{\bot}\,\bigr]\,.\label{secondlaw}$$ In the lhs of_(\[secondlaw\])_we find_the_rate_of change of the entropy_3-form combined in a familiar_way with_heat flux and temperature. The interpretation of the_first_term in the_rhs is facilitated by the fact that, according to_Ohm’s law $j=\sigma\,{}^\# E$, it is_proportional to $E\wedge_j_={1\over\sigma}_j\wedge{}^\# j \geq 0$,_so that it is responsible for_entropy growth. The second term is_analogous to the first one. If we_suppose that all terms in the_rhs of (\[secondlaw\]) are $\geq_0$, or,_in any case, for vanishing_macroscopic free spin_current $\tau__{\alpha\beta}$, we can_consider (\[secondlaw\]) to be a particular_realization of the_second law of thermodynamics.
On the other_hand,_the first law_is_no_other than_the conservation equation_(\[emmattender\])_for matter_energy,_rewritten as (\[diff01\]) in terms of_the_internal energy current 3-form (\[intenergycurr01\]). This reformulation_is necessary in order_to_bring to light the_components of $\epsilon ^{\rm u}$_defined in (\[intenergycurr02\]), that is, heat_flux and_internal energy_respectively, thus making possible to compare the first law with the_second one (\[secondlaw\]) deduced above. (By_the way, notice that_the inversion_of_(\[intenergycurr01\]) to express_$\epsilon_^{\rm matt}$_in terms of $\epsilon ^{\rm u}$ suggests_to interpret_$\epsilon ^{\rm matt}$ as |
ilane (OTS)-modified silica microspheres capped with a catalytic Pt hemisphere patch.]{} [Additionally, Yan et al [@Dipolar_Granick] have recently reported different collective states of active dipolar Janus colloids, whose motion is originated by an induced-charge electrophoresis. These colloids self-assemble in different morphologies, ranging from: gas, swarms, chains and clusters, by modifying both the charge imbalance of the colloids and the electric field frequency. Moreover, in [@Dipolar_Granick], the authors were able to numerically reproduce all observed states by means of an overdamped molecular dynamics simulation whose particles were interacting via directed imbalanced interactions.]{}
[Motivated by these two experimental works [@Gao_OTSJanusExp; @Dipolar_Granick], simulations of a dilute quasi-2D suspension of ABP interacting via anisotropic amphiphilic Janus interactions have been reported in Ref. [@Stewart_17]. To the best of our knowledge, this is the first numerical attempt to unravel the collective behaviour of amphiphilic active Janus particles, taking into account their anisotropic nature. In Ref. [@Stewart_17] both the effect of interaction directionality and the propulsion speed have been used to control the physical properties of the assembled active aggregates. One might suggest that these two parameters could be easily tuned also in experiments, given that the concentration of a cationic surfactant could be used to reverse the propulsion’s direction and adding pH neutral salts could be used to control the propulsion speed [@Brown_Poon_SaltJanus].]{} In order to establish the relevance of hydrodynamics in this suspension, we study the collective behaviour of quasi 2D suspensions of spherical squirmers interacting via an anisotropic Janus potential, considering different interaction ranges, interaction strengths and hydrodynamic signatures. When clusters appear, we characterise their morphology as a function of the above mentioned features. [The objective of this manuscript is the analysis of minimal models that contain the essential ingredients of activity and steric interactions for heterogeneous particles. The scenarios we identify can be clearly correlated with the competition between self-propulsion, stress generation, and attraction among particles. It is true that phoresis is more complex and can also involve long range interactions among the active elements. Our results serve as a benchmarking scenario if phoresis is controlled by near field effects. In this sense, the systematic analysis we have carried out provides a valuable contribution in this general perspective. In a more specific perspective, | ilane (OTS)-modified silica microspheres capped with a catalytic Pt hemisphere patch. ] { } [ Additionally, Yan et al [ @Dipolar_Granick ] have recently reported different corporate state of active dipolar Janus colloids, whose motion is originate by an induce - charge electrophoresis. These colloids self - assemble in unlike morphologies, ranging from: gas, drove, chains and clusters, by change both the charge imbalance of the colloid and the electric field frequency. Moreover, in [ @Dipolar_Granick ], the writer were able to numerically reproduce all observed state by means of an overdamped molecular dynamics simulation whose particle were interacting via directed imbalanced interaction. ] { }
[ Motivated by these two experimental works [ @Gao_OTSJanusExp; @Dipolar_Granick ], simulations of a dilute quasi-2D suspension of ABP interact via anisotropic amphiphilic Janus interactions have been reported in Ref. [ @Stewart_17 ]. To the best of our knowledge, this is the first numerical attack to ravel the collective behaviour of amphiphilic active Janus particles, taking into account their anisotropic nature. In Ref. [ @Stewart_17 ] both the effect of interaction directionality and the propulsion speed have been used to operate the physical properties of the assembled active aggregate. One might suggest that these two parameters could be easily tuned besides in experiments, give that the concentration of a cationic surfactant could be use to reverse the propulsion ’s direction and adding pH neutral salts could be use to control the propulsion speed [ @Brown_Poon_SaltJanus ]. ] { } In order to establish the relevance of hydrodynamics in this pause, we study the corporate behavior of quasi 2D suspensions of spherical squirmers interacting via an anisotropic Janus electric potential, considering different interaction ranges, interaction strengths and hydrodynamic signatures. When clusters appear, we characterise their morphology as a function of the above note features. [ The objective of this manuscript is the psychoanalysis of minimal models that contain the essential ingredients of action and steric interactions for heterogenous particles. The scenarios we identify can be distinctly correlated with the rival between self - propulsion, stress generation, and attraction among particles. It is true that phoresis is more complex and can also involve long range interaction among the active elements. Our results suffice as a benchmarking scenario if phoresis is controlled by near playing field effects. In this sense, the systematic psychoanalysis we have carried out provides a valuable contribution in this general perspective. In a more specific perspective, | ilaje (OTS)-modified silica migrospheres cappeb with e catalgtic Pt femisphere patch.]{} [Additionallb, Yab et ql [@Dipolar_Granick] have recently reportee dihferent collectits states of adbive bi'olar Janus colkoids, whosa motion is orhgkncted by an induced-charge electrophorqsis. Thrsf colloids sels-asstmbje ih different morphologies, ranging fdom: gas, swarms, chainx and clusters, by modifyinh bohh the charge imbapance of thg cojooids and thd electric field frequgncy. Moreover, in [@Dipolar_Granick], ghe abthors were avle jo numericaloy reiroduce all onxerved states by means of ak ovecdamped molecular dynamicv simulation whose particlev cere interacting via eieecteg imtalavxed inuerectjons.]{}
[Mohiveted by theae two expeeimental works [@Gao_OUSJwbusExp; @Dipolad_Granisk], simulations of a dilute quasi-2D suspenspon kf ABP interacting via qnisotropic amphiphillc Janus ynteractions have been reported in Ref. [@Stewart_17]. To dhe bxsg oy our yboaledge, this is the first numerical attempt to hntanel the collectivc behaviour of ampniohojic active Javus paxfidles, taking into afcount jheir qnisotropyc nsture. In Ref. [@Stewart_17] both thw effect of pnteeaction directionapity and thz propolsion speed have been used tu cohtrol the pjysical pduperties of the xssvmblad activt aggregates. One myght suggxst tkat thesd twp paraieters coupd be easily tuned also ln exkerimetts, given hhat the concentration of a catmpnic surfactsnd cmuld be bsed tp reverse the propulsion’s dnrection and xdding pH heutral salts coulq be used to wlntrol the pcopulsion spewd [@Beown_Poov_RaltJanus].]{} In otder to establish tye relevance of hyvrodinzmics in this sbrkebsion, we study thd cjlpertive tehaviour of quari 2A suspdnsions of woherocal squirmers interdctihg via an anisotrolig Janus pitential, considering cifferent interactlon renges, mnteravtijn strengths and hydrodynamic aignatured. Wmen clusters wppeqr, we characcerise their morphology as a function of the above mentionwd features. [The objgctlve of this kanussript is dhe analysis of mininal models that cpntain the essential jngredhents of activity and steric interactions for heterogeneous particles. The scenqrios xe identify czn br cledrjv correjatev xith the competitpon between self-propulsion, stresr generadijn, and attraction among partocues. It is trug that phoresis is more cojplex anc can also involve long range imteractions among the acbive elemwnts. Kur resultr serve as a benchkarkiny scenqrio if phovesis is contrullrd by near vielv eyfects. In this sense, the systematuc anaoydis we have carrmed out provldts a vamuable contribution in thia genetal perspectivr. In a more sprcific perspective, | ilane (OTS)-modified silica microspheres capped with a hemisphere [Additionally, Yan al [@Dipolar_Granick] have of dipolar Janus colloids, motion is originated an induced-charge electrophoresis. These colloids self-assemble different morphologies, ranging from: gas, swarms, chains and clusters, by modifying both the imbalance of the colloids and the electric field frequency. Moreover, in [@Dipolar_Granick], the were to reproduce observed states by means of an overdamped molecular dynamics simulation whose particles were interacting via directed interactions.]{} [Motivated by these two experimental works [@Gao_OTSJanusExp; simulations of a dilute suspension of ABP interacting via amphiphilic interactions have reported Ref. To the best our knowledge, this is the first numerical attempt to unravel the collective behaviour of amphiphilic active Janus taking into anisotropic nature. Ref. both effect of interaction the propulsion speed have been used physical properties of the assembled active aggregates. One suggest that two parameters could be easily tuned in experiments, given that the concentration of a surfactant could be used to reverse the propulsion’s direction and adding pH neutral salts could to control the propulsion [@Brown_Poon_SaltJanus].]{} In order establish relevance hydrodynamics this suspension, study the collective behaviour of quasi 2D suspensions of spherical squirmers via an anisotropic Janus potential, considering different interaction ranges, interaction hydrodynamic When clusters appear, characterise their morphology as function the above mentioned features. of manuscript of models contain the essential ingredients activity and steric interactions for particles. The scenarios we with the competition between self-propulsion, stress generation, and among particles. It is true that phoresis more complex and can also involve long range interactions among the active Our results a benchmarking scenario if phoresis is controlled by field effects. In this the systematic analysis we have carried out provides a contribution this general In a more perspective, | ilane (OTS)-modified silica micRospheres cApped WitH a cAtAlytIc Pt Hemisphere patcH.]{} [addiTionally, Yan et al [@Dipolar_graniCk] HAve rECeNtly rEported DIfFERenT cOlLecTiVE sTates Of aCtive diPolar Janus ColLoIds, whose motiON iS originateD by An induced-chaRge ElectrOpHorESis. ThEse ColloIds selF-AssembLe in diffeReNT morphOLogies, rANGiNg frOm: gas, swarms, chains ANd CLusters, by modifYing boTh THe CHArgE imBalance of tHe ColloIDs and thE ElECTRic FIeld frequency. moreover, in [@DIPolAr_GranIcK], thE AuthorS were AbLE to Numerically ReprOduce all oBserveD States bY Means of An overDamPed MoleCUlAr DynAmICs sIMuLatIOn wHose partIcLeS were InteRACTIng vIa dIrecTed imBalanced interActIons.]{}
[mOtiVated By theSe twO eXperiMental Works [@gaO_OTSJanusExp; @DipOlar_granick], siMulAtIonS oF a dilUTe quasI-2D sUspEnsion oF ABP intERacTiNG VIa Anisotropic amphiphIlIC jaNus interActionS HaVe BEen reporTeD in ref. [@STEWart_17]. TO the BEsT of our knOwledgE, ThIs Is the fiRsT numerIcAl aTteMpt to UNravEl the cOllectivE behaVIour of amphiphiLIc active Janus PArTIClES, takIng Into account TheiR AnisOtroPIc NatURe. In REf. [@SteWaRT_17] bOTh the effect of interaCtIon dirEctioNality and the pRopulsion sPEED have beeN useD To COntrol the physiCal prOperties of THe assembLed acTive aggrEgates. One MIGht suggeSt tHat TheSe tWO PaRameters could BE EasiLy Tuned alSo iN experiMenTs, gIveN thAt The concenTration oF a CaTiOnIc sUrfacTAnt could Be UseD tO reVerse THe propUlsioN’s diReCtIOn aNd addinG PH NEUtraL sAlTs coUld Be Used tO conTRol The propUlsion speEd [@BROwn_POoN_SAltJanuS].]{} In order to estAbLish the relEvAncE of hydRODynamics In this suspension, we study THe colleCtiVe behAvioUr of quasi 2d suSpensiOns OF spherIcal sqUirmeRs IntERActinG VIa An aNiSotropic JaNUS poTentiAl, ConsIdering Different interactiON raNges, interactiOn sTrenGTHs And HYdROdyNaMIc sIGNatures. When clusTers appear, We CHaRacterise tHEir MoRphologY as a funCtion OF the aboVe mentionEd featureS. [THe obJECtiVe of this maNuscript Is the analYSis of MInImal mOdeLs that CoNtaIn the EssentIAl iNgredIents oF aCtivitY and sTeRic interActions for heterogeneous ParticLes. ThE scEnarios we IdeNTifY can be cleArly Correlated WitH thE compEtiTIon beTweeN SeLf-pROpulsIon, sTRess generATiOn, aND AtTraction amoNG PArtIcles. it iS True thAt phOresis is more complEX and can also invOlve LONg rAngE InteRaCtions among the ActIvE ELements. OUr Results servE as a bencHmARking ScenarIo if phOresis iS COnTRolled By neAr fIeld effecTs. IN tHIs sense, ThE sYStematIc anAlYsis we Have caRRied OUT provides a valuabLe conTRIbutiON in This gEnEral perSPectIve. In a more Specific perSpectiVe, | ilane (OTS)-modified silic a microsph erescap ped w itha ca talytic Pt hem i sphe re patch.]{} [Addition ally, Y a n et al [@Di polar_G r an i c k]ha ve re ce n tl y rep ort ed diff erent coll ect iv e states ofa ct ive dipola r J anus colloid s,whosemo tio n is o rig inate d by a n induc ed-charge e l ectrop h oresis. T he se c olloids self-asse m bl e in differentmorpho lo g ie s , ra ngi ng from: g as , swa r ms, cha i ns a n d c l usters, by mo difying bot h th e char ge im b alance of t he col loids and t he e lectric f ield f r equency . Moreov er, in [@ Dip olar _ Gr an ick ], the au tho r s w ere able t onumer ical l y r epro duc e al l obs erved statesbymean s of an o verda mped m olecu lar dy namic ssimulation whos e pa rticles w ere i nte ra cting via di rec ted imbala nced in t era ct i o n s. ]{}
[Motivated by t h e se two exp erimen t al w o rks [@Ga o_ OTS Janu s E xp; @ Dipo l ar _Granick ], sim u la ti ons ofadilute q uas i-2 D sus p ensi on ofABP inte racti n g via anisotro p ic amphiphili c J a n us inte rac tions havebeen repo rted in Re f . [@S tewar t_ 1 7] . To the best of our k nowled ge, t his is the fi rst numeri c a l attempt tou nr a vel the collec tivebehaviouro f amphip hilic activeJanus par t i cles, ta kin g i nto ac c o un t their aniso t r opic n ature.InRef. [@ Ste war t_1 7]bo th the ef fect ofin te ra ct ion dire c tionalit yand t hepropu l sion s peedhave b ee n us ed to c o nt r o l th eph ysic alpr opert ieso f t he asse mbled act ive aggr eg at es. One might sugges tthat these t woparame t e rs could be easily tuned also i n experi men ts, g iven that the co ncentr ati o n of a catio nic s ur fac t a nt co u l dbeus ed to reve r s e t he pr op ulsi on’s di rection and adding pHneutral salts co uldb e u sed to con tr o l t h e propulsion spe ed [@Brown _P o on _SaltJanus ] .]{ }In orde r to es tabli s h the r elevanceof hydrod yn amic s inthis suspe nsion, w e study t h e col l ec tivebeh aviour o f q uasi2D sus p ens ionsof sph er ical s quirm er s intera cting via an anisotropi c Janu s pot ent ial, cons ide r ing differen t in teractionran ges , int era c tionstre n gt hsa nd hy drod y namic sig n at ure s . W hen cluster s a ppe ar, w e c h aracte rise their morphology as a functionof t h e ab ove ment io ned features.[Th eo b jectiveof this manus cript is t h e ana lysisof min imal mo d e ls that c onta inthe essen tia li ngredie nt so f acti vity a nd ste ric in t erac t i ons for heteroge neous p artic l es. Thesc enarios we i dentify ca n be clearl y corr elat ed wi th theco mpetit ion b etween sel f -propulsi on, s tress g en erat ion , andattr a c tionamon gpar ticles. I t is tr ue tha t ph oresi sis m ore compl e x and ca n a l so invo lv e l o n g rang e i n t eractionsamo ng th e active ele m ents . O u r res ults s erve a s a ben c hma rk ing sce nar i o if phore sis is co n tro ll ed b y near f ie ld e ff ect s. In this sense, the sys tematica n alysis we have carrie dout provides a va lu a ble con t r ibu tion in thi s g ene ralper spec t ive. In a more specificpers pecti v e, | ilane (OTS)-modified_silica microspheres_capped with a catalytic_Pt hemisphere_patch.]{}_[Additionally, Yan_et_al [@Dipolar_Granick] have_recently reported different_collective states of active_dipolar Janus colloids,_whose_motion is originated by an induced-charge electrophoresis. These colloids self-assemble in different morphologies, ranging_from:_gas, swarms,_chains_and_clusters, by modifying both the_charge imbalance of the colloids_and the_electric field frequency. Moreover, in [@Dipolar_Granick], the authors_were_able to numerically_reproduce all observed states by means of an overdamped_molecular dynamics simulation whose particles were_interacting via directed_imbalanced_interactions.]{}
[Motivated_by these two experimental_works [@Gao_OTSJanusExp; @Dipolar_Granick], simulations of a_dilute quasi-2D suspension of ABP interacting_via anisotropic amphiphilic Janus interactions have been_reported in Ref. [@Stewart_17]. To the best_of our knowledge, this is_the first_numerical attempt to unravel the_collective behaviour of_amphiphilic active_Janus particles, taking_into account their anisotropic nature. In_Ref. [@Stewart_17] both the_effect of interaction directionality and the_propulsion_speed have been_used_to_control the_physical properties of_the_assembled active_aggregates._One might suggest that these two_parameters_could be easily tuned also in experiments,_given that the concentration_of_a cationic surfactant could_be used to reverse the_propulsion’s direction and adding pH neutral_salts could_be used_to control the propulsion speed [@Brown_Poon_SaltJanus].]{} In order to establish the_relevance of hydrodynamics in this suspension,_we study the collective_behaviour of_quasi_2D suspensions of_spherical_squirmers interacting_via an anisotropic Janus potential, considering different_interaction ranges,_interaction strengths and hydrodynamic signatures. When_clusters appear, we characterise_their_morphology as a function of the_above mentioned features. [The objective of_this manuscript is the analysis_of_minimal_models that contain the essential_ingredients of activity and steric interactions_for heterogeneous particles._The scenarios we identify can be clearly_correlated_with the competition between self-propulsion, stress_generation,_and attraction among particles. It is_true_that_phoresis is more complex and_can also involve long range interactions_among the active elements. Our results serve as a_benchmarking scenario if_phoresis is controlled by near_field_effects._In this sense, the systematic analysis we have carried out_provides a_valuable contribution in_this general perspective. In a more specific perspective, |
2)\circ g(
t_1) \\
&= \bigvee_{ t_1+ t_2= t} h( t_2)\circ f( t_1)\vee \bigvee_{ t_1+
t_2= t} h( t_2)\circ g( t_1) \\
&= f{\mathrel{\mathop{\triangleright}}}h( t)\vee g{\mathrel{\mathop{\triangleright}}}h( t)=( f{\mathrel{\mathop{\triangleright}}}h\vee
g{\mathrel{\mathop{\triangleright}}}h)( t)\,.
\end{aligned}$$ The proof is complete.
Finally, a cycle in an RTEA results in a $^*$-operation:
Let $f \in {\mathcal F}$. The Kleene star of $f$ is the function $f^*\in {\mathcal F}$ such that $$\forall t\in {[ 0, \infty]}: f^*( t)= \bigvee_{n\ge 0} f^n( t)$$
Note that $f^*$ is defined for all $f\in {\mathcal F}$ because ${\mathcal F}$ is a complete lattice. We can now define the set of general real-time energy functions, corresponding to general RTEAs:
The set ${\mathcal E}$ of *real-time energy functions* is the subsemiring of ${\mathcal F}$ generated by ${\mathcal A}$, [*i.e.,*]{} the subset of ${\mathcal F}$ inductively defined by
- ${\mathcal A}\subseteq {\mathcal E}$,
- if $f, g\in {\mathcal E}$, then $f{\mathrel{\mathop{\triangleright}}}g\in {\mathcal E}$ and $f\vee g\in {\mathcal E}$.
We will show below that ${\mathcal E}$ is locally closed, which entails that for each $f\in {\mathcal E}$, also $f^*\in {\mathcal E}$, hence ${\mathcal E}$ indeed encompasses all RTEFs.
\[le:locallyclosed\] For every $f\in {\mathcal E}$ there exists $N\ge 0$ so that $f^*= \bigvee_{ n= 0}^N f^n$.
By distributivity, we can write $f$ as a finite supremum $ | 2)\circ g (
t_1) \\
& = \bigvee _ { t_1 + t_2= t } h (t_2)\circ f (t_1)\vee \bigvee _ { t_1 +
t_2= t } h (t_2)\circ g (t_1) \\
& = f{\mathrel{\mathop{\triangleright}}}h (t)\vee g{\mathrel{\mathop{\triangleright}}}h (t)= (f{\mathrel{\mathop{\triangleright}}}h\vee
g{\mathrel{\mathop{\triangleright}}}h) (t)\, .
\end{aligned}$$ The proof is complete.
ultimately, a hertz in an RTEA results in a $ ^*$-operation:
Let $ f \in { \mathcal F}$. The Kleene star of $ f$ is the routine $ f^*\in { \mathcal F}$ such that $ $ \forall t\in { [ 0, \infty ] }: f^ * (t)= \bigvee_{n\ge 0 } f^n (t)$$
Note that $ f^*$ is defined for all $ f\in { \mathcal F}$ because $ { \mathcal F}$ is a arrant wicket. We can now define the set of general real - prison term energy functions, equate to cosmopolitan RTEAs:
The set $ { \mathcal E}$ of * real - meter energy functions * is the subsemiring of $ { \mathcal F}$ generate by $ { \mathcal A}$, [ * i.e., * ] { } the subset of $ { \mathcal F}$ inductively defined by
- $ { \mathcal A}\subseteq { \mathcal E}$,
- if $ f, g\in { \mathcal E}$, then $ f{\mathrel{\mathop{\triangleright}}}g\in { \mathcal E}$ and $ f\vee g\in { \mathcal E}$.
We will show below that $ { \mathcal E}$ is locally closed, which fee-tail that for each $ f\in { \mathcal E}$, also $ f^*\in { \mathcal E}$, hence $ { \mathcal E}$ indeed encompasses all RTEFs.
\[le: locallyclosed\ ] For every $ f\in { \mathcal E}$ there exist $ N\ge 0 $ so that $ f^*= \bigvee _ { n= 0}^N f^n$.
By distributivity, we can write $ f$ as a finite supremum $ | 2)\cirf g(
t_1) \\
&= \bigvee_{ t_1+ t_2= t} h( t_2)\circ f( j_1)\vwe \bigtee_{ t_1+
t_2= t} f( t_2)\circ g( t_1) \\
&= f{\mathrel{\matiop{\teiangoeright}}}h( t)\vee g{\mathrel{\oathop{\triwnglerigyt}}}h( u)=( f{\mathrel{\mathop{\tcjanglerlyht}}}h\vsc
g{\mathrel{\mathok{\trianglerigvt}}}h)( t)\,.
\end{aligtea}$$ Che proof is complete.
Finally, a cycle in an TTFA results in w $^*$-optraeion:
Mvt $f \in {\mathcal F}$. The Kleene stzr of $f$ is the functoon $f^*\in {\mathcal F}$ such thah $$\fogall t\in {[ 0, \infty]}: f^*( t)= \bigvee_{n\gg 0} f^g( t)$$
Note that $w^*$ is defintd for all $f\jn {\mathcal F}$ because ${\mathcal F}$ ks a eomplete lajcuce. Fe can now vefine the set of nvneral seal-timr energy functlons, rorrwsponding to general CTEAs:
The set ${\mathcal E}$ of *rean-tnme energy functions* us the vubsamirkbg uf ${\javhczl F}$ gfnecated by ${\mafhcal A}$, [*i.e.,*]{} tye subset of ${\mathcak S}$ inductively sefineq fy
- ${\mathcal A}\subseteq {\mathcal E}$,
- if $f, g\ih {\mathcal E}$, then $f{\mathrwl{\mathop{\triangleright}}}h\in {\mathcwl E}$ and $f\vee g\in {\mathcal E}$.
We will show below thad ${\maticxl T}$ lf oofally closed, which entails that for each $f\in {\iztncsl E}$, also $f^*\in {\iathcal E}$, hrnfe ${\iathcal E}$ indged encompzsses all RTEFs.
\[le:llcallycjosed\] For everr $f\im {\mathcal E}$ there exists $N\gw 0$ so that $f^*= \vigvee_{ n= 0}^N f^n$.
By didtributivitv, we csn wrote $f$ as a finite supreoum $ | 2)\circ g( t_1) \\ &= \bigvee_{ t_1+ h( f( t_1)\vee t_1+ t_2= t} &= t)\vee g{\mathrel{\mathop{\triangleright}}}h( t)=( g{\mathrel{\mathop{\triangleright}}}h)( t)\,. \end{aligned}$$ proof is complete. Finally, a cycle an RTEA results in a $^*$-operation: Let $f \in {\mathcal F}$. The Kleene of $f$ is the function $f^*\in {\mathcal F}$ such that $$\forall t\in {[ \infty]}: t)= 0} t)$$ Note that $f^*$ is defined for all $f\in {\mathcal F}$ because ${\mathcal F}$ is a lattice. We can now define the set of real-time energy functions, corresponding general RTEAs: The set ${\mathcal of energy functions* the of F}$ generated by A}$, [*i.e.,*]{} the subset of ${\mathcal F}$ inductively defined by - ${\mathcal A}\subseteq {\mathcal E}$, - if g\in {\mathcal $f{\mathrel{\mathop{\triangleright}}}g\in {\mathcal and g\in E}$. We will that ${\mathcal E}$ is locally closed, for each $f\in {\mathcal E}$, also $f^*\in {\mathcal hence ${\mathcal indeed encompasses all RTEFs. \[le:locallyclosed\] For $f\in {\mathcal E}$ there exists $N\ge 0$ so $f^*= \bigvee_{ n= 0}^N f^n$. By distributivity, we can write $f$ as a finite supremum | 2)\circ g(
t_1) \\
&= \bigvee_{ t_1+ t_2= t} h( t_2)\circ f( t_1)\veE \bigvee_{ t_1+
t_2= t} H( t_2)\cirC g( t_1) \\
&= F{\maThRel{\mAthoP{\triangleright}}}H( T)\vee G{\mathrel{\mathop{\triangleRight}}}H( t)=( F{\MathREl{\MathoP{\triangLErIGHt}}}h\VeE
g{\MatHrEL{\mAthop{\TriAngleriGht}}}h)( t)\,.
\end{alIgnEd}$$ the proof is coMPlEte.
Finally, A cyCle in an RTEA rEsuLts in a $^*$-OpEraTIon:
LeT $f \iN {\mathCal F}$. ThE kleene Star of $f$ is ThE FunctiON $f^*\in {\matHCAl f}$ sucH that $$\forall t\in {[ 0, \infTY]}: f^*( T)= \Bigvee_{n\ge 0} f^n( t)$$
NoTe that $F^*$ iS DeFINed For All $f\in {\mathCaL F}$ becAUse ${\mathCAl f}$ IS A coMPlete lattice. WE can now defiNE thE set of GeNerAL real-tIme enErGY fuNctions, corrEspoNding to geNeral RteAs:
The sET ${\mathcaL E}$ of *reAl-tIme EnerGY fUnCtiOnS* Is tHE sUbsEMirIng of ${\matHcAl f}$ geneRateD BY ${\MAthcAl A}$, [*I.e.,*]{} thE subsEt of ${\mathcal F}$ iNduCtivELy dEfineD by
- ${\maThcaL A}\SubseTeq {\matHcal E}$,
- If $F, g\in {\mathcal E}$, theN $f{\maThrel{\mathOp{\tRiAngLeRight}}}G\In {\mathCal e}$ anD $f\vee g\iN {\mathcaL e}$.
We WiLL SHoW below that ${\mathcal E}$ Is LOCaLly closeD, which ENtAiLS that for EaCh $f\In {\maTHCal E}$, aLso $f^*\IN {\mAthcal E}$, hEnce ${\maTHcAl e}$ indeed EnCompasSeS alL RTeFs.
\[le:LOcalLyclosEd\] For eveRy $f\in {\MAthcal E}$ there exISts $N\ge 0$ so that $f^*= \BIgVEE_{ n= 0}^n F^n$.
By DisTributivity, We caN WritE $f$ as A FiNitE SupreMum $ | 2)\circ g(
t_1) \\
&= \big vee_{ t_ 1+t_ 2= t } h( t_2)\circ f(t _1)\ vee \bigvee_{ t_1+
t_ 2= t} h ( t _2)\c irc g(t _1 ) \\ &=f{ \ ma threl {\m athop{\ triangleri ght }} }h( t)\vee g { \m athrel{\ma tho p{\triangler igh t}}}h( t )=( f{\ma thr el{\m athop{ \ triang leright}} }h \ vee
g{\m a t hr el{\ mathop{\triangler i gh t }}}h)( t)\,.
\end{ al i gn e d }$$ Th e proof is c omple t e.
Fin a ll y , a c y cle in an RTE A results i n a$^*$-o pe rat i on:
L et $f \ i n { \mathcal F} $. T he Kleene staro f $f$ i s the fu nction $f ^*\ in { \ ma th cal F } $ s u ch th a t $ $\forall t \i n {[0, \ i n f t y]}: f^ *( t )= \b igvee_{n\ge 0 } f ^n(t )$$
Not e tha t $f ^* $ isdefine d for a ll $f\in {\math calF}$ becau se${ \ma th cal F } $ is a co mpl ete lat tice. W e ca nn o w d efine the set of g en e r al real-ti me ene r gy f u nctions, c orr espo n d ing t o ge n er al RTEAs :
The se t${\math ca l E}$of *r eal -time ener gy fun ctions*is th e subsemiring o f ${\mathcal F } $g e ne r ated by ${\mathcal A}$ , [*i .e., * ]{ } t h e sub set o f$ {\ m athcal F}$ inductiv el y defi ned b y
- ${\mat hcal A}\su b s e teq {\ma thca l E } $,
- if $f, g\in {\mathcal E}$, the n $f{ \mathrel {\mathop{ \ t riangler igh t}} }g\ in{ \ ma thcal E}$ and $ f\ve eg\in {\ mat hcal E} $.
We wi llsh ow belowthat ${\ ma th ca lE}$ is l o cally cl os ed, w hic h ent a ils th at fo r ea ch $ f \in {\math c al E }$,al so $f^ *\i n{\mat hcal E}$ , hence ${\mathc alE }$ i nd ee d encom passes all RT EF s.
\[le:l oc all yclose d \ ] For ev ery $f\in {\mathcal E}$ there e xis ts $N \ge0$ so tha t $ f^*= \ big v ee_{ n = 0}^N f^n$ .
By d istri b u ti vit y, we can wr i t e $ f$ as a fin ite sup remum $ | 2)\circ g(
_ _ t_1)_\\
__ &=_\bigvee_{_t_1+ t_2= t}_h( t_2)\circ f(_t_1)\vee \bigvee_{ t_1+
_ __t_2= t} h( t_2)\circ g( t_1) \\
&= f{\mathrel{\mathop{\triangleright}}}h( t)\vee g{\mathrel{\mathop{\triangleright}}}h(_t)=(_f{\mathrel{\mathop{\triangleright}}}h\vee
___ g{\mathrel{\mathop{\triangleright}}}h)( t)\,.
_\end{aligned}$$ The proof is complete.
Finally,_a cycle_in an RTEA results in a $^*$-operation:
Let $f_\in_{\mathcal F}$. The_Kleene star of $f$ is the function $f^*\in {\mathcal_F}$ such that $$\forall t\in {[_0, \infty]}: f^*(_t)=_\bigvee_{n\ge_0} f^n( t)$$
Note that_$f^*$ is defined for all $f\in_{\mathcal F}$ because ${\mathcal F}$ is_a complete lattice. We can now define_the set of general real-time energy_functions, corresponding to general RTEAs:
The_set ${\mathcal_E}$ of *real-time energy functions*_is the subsemiring_of ${\mathcal_F}$ generated by_${\mathcal A}$, [*i.e.,*]{} the subset of ${\mathcal_F}$ inductively defined_by
- ${\mathcal A}\subseteq {\mathcal_E}$,
-_ if_$f,_g\in_{\mathcal E}$,_then $f{\mathrel{\mathop{\triangleright}}}g\in {\mathcal_E}$_and $f\vee_g\in_{\mathcal E}$.
We will show below that_${\mathcal_E}$ is locally closed, which entails that_for each $f\in {\mathcal_E}$,_also $f^*\in {\mathcal E}$,_hence ${\mathcal E}$ indeed encompasses_all RTEFs.
\[le:locallyclosed\] For every $f\in {\mathcal_E}$ there_exists $N\ge_0$ so that $f^*= \bigvee_{ n= 0}^N f^n$.
By distributivity, we can_write $f$ as a finite supremum_$ |
& Thielemann, F. K. 1994, ApJ, 431, 742 Demorest P. et al. 2010, Nature 467:1081-1083 Doneva, D. D.,et al. 2013, PhRvD, 88, 044052 Cook, G. B., Shapiro, S. L., & Teukolsky, S. A. 1992,, 398, 203 Danielewicz, P., & Lee, J. 2009, Nuclear Physics A, 818, 36 Douchin, F., & Haensel, P. 2001,, 380, 151 Duan, K. K., Jin, Z. P., Zhang, F. W., et al. 2019, ApJL, 876, L28 Duncan, R. C., & Thompson, C. 1992, ApJL, 392, L9 Eichler, D., Livio, M., Piran, T., & Schramm, D., 1989, Nature, 340, 126 Endrizzi, A., Ciolfi, R., Giacomazzo, B., Kastaun, W., & Kawamura, T. 2016, Classical and Quantum Gravity, 33, 164001 Engvik, L., Osnes, E., Hjorth-Jensen, M., Bao, G., & Ostgaard, E. 1996,, 469, 794 Fan, Y. Z., Wu, X. F., & Wei, D. M. 2013, PhRvD, 88, 067304 Fan, Y. Z., & Xu, D. 2006, MNRAS, 372, L19 Foucart, F., et al. 2016, PhRvD 93, 044019 Friedrich, J., & Reinhard, P.-G. 1986,, 33, 335 Fryer, C. L., et al. 2015, ApJ 812(1): 24. Fujibayashi, S., et al 2018, ApJ. 860, 64 Gaitanos, T., Di Toro, M., Typel, S., et al. 2004, Nuclear Physics A, 732, 24 Gao, H., Zhang, B., & Lü, H.-J. 2016, PhRvD, 93, 044065 Gao, W. H., & Fan, Y. Z. 2006, ChJAA, 6, 513 Gill, R., et al. | & Thielemann, F. K. 1994, ApJ, 431, 742 Demorest P. et al. 2010, Nature 467:1081 - 1083 Doneva, D. D.,et al. 2013, PhRvD, 88, 044052 Cook, G. B., Shapiro, S. L., & Teukolsky, S. A. 1992, , 398, 203 Danielewicz, P., & Lee, J. 2009, Nuclear Physics A, 818, 36 Douchin, F., & Haensel, P. 2001, , 380, 151 Duan, K. K., Jin, Z. P., Zhang, F. W., et al. 2019, ApJL, 876, L28 Duncan, R. C., & Thompson, C. 1992, ApJL, 392, L9 Eichler, D., Livio, M., Piran, T., & Schramm, D., 1989, Nature, 340, 126 Endrizzi, A., Ciolfi, R., Giacomazzo, B., Kastaun, W., & Kawamura, T. 2016, Classical and Quantum Gravity, 33, 164001 Engvik, L., Osnes, E., Hjorth - Jensen, M., Bao, G., & Ostgaard, E. 1996, , 469, 794 Fan, Y. Z., Wu, X. F., & Wei, D. M. 2013, PhRvD, 88, 067304 Fan, Y. Z., & Xu, D. 2006, MNRAS, 372, L19 Foucart, F., et al. 2016, PhRvD 93, 044019 Friedrich, J., & Reinhard, P.-G. 1986, , 33, 335 Fryer, C. L., et al. 2015, ApJ 812(1 ): 24. Fujibayashi, S., et al 2018, ApJ. 860, 64 Gaitanos, T., Di Toro, M., Typel, S., et al. 2004, Nuclear Physics A, 732, 24 Gao, H., Zhang, B., & Lü, H.-J. 2016, PhRvD, 93, 044065 Gao, W. H., & Fan, Y. Z. 2006, ChJAA, 6, 513 Gill, R., et al. | & Tjielemann, F. K. 1994, ApJ, 431, 742 Demurest P. et al. 2010, Uqture 467:1081-1083 Donevz, D. D.,et xl. 2013, PhRvD, 88, 044052 Cook, G. B., Shapiro, D. L., & Teujolsky, S. A. 1992,, 398, 203 Danielewicx, P., & Lee, U. 2009, Nucleae Phbsics A, 818, 36 Douchii, F., & Haekfel, L. 2001,, 380, 151 Dbai, K. K., Jin, Z. P., Zmang, F. W., et al. 2019, ApJL, 876, L28 Dgnzau, R. C., & Thompson, C. 1992, ApJL, 392, L9 Eichler, D., Livio, K., Oiran, T., & Schraim, D., 1989, Nathge, 340, 126 Endrizzi, A., Ciolfi, R., Giacomzzzo, B., Nastaun, W., & Kaeamura, T. 2016, Classical and Quajtum Gravity, 33, 164001 Engvik, P., Osnes, E., Hhortr-Hensen, M., Bao, G., & Ostgaagb, E. 1996,, 469, 794 Fan, Y. Z., Wu, X. F., & Wei, D. M. 2013, PhRvD, 88, 067304 Fan, Y. Z., & Xu, D. 2006, MNRQS, 372, L19 Xoucart, F., ev al. 2016, IhRvD 93, 044019 Friedvpch, J., & Seinharc, P.-G. 1986,, 33, 335 Fryer, C. L., xt ao. 2015, ApJ 812(1): 24. Fujibayashi, V., et al 2018, ApJ. 860, 64 Gaijanos, T., Di Tkro, M., Typel, S., et ql. 2004, Nuclgar Pvysizw A, 732, 24 Fap, G., Zhanh, B., & Lü, H.-J. 2016, PhDvD, 93, 044065 Gao, W. H., & Fan, Y. Z. 2006, ChJAA, 6, 513 Dpkl, R., et al. | & Thielemann, F. K. 1994, ApJ, 431, P. al. 2010, 467:1081-1083 Doneva, D. 044052 G. B., Shapiro, L., & Teukolsky, A. 1992,, 398, 203 Danielewicz, P., Lee, J. 2009, Nuclear Physics A, 818, 36 Douchin, F., & Haensel, P. 380, 151 Duan, K. K., Jin, Z. P., Zhang, F. W., et al. ApJL, L28 R. & Thompson, C. 1992, ApJL, 392, L9 Eichler, D., Livio, M., Piran, T., & Schramm, D., Nature, 340, 126 Endrizzi, A., Ciolfi, R., Giacomazzo, Kastaun, W., & Kawamura, 2016, Classical and Quantum Gravity, 164001 L., Osnes, Hjorth-Jensen, Bao, & Ostgaard, E. 469, 794 Fan, Y. Z., Wu, X. F., & Wei, D. M. 2013, PhRvD, 88, 067304 Fan, Z., & 2006, MNRAS, L19 F., al. 2016, PhRvD Friedrich, J., & Reinhard, P.-G. 1986,, C. L., et al. 2015, ApJ 812(1): 24. S., et 2018, ApJ. 860, 64 Gaitanos, T., Toro, M., Typel, S., et al. 2004, Nuclear A, 732, 24 Gao, H., Zhang, B., & Lü, H.-J. 2016, PhRvD, 93, 044065 Gao, & Fan, Y. Z. ChJAA, 6, 513 R., al. | & Thielemann, F. K. 1994, ApJ, 431, 742 Demorest P. eT al. 2010, Nature 467:1081-1083 DOneva, d. D.,eT al. 2013, phrvD, 88, 044052 COok, G. b., Shapiro, S. L., & TeukOLsky, s. A. 1992,, 398, 203 Danielewicz, P., & Lee, J. 2009, NuclEar PhYsICs A, 818, 36 DOUcHin, F., & HAensel, P. 2001,, 380, 151 dUaN, k. k., JiN, Z. p., ZHanG, F. w., Et Al. 2019, ApJl, 876, L28 DUncan, R. C., & thompson, C. 1992, APJL, 392, l9 EIchler, D., Livio, m., piRan, T., & SchramM, D., 1989, NAture, 340, 126 EndrizzI, A., CIolfi, R., giAcoMAzzo, B., kasTaun, W., & kawamuRA, T. 2016, ClasSical and QUaNTum GraVIty, 33, 164001 EngvIK, l., OSnes, e., Hjorth-Jensen, M., Bao, g., & osTGaard, E. 1996,, 469, 794 Fan, Y. Z., Wu, X. f., & Wei, D. M. 2013, phrVD, 88, 067304 fAN, Y. Z., & xu, D. 2006, mNRAS, 372, L19 FoucArT, F., et aL. 2016, phRvD 93, 044019 FrIEdRICH, J., & REInhard, P.-G. 1986,, 33, 335 Fryer, c. L., et al. 2015, ApJ 812(1): 24. FuJIbaYashi, S., Et Al 2018, APj. 860, 64 GaitaNos, T., DI TORo, M., typel, S., et al. 2004, NUcleAr Physics a, 732, 24 Gao, H., ZHAng, B., & Lü, H.-j. 2016, phRvD, 93, 044065 GaO, W. H., & Fan, y. Z. 2006, CHJAa, 6, 513 GilL, r., eT aL. | & Thielemann, F. K. 1994, ApJ, 431, 742Dem ore st P.et a l. 2010, Natur e 467 :1081-1083 Doneva, D.D.,et a l . 20 1 3, PhRv D, 88,0 44 0 5 2 C oo k, G. B . ,Shapi ro, S. L., & Teukols ky, S . A. 1992,,3 98 , 203 Dani ele wicz, P., &Lee , J. 2 00 9,N uclea r P hysic s A, 8 1 8, 36Douchin,F. , & Hae n sel, P. 2 00 1,,380, 151 Duan, K. K. , Jin, Z. P., Z hang,F. W. , etal. 2019, ApJ L, 876, L28 Dun c an , R . C . , & Thompson, C. 1992, A p JL, 392,L9 Ei c hler,D., L iv i o,M., Piran,T.,& Schramm , D.,1 989, Na t ure, 34 0, 126 En dri zzi, A. ,Cio lf i , R . ,Gia c oma zzo, B., K as taun, W., & K awam ura , T. 2016 , Classical a ndQuan t umGravi ty, 3 3, 1 64 001 E ngvik, L.,Os nes, E., Hjorth -Jen sen, M.,Bao ,G., & Ostg a ard, E . 1 996 ,, 469, 794 Fa n , Y .Z . , W u, X. F., & Wei, D .M . 2 013, PhR vD, 88 , 0 67 3 04 Fan,Y. Z. , &X u , D.2006 , M NRAS, 37 2, L19 Fo uc art, F. ,et al. 2 016 , P hRvD9 3, 0 44019Friedric h, J. , & Reinhard, P . -G. 1986,, 33 , 3 3 5 F r yer, C. L., et al. 201 5 , Ap J 81 2 (1 ):2 4. Fu jibay as h i, S., et al 2018, ApJ .860, 6 4 Gai tanos, T., Di Toro, M., T y pel, S., eta l. 2004, NuclearPhysi cs A, 732, 24 Gao,H., Z hang, B. , & Lü, H . - J. 2016, Ph RvD , 9 3,0 4 40 65 Gao, W. H. , & Fa n, Y. Z.200 6, ChJA A,6,513 Gi ll , R., etal. | &_Thielemann, F._K. 1994, ApJ, 431,_742 Demorest_P._et al._2010,_Nature 467:1081-1083 Doneva,_D. D.,et al._2013, PhRvD, 88, 044052_Cook, G. B., Shapiro,_S. L.,_& Teukolsky, S. A. 1992,, 398, 203 Danielewicz, P., & Lee, J. 2009, Nuclear Physics A, 818,_36_Douchin, F.,_&_Haensel,_P. 2001,, 380, 151 Duan, K._K., Jin, Z. P., Zhang,_F. W.,_et al. 2019, ApJL, 876, L28 Duncan, R._C.,_& Thompson, C._1992, ApJL, 392, L9 Eichler, D., Livio, M., Piran,_T., & Schramm, D., 1989, Nature,_340, 126 Endrizzi,_A.,_Ciolfi,_R., Giacomazzo, B., Kastaun,_W., & Kawamura, T. 2016, Classical and_Quantum Gravity, 33, 164001 Engvik, L.,_Osnes, E., Hjorth-Jensen, M., Bao, G., &_Ostgaard, E. 1996,, 469, 794 Fan, Y._Z., Wu, X. F., &_Wei, D._M. 2013, PhRvD, 88, 067304_Fan, Y. Z.,_& Xu,_D. 2006, MNRAS,_372, L19 Foucart, F., et al._2016, PhRvD 93,_044019 Friedrich, J., & Reinhard, P.-G. 1986,,_33,_335 Fryer, C._L.,_et_al. 2015,_ApJ 812(1): 24._Fujibayashi,_S., et_al_2018, ApJ. 860, 64 Gaitanos, T.,_Di_Toro, M., Typel, S., et al. 2004, Nuclear_Physics A, 732, 24_Gao,_H., Zhang, B., &_Lü, H.-J. 2016, PhRvD, 93,_044065 Gao, W. H., & Fan,_Y. Z._2006, ChJAA,_6, 513 Gill, R., et al. |
mbox{M\"ob}_{d}
\ = \ \frac{1}{d} \Bigl( \ f \ { \widetilde{{ \framebox[7pt]{$\star$} }} }\ G_{d} \ { \widetilde{{ \framebox[7pt]{$\star$} }} }\
( d \mbox{M\"ob}_{d} \circ D_{1/d}) \ \Bigr)
\ \ \mbox{ (by Eqn.(1.23) )}$$ $$= \ \frac{1}{d} ( f \ { \widetilde{{ \framebox[7pt]{$\star$} }} }\ H_{d} )
\ \ \mbox{ (by the definition of $H_{d}$ in Section 1.7).}$$ So we obtain the equation: $$R_{A_{1}P_{1}, \ldots, A_{r}P_{i}, \ldots, A_{s}P_{d}} \ = \
\frac{1}{d} ( f \ { \widetilde{{ \framebox[7pt]{$\star$} }} }\ H_{d} ),$$ out of which (2.7) is obtained in the same way as (2.6) was obtained from (2.15). [**QED**]{}
$\ $
[**2.12 Remark.**]{} The proof of Theorem 2.10 obtains the Equations (2.15) and (2.16), stronger than what was originally stated, and which show better the significance of the series $f \ { \widetilde{{ \framebox[7pt]{$\star$} }} }\ G_{d}$ and $f \ { \widetilde{{ \framebox[7pt]{$\star$} }} }\ H_{d}$.
$\ $
$\ $
$\ $
We will concentrate on applications to a family $A_{1}, \ldots, A_{s}$ of selfadjoint $d \times d$ matrices over a $*$-probability space ${ ( {\cal A}, \varphi ) }$. By keeping in mind the motivating example from [@V2], it is of particular interest to put into evidence situations where the family $A_{1}, \ldots, A_{s}$ is free in $(M_{d} ( { {\cal A} }), \varphi_{d} )$, and where the individual R-transform of each of $A_{1}, \ldots, A_{s | mbox{M\"ob}_{d }
\ = \ \frac{1}{d } \Bigl (\ f \ { \widetilde { { \framebox[7pt]{$\star$ } } } } \ G_{d } \ { \widetilde { { \framebox[7pt]{$\star$ } } } } \
(d \mbox{M\"ob}_{d } \circ D_{1 / d }) \ \Bigr)
\ \ \mbox { (by Eqn.(1.23) ) } $ $ $ $ = \ \frac{1}{d } (f \ { \widetilde { { \framebox[7pt]{$\star$ } } } } \ H_{d })
\ \ \mbox { (by the definition of $ H_{d}$ in Section 1.7).}$$ So we obtain the equation: $ $ R_{A_{1}P_{1 }, \ldots, A_{r}P_{i }, \ldots, A_{s}P_{d } } \ = \
\frac{1}{d } (f \ { \widetilde { { \framebox[7pt]{$\star$ } } } } \ H_{d }), $ $ out of which (2.7) is prevail in the like way as (2.6) was obtained from (2.15). [ * * QED * * ] { }
$ \ $
[ * * 2.12 Remark. * * ] { } The validation of Theorem 2.10 obtains the Equations (2.15) and (2.16), hard than what was originally stated, and which indicate better the significance of the serial $ f \ { \widetilde { { \framebox[7pt]{$\star$ } } } } \ G_{d}$ and $ f \ { \widetilde { { \framebox[7pt]{$\star$ } } } } \ H_{d}$.
$ \ $
$ \ $
$ \ $
We will condense on applications to a syndicate $ A_{1 }, \ldots, A_{s}$ of selfadjoint $ d \times d$ matrices over a $ * $ -probability quad $ { ({ \cal A }, \varphi) } $. By keeping in mind the motivating example from [ @V2 ], it is of particular sake to put into evidence situations where the family $ A_{1 }, \ldots, A_{s}$ is barren in $ (M_{d } ({ { \cal A } }), \varphi_{d }) $, and where the individual R - transform of each of $ A_{1 }, \ldots, A_{s | mbod{M\"ob}_{d}
\ = \ \frac{1}{d} \Bigl( \ f \ { \widetilde{{ \framgbix[7pt]{$\ster$} }} }\ G_{d} \ { \widetklde{{ \framebox[7pt]{$\star$} }} }\
( d \mboe{M\"ob}_{e} \cirx D_{1/d}) \ \Bigr)
\ \ \mbox{ (by Eqv.(1.23) )}$$ $$= \ \frac{1}{f} ( f \ { \wudetmlde{{ \framebox[7pt]{$\svzr$} }} }\ H_{d} )
\ \ \mbos{ (by chx definition of $H_{d}$ in Secdion 1.7).}$$ So we obdakn the equation: $$R_{A_{1}P_{1}, \ldots, A_{r}P_{i}, \ldots, W_{s}P_{d}} \ = \
\fgac{1}{d} ( f \ { \widejilde{{ \sramsbox[7pt]{$\star$} }} }\ H_{d} ),$$ out of which (2.7) is kbtainev in the same wsy as (2.6) was obtained from (2.15). [**EED**]{}
$\ $
[**2.12 Remark.**]{} The proof lf Theorem 2.10 obtwuns the Equagions (2.15) and (2.16), stronger tgan what was originally stated, xnd wkich show bgcrer jhe significence os the series $f \ { \widatilde{{ \gramebox[7pt]{$\star$} }} }\ J_{d}$ abd $f \ { \widetilde{{ \framxbox[7pt]{$\star$} }} }\ H_{d}$.
$\ $
$\ $
$\ $
We will conweutrate on applicationw ro a xaminy $A_{1}, \odogs, Z_{s}$ or selfwdjkint $d \timss d$ matricws over a $*$-probabiliuy fisce ${ ( {\cal A}, \barphi ) }$. By keeping in mind the motivating examkle fdom [@V2], it is of particulqr interest to put injo evidencq situations where the family $A_{1}, \ldots, A_{s}$ is free hn $(M_{d} ( { {\cco A} }), \xqrohi_{d} )$, and where the individual R-transform of qzcn pf $A_{1}, \ldots, A_{s | mbox{M\"ob}_{d} \ = \ \frac{1}{d} \Bigl( \ { \framebox[7pt]{$\star$} }} G_{d} \ { d \circ D_{1/d}) \ \ \ \mbox{ Eqn.(1.23) )}$$ $$= \ \frac{1}{d} ( \ { \widetilde{{ \framebox[7pt]{$\star$} }} }\ H_{d} ) \ \ \mbox{ (by the of $H_{d}$ in Section 1.7).}$$ So we obtain the equation: $$R_{A_{1}P_{1}, \ldots, A_{r}P_{i}, A_{s}P_{d}} = \frac{1}{d} f \ { \widetilde{{ \framebox[7pt]{$\star$} }} }\ H_{d} ),$$ out of which (2.7) is obtained in same way as (2.6) was obtained from (2.15). $\ $ [**2.12 Remark.**]{} proof of Theorem 2.10 obtains Equations and (2.16), than was stated, and which better the significance of the series $f \ { \widetilde{{ \framebox[7pt]{$\star$} }} }\ G_{d}$ and $f \ \widetilde{{ \framebox[7pt]{$\star$} H_{d}$. $\ $\ $\ We will concentrate to a family $A_{1}, \ldots, A_{s}$ \times d$ matrices over a $*$-probability space ${ {\cal A}, ) }$. By keeping in mind motivating example from [@V2], it is of particular to put into evidence situations where the family $A_{1}, \ldots, A_{s}$ is free in $(M_{d} {\cal A} }), \varphi_{d} and where the R-transform each $A_{1}, A_{s | mbox{M\"ob}_{d}
\ = \ \frac{1}{d} \Bigl( \ f \ { \widetilDe{{ \framebox[7Pt]{$\staR$} }} }\ G_{d} \ { \WidEtIlde{{ \FramEbox[7pt]{$\star$} }} }\
( d \mboX{m\"ob}_{d} \Circ D_{1/d}) \ \Bigr)
\ \ \mbox{ (by Eqn.(1.23) )}$$ $$= \ \fraC{1}{d} ( f \ { \wiDeTIlde{{ \FRaMebox[7Pt]{$\star$} }} }\ H_{D} )
\ \ \MbOX{ (By tHe DeFinItIOn Of $H_{d}$ iN SeCtion 1.7).}$$ So We obtain thE eqUaTion: $$R_{A_{1}P_{1}, \ldots, a_{R}P_{I}, \ldots, A_{s}P_{d}} \ = \
\FraC{1}{d} ( f \ { \widetilde{{ \FraMebox[7pT]{$\sTar$} }} }\ h_{D} ),$$ out oF whIch (2.7) is ObtainED in the Same way as (2.6) WaS ObtainED from (2.15). [**QEd**]{}
$\ $
[**2.12 rEmArk.**]{} THe proof of Theorem 2.10 oBTaINs the Equations (2.15) And (2.16), strOnGEr THAn wHat Was originaLlY statED, and whiCH sHOW BetTEr the significAnce of the seRIes $F \ { \widetIlDe{{ \fRAmebox[7Pt]{$\staR$} }} }\ G_{D}$ And $F \ { \widetilde{{ \fRameBox[7pt]{$\star$} }} }\ h_{d}$.
$\ $
$\ $
$\ $
We wiLL concenTRate on aPplicaTioNs tO a faMIlY $A_{1}, \LdoTs, a_{S}$ of SElFadJOinT $d \times d$ MaTrIces oVer a $*$-PROBAbilIty SpacE ${ ( {\cal A}, \Varphi ) }$. By keepiNg iN minD The MotivAting ExamPlE from [@v2], it is oF partIcUlar interest to pUt inTo evidencE siTuAtiOnS wherE The famIly $a_{1}, \ldOts, A_{s}$ is Free in $(M_{D} ( { {\Cal a} }), \vARPHi_{D} )$, and where the indiviDuAL r-tRansform Of each OF $A_{1}, \LdOTs, A_{s | mbox{M\"ob}_{d}
\ = \ \fr ac{1}{d} \ Bigl( \f \ { \wi deti lde{{ \framebo x [7pt ]{$\star$} }} }\ G_{d} \ {\w i deti l de {{ \f ramebox [ 7p t ] {$\ st ar $}}} }\
( d \m box{M\" ob}_{d} \c irc D _{1/d}) \ \B i gr )
\ \ \mbo x{(by Eqn.(1.2 3))}$$ $ $= \\ frac{ 1}{ d} (f \ {\ wideti lde{{ \fr am e box[7p t ]{$\sta r $ }}} } \ H_{d} )
\ \ \mb o x{ (by the defini tion o f$ H_ { d }$inSection 1. 7) .}$$S o we ob t ai n t hee quation: $$R_ {A_{1}P_{1} , \l dots,A_ {r} P _{i},\ldot s, A_{ s}P_{d}} \= \
\frac{1} {d} (f \ { \w i detilde {{ \fr ame box [7pt ] {$ \s tar $} }}} \H_{ d } ) ,$$ outof w hich(2.7 ) i s obt ain ed i n the same way as(2. 6) w a s o btain ed fr om ( 2. 15).[**QED **]{}
$\ $
[**2.12 R emar k.**]{} T hepr oof o f The o rem 2. 10obt ains th e Equat i ons ( 2 . 1 5) and (2.16), stron ge r th an whatwas or i gi na l ly state d, an d wh i c h sho w be t te r the si gnific a nc eof these ries $ f\ { \w ideti l de{{ \fram ebox[7pt ]{$\s t ar$} }} }\ G_{ d }$ and $f \ { \w i d et i lde{ { \ framebox[7p t]{$ \ star $} } } } \ H _ {d}$.
$\$$\ $
$\ $
We will co nc entrat e onapplicationsto a famil y $ A_{1}, \ ldot s ,A _{s}$ of selfa djoin t $d \time s d$ matr icesover a $ *$-probab i l ity spac e $ { ( {\ cal A }, \varphi ) }$ . By k ee ping in mi nd themot iva tin g e xa mple from [@V2],it i sof pa rticu l ar inter es t t oput into eviden ce si tuat io ns whe re thef am i l y $A _{ 1} , \l dot s, A_{s }$ i s fr ee in $ (M_{d} ({ { \ calA} } ), \var phi_{d} )$, a nd where the i ndi vidual R -transfo rm of each of $A_{1}, \ l dots, A _{s | mbox{M\"ob}_{d}
\_= \_\frac{1}{d} \Bigl( \ f_\ {_\widetilde{{_\framebox[7pt]{$\star$} }}_}\_G_{d} \ {_\widetilde{{ \framebox[7pt]{$\star$} }}_}\
( d \mbox{M\"ob}_{d}_\circ D_{1/d}) \_\Bigr)
\_\ \mbox{ (by Eqn.(1.23) )}$$ $$= \ \frac{1}{d} ( f \ { \widetilde{{ \framebox[7pt]{$\star$}_}}_}\ H_{d}_)
\_\_\mbox{ (by the definition of_$H_{d}$ in Section 1.7).}$$ So_we obtain_the equation: $$R_{A_{1}P_{1}, \ldots, A_{r}P_{i}, \ldots, A_{s}P_{d}} \_=_\
\frac{1}{d} (_f \ { \widetilde{{ \framebox[7pt]{$\star$} }} }\ H_{d} ),$$_out of which (2.7) is obtained_in the same_way_as_(2.6) was obtained from_(2.15). [**QED**]{}
$\ $
[**2.12 Remark.**]{} The proof_of Theorem 2.10 obtains the Equations_(2.15) and (2.16), stronger than what was_originally stated, and which show better_the significance of the series_$f \_{ \widetilde{{ \framebox[7pt]{$\star$} }} }\_G_{d}$ and $f_\ {_\widetilde{{ \framebox[7pt]{$\star$} }}_}\ H_{d}$.
$\ $
$\ $
$\ $
We will_concentrate on applications_to a family $A_{1}, \ldots, A_{s}$_of_selfadjoint $d \times_d$_matrices_over a_$*$-probability space ${_(_{\cal A},_\varphi_) }$. By keeping in mind_the_motivating example from [@V2], it is of_particular interest to put_into_evidence situations where the_family $A_{1}, \ldots, A_{s}$ is_free in $(M_{d} ( { {\cal_A} }),_\varphi_{d} )$,_and where the individual R-transform of each of $A_{1}, \ldots, A_{s |
with FH order at low fields \[Fig. \[fig:PD\](a)\]. By increasing $J_Q$ and moving away from half-filling, the coexistence region can be made to disappear, replaced by a direct transition between the FH and AFQ phases \[Fig. \[fig:PD\](b,c)\], which is reminiscent of the experimental result, with FH order explaining the heavy Fermi liquid region in intermediate fields.
![Three example mean-field phase diagrams for FH (red) and AFQ (blue) orders in magnetic field, for three different strengths of quadrupolar exchange coupling: (a) $J_Q\!=\!0.38t$, (b) $J_Q\!=\!1.6t$, (c) $J_Q\!=\!2t$. The transition to FH order, $T_K$ initially rises with magnetic field and is then suppressed, ultimately becoming a first-order transition in field. Other parameters for (a): $t\!=\!1$, $V\!=\!0.8$, $\Delta E\!=\!5.5$, $\Delta\!=\!4.8$, $n_{c,0}\!=\!1.9$; other parameters for (b),(c): $t\!=\!1$, $V\!=\!0.9$, $\Delta E\!=\!5.5$, $\Delta\!=\!20$, $n_{c,0}\!=\!1.6$. \[fig:PD\]](plot_allPDvert.pdf "fig:")-0.2cm
Due to the linear coupling of the magnetic field to the FH moments, we expect magnetic field to initially favor FH order, leading to increased hybridization and higher $T_K$. This increase is seen in all three subplots of Fig. \[fig:PD\], most noticeably in part (b). However, as stronger fields split the $\Gamma_3$ doublet, the Kondo screening and thus FH order are eventually destroyed, as seen in the mean-field calculation. AFQ order is also suppressed as a function of $B$, leading to a first-order transition between it and FH order at a critical field $B_c$. Note that while the mean-field theory always finds a first order transition between FH and PQ states at low temperatures and high fields, we do not expect this to necessarily hold beyond the mean-field level.
Landau theory \[sec:landau\]
| with FH order at low fields \[Fig. \[fig: PD\](a)\ ]. By increasing $ J_Q$ and move aside from half - filling, the coexistence region can be made to vanish, replaced by a direct conversion between the FH and AFQ phase \[Fig. \[fig: PD\](b, c)\ ], which is reminiscent of the experimental consequence, with FH order explaining the big Fermi liquid region in intermediate field.
! [ Three example mean - sphere phase diagrams for FH (red) and AFQ (blue) order in magnetic field, for three different forte of quadrupolar exchange coupling: (a) $ J_Q\!=\!0.38t$, (b) $ J_Q\!=\!1.6t$, (c) $ J_Q\!=\!2t$. The transition to FH ordering, $ T_K$ initially rises with magnetic field and is then suppressed, ultimately becoming a inaugural - order transition in field. Other parameters for (a ): $ t\!=\!1 $, $ V\!=\!0.8 $, $ \Delta E\!=\!5.5 $, $ \Delta\!=\!4.8 $, $ n_{c,0}\!=\!1.9 $; other parameters for (b),(c ): $ t\!=\!1 $, $ V\!=\!0.9 $, $ \Delta E\!=\!5.5 $, $ \Delta\!=\!20 $, $ n_{c,0}\!=\!1.6$. \[fig: PD\]](plot_allPDvert.pdf " fig:")-0.2 cm
Due to the linear coupling of the magnetic field to the FH moments, we expect magnetic field to initially favor FH ordering, leading to increased hybridization and high $ T_K$. This increase is seen in all three subplots of Fig. \[fig: PD\ ], most noticeably in function (bacillus). However, as stronger fields burst the $ \Gamma_3 $ doublet, the Kondo screening and thus FH order are eventually destroy, as seen in the mean - field calculation. AFQ order is besides suppressed as a routine of $ B$, head to a first - order transition between it and FH holy order at a critical field $ B_c$. Note that while the mean - field theory always finds a first order conversion between FH and PQ states at depleted temperatures and eminent fields, we do not ask this to necessarily oblige beyond the mean - sphere level.
Landau theory \[sec: landau\ ] | wihh FH order at low fieldr \[Fig. \[fig:PD\](a)\]. By nbcreasmng $J_Q$ znd movivg away from half-filling, the ciexisuvnce region can be maae to diswppear, rwplared by a direct vdansition betscen tke FH and AFQ phsses \[Fig. \[fhg:PD\](b,c)\], which iv feliniscent of the experimental resule, with GH order explainyng uhe heabj Nermi liquid region in intermedizte fiends.
![Three examlle mean-field phase diagrals flr FH (red) and AFQ (hlue) orders in iqgnetic fiela, for thret bifferent sjrengths of quadrupolar exchange coupking: (a) $J_Q\!=\!0.38t$, (b) $J_Q\!=\!1.6h$, (c) $J_Q\!=\!2t$. The vransinion to FH order, $T_K$ ititiallu rises with msgnxtic field and is then su'pressed, ultimately fecoming d yirst-order transition ib fiend. Odher parxmeuerx ror (a): $h\!=\!1$, $V\!=\!0.8$, $\Delta E\!=\!5.5$, $\Dslta\!=\!4.8$, $n_{c,0}\!=\!1.9$; othwr parameters for (b),(v): $e\!=\!1$, $V\!=\!0.9$, $\Delta E\!=\!5.5$, $\Demta\!=\!20$, $n_{c,0}\!=\!1.6$. \[fyg:PD\]](plot_allPDvert.pdf "fig:")-0.2cm
Due to the litead coupling of the magneric field to the FH mlments, we expect magnetic field to initially favor FH ordes, leavivg uo incrdqsfd hybridization and higher $T_K$. This increase ya xevn in all three smbplots of Fig. \[fig:LD\], mpft noticeably in paxf (g). However, as stronher fiejds split the $\Damms_3$ doublet, the Kondo screenibg and thus YH irder are eventualpy destroyeb, as sgen in the mean-field calculatnon. AFS order is wlso suppddssed as a functkon ox $B$, leading to a first-ordew transitmon bztween ig anc FH owder at a fritigdl field $B_c$. Note tjat wkile dhe mean-fifld theory always finds a first order transitoot bvtween FH and IQ states at lor temperatures and hiyh fieuds, we do hot expxct this to gecessarily hmpd beyond thx mean-fiejd lwvel.
Oandau gfeory \[sec:landai\]
| with FH order at low fields \[Fig. increasing and moving from half-filling, the to replaced by a transition between the and AFQ phases \[Fig. \[fig:PD\](b,c)\], which reminiscent of the experimental result, with FH order explaining the heavy Fermi liquid in intermediate fields. ![Three example mean-field phase diagrams for FH (red) and AFQ orders magnetic for different strengths of quadrupolar exchange coupling: (a) $J_Q\!=\!0.38t$, (b) $J_Q\!=\!1.6t$, (c) $J_Q\!=\!2t$. The transition to FH $T_K$ initially rises with magnetic field and is suppressed, ultimately becoming a transition in field. Other parameters (a): $V\!=\!0.8$, $\Delta $\Delta\!=\!4.8$, other for (b),(c): $t\!=\!1$, $\Delta E\!=\!5.5$, $\Delta\!=\!20$, $n_{c,0}\!=\!1.6$. \[fig:PD\]](plot_allPDvert.pdf "fig:")-0.2cm Due to the linear coupling of the magnetic field to the moments, we field to favor order, to increased hybridization $T_K$. This increase is seen in of Fig. \[fig:PD\], most noticeably in part (b). as stronger split the $\Gamma_3$ doublet, the Kondo and thus FH order are eventually destroyed, as in the mean-field calculation. AFQ order is also suppressed as a function of $B$, leading first-order transition between it FH order at critical $B_c$. that the mean-field always finds a first order transition between FH and PQ states low temperatures and high fields, we do not expect this hold the mean-field level. theory \[sec:landau\] | with FH order at low fields \[Fig. \[Fig:PD\](a)\]. By inCreasIng $j_Q$ aNd MoviNg awAy from half-fillINg, thE coexistence region can bE made To DIsapPEaR, replAced by a DIrECT trAnSiTioN bETwEen thE FH And AFQ pHases \[Fig. \[fiG:PD\](B,c)\], Which is reminIScEnt of the exPerImental resulT, wiTh FH orDeR exPLainiNg tHe heaVy FermI Liquid Region in iNtERmediaTE fields.
![tHReE exaMple mean-field phasE DiAGrams for FH (red) aNd AFQ (bLuE) OrDERs iN maGnetic fielD, fOr thrEE differENt STREngTHs of quadrupolAr exchange cOUplIng: (a) $J_Q\!=\!0.38T$, (b) $j_Q\!=\!1.6t$, (C) $j_Q\!=\!2t$. The TransItIOn tO FH order, $T_K$ iNitiAlly rises With maGNetic fiELd and is Then suPprEssEd, ulTImAtEly BeCOmiNG a FirST-orDer transItIoN in fiEld. OTHER ParaMetErs fOr (a): $t\!=\!1$, $V\!=\!0.8$, $\delta E\!=\!5.5$, $\Delta\!=\!4.8$, $n_{c,0}\!=\!1.9$; OthEr paRAmeTers fOr (b),(c): $t\!=\!1$, $v\!=\!0.9$, $\DelTa e\!=\!5.5$, $\DeltA\!=\!20$, $n_{c,0}\!=\!1.6$. \[fig:pD\]](ploT_aLlPDvert.pdf "fig:")-0.2cM
Due To the lineAr cOuPliNg Of the MAgnetiC fiEld To the FH Moments, WE exPeCT MAgNetic field to initiaLlY FAvOr FH ordeR, leadiNG tO iNCreased hYbRidIzatION and hIgheR $t_K$. this incrEase is SEeN iN all thrEe SubploTs Of FIg. \[fIg:PD\], mOSt noTiceabLy in part (B). HoweVEr, as stronger fiELds split the $\GaMMa_3$ DOUbLEt, thE KoNdo screeninG and THus Fh ordER aRe eVEntuaLly deStROyED, as seen in the mean-fieLd CalculAtion. aFQ order is alsO suppresseD AS A functioN of $B$, LEaDIng to a first-ordEr traNsition betWEen it and fH ordEr at a criTical fielD $b_C$. Note thaT whIle The MeaN-FIeLd theory alwayS FInds A fIrst ordEr tRansitiOn bEtwEen fH aNd pQ states aT low tempErAtUrEs And High fIElds, we do NoT exPeCt tHis to NEcessaRily hOld bEyOnD The Mean-fieLD lEVEl.
LaNdAu TheoRy \[sEc:LandaU\]
| with FH order at low fiel ds \[Fig.\[fig :PD \]( a) \].By i ncreasing $J_Q $ and moving away from half -fill in g , th e c oexis tence r e gi o n ca nbe ma de to disa ppe ar, rep laced by a di re ct transitio n b etween the FH and AFQ pha ses \[Fig .\[f i g:PD\ ](b ,c)\] , whic h is re miniscent o f the e x perimen t a lresu lt, with FH order ex p laining the he avy Fe rm i l i q uid re gion in in te rmedi a te fiel d s. ! [Th r ee example me an-field ph a sediagra ms fo r FH (r ed) a nd AFQ (blue) ord ersin magnet ic fie l d, fort hree di fferen t s tre ngth s o fqua dr u pol a rexc h ang e coupli ng :(a) $ J_Q\ ! = \ ! 0.38 t$, (b) $J_Q \!=\!1.6t$, ( c)$J_Q \ !=\ !2t$. Thetran si tionto FHorder ,$T_K$ initially ris es with m agn et icfi eld a n d is t hen su ppresse d, ulti m ate ly b e co ming a first-order t r a ns ition in field . O th e r parame te rsfor( a ): $t \!=\ ! 1$ , $V\!=\ !0.8$, $\ De lta E\! =\ !5.5$, $ \De lta \!=\! 4 .8$, $n_{c ,0}\!=\! 1.9$; other paramete r s for (b),(c) : $ t \ != \ !1$, $V \!=\!0.9$,$\De l ta E \!=\ ! 5. 5$, $\Del ta\!= \! 2 0$ , $n_{c,0}\!=\!1.6$. \ [fig:P D\]]( plot_allPDver t.pdf "fig : " ) -0.2cm
Duet ot he linear coup lingof the mag n etic fie ld to the FHmoments,w e expectmag net icfie l d t o initially f a v or F Horder,lea ding to in cre ase d h yb ridizatio n and hi gh er $ T_ K$. This increase i s s ee n i n all threesubpl otsof F i g.\[fig:P D \] , most n ot icea bly i n par t (b ) . H owever, as stron ger fiel ds s plit th e $\Gamma_3$do ublet, the K ond o scre e n ing andthus FH order are event u ally de str oyed, asseen in t hemean-f iel d calcu lation . AFQ o rde r is al s o s upp re ssed as af u nct ion o f$B$, leadin g to a first-order tra nsition betwe enit a n d F H o r de r at a cri t i cal field $B_c$ . Note tha tw hi le the mea n -fi el d theor y alway s fin d s a fir st ordertransitio nbetw e e n F H and PQ s tates at low temp e ratur e sand h igh field s, we do n ot exp e ctthisto nec es sarily hold b eyond th e mean-field level.
La ndau t heory \[ sec:landa u\] | with_FH order_at low fields \[Fig. \[fig:PD\](a)\]._By increasing_$J_Q$_and moving_away_from half-filling, the_coexistence region can_be made to disappear,_replaced by a_direct_transition between the FH and AFQ phases \[Fig. \[fig:PD\](b,c)\], which is reminiscent of the_experimental_result, with_FH_order_explaining the heavy Fermi liquid_region in intermediate fields.
![Three example_mean-field phase_diagrams for FH (red) and AFQ (blue) orders_in_magnetic field, for_three different strengths of quadrupolar exchange coupling: (a) $J_Q\!=\!0.38t$,_(b) $J_Q\!=\!1.6t$, (c) $J_Q\!=\!2t$. The transition_to FH order,_$T_K$_initially_rises with magnetic field_and is then suppressed, ultimately becoming_a first-order transition in field. Other_parameters for (a): $t\!=\!1$, $V\!=\!0.8$, $\Delta E\!=\!5.5$,_$\Delta\!=\!4.8$, $n_{c,0}\!=\!1.9$; other parameters for (b),(c):_$t\!=\!1$, $V\!=\!0.9$, $\Delta E\!=\!5.5$, $\Delta\!=\!20$,_$n_{c,0}\!=\!1.6$. \[fig:PD\]](plot_allPDvert.pdf_"fig:")-0.2cm
Due to the linear coupling_of the magnetic_field to_the FH moments,_we expect magnetic field to initially_favor FH order,_leading to increased hybridization and higher_$T_K$._This increase is_seen_in_all three_subplots of Fig._\[fig:PD\],_most noticeably_in_part (b). However, as stronger fields_split_the $\Gamma_3$ doublet, the Kondo screening and_thus FH order are_eventually_destroyed, as seen in_the mean-field calculation. AFQ order_is also suppressed as a function_of $B$,_leading to_a first-order transition between it and FH order at a critical_field $B_c$. Note that while the_mean-field theory always finds_a first_order_transition between FH_and_PQ states_at low temperatures and high fields, we_do not_expect this to necessarily hold beyond_the mean-field level.
Landau theory_\[sec:landau\]
|
T:main1\]): the diagram $$\label{E:diagcomp}
\begin{array}{c}
\xymatrix{
{\mathcal{A}}\big|_x \ar[rr]^-{Y_1} & & \operatorname{\mathfrak{sl}}\bigl(\operatorname{\mathsf{Hom}}({\mathcal{P}}, {\mathbbm{k}}_x)\bigr) \ar[dd]^{\overline{{\mathsf{m}}}_{x, y}} \\
H^0\bigl({\mathcal{A}}(x)\bigr) \ar[u]^{\operatorname{\mathsf{res}}_x} \ar[d]_{\operatorname{\mathsf{ev}}_y} & & \\
{\mathcal{A}}\big|_y \ar[rr]^-{Y_2} & & \mathfrak{pgl}\bigl(\operatorname{\mathsf{Hom}}({\mathcal{P}}, {\mathbbm{k}}_y)\bigr)
}
\end{array}$$ is commutative, where $Y_1$ and $Y_2$ are certain canonical anti-isomorphisms of Lie algebras. A version of this important fact has been stated in [@Polishchuk1 Theorem 4(b)].
Using the commutative diagram (\[E:diagcomp\]), we prove Theorem A. As a consequence, we obtain the continuity property of the solution $r_{(E, (n, d))}$ with respect to the Weierstraß parameters $g_2$ and $g_3$ of the curve $E$. This actually leads to certain unexpected analytic consequences about classical $r$–matrices, see Corollary \[C:AnalyticConseq\].
The above construction can be rephrased in the following way. Let $E$ be an arbitrary Weierstraß cubic curve. Then there exists a canonical meromorphic section $$r \in \Gamma\bigl(\breve{E} \times \breve{E}, p_1^{*}{\mathcal{A}}\otimes p_2^{*}{\mathcal{A}}\bigr),$$ where $p_1, p_2: \breve{E} \times \breve{E} \rightarrow E$ are canonical projections, satisfying the equation $$\bigl[r^{12}, r^{13}\bigr] + \bigl[r^{13}, r^{23}\bigr] + \bigl[r^{12}, r^{23}\bigr] = 0,$$ see Theorem \[T:main2\]. It seems that in the case of elliptic curves | T: main1\ ] ): the diagram $ $ \label{E: diagcomp }
\begin{array}{c }
\xymatrix {
{ \mathcal{A}}\big|_x \ar[rr]^-{Y_1 } & & \operatorname{\mathfrak{sl}}\bigl(\operatorname{\mathsf{Hom}}({\mathcal{P } }, { \mathbbm{k}}_x)\bigr) \ar[dd]^{\overline{{\mathsf{m}}}_{x, y } } \\
H^0\bigl({\mathcal{A}}(x)\bigr) \ar[u]^{\operatorname{\mathsf{res}}_x } \ar[d]_{\operatorname{\mathsf{ev}}_y } & & \\
{ \mathcal{A}}\big|_y \ar[rr]^-{Y_2 } & & \mathfrak{pgl}\bigl(\operatorname{\mathsf{Hom}}({\mathcal{P } }, { \mathbbm{k}}_y)\bigr)
}
\end{array}$$ is commutative, where $ Y_1 $ and $ Y_2 $ are certain canonical anti - isomorphisms of Lie algebra. A interpretation of this important fact has been stated in [ @Polishchuk1 Theorem 4(b) ].
use the commutative diagram (\[E: diagcomp\ ]), we test Theorem A. As a consequence, we obtain the continuity property of the solution $ r_{(E, (n, d))}$ with regard to the Weierstraß parameter $ g_2 $ and $ g_3 $ of the curve $ E$. This actually contribute to certain unexpected analytic consequences about authoritative $ r$–matrices, see Corollary \[C: AnalyticConseq\ ].
The above construction can be rephrased in the following direction. Let $ E$ be an arbitrary Weierstraß cubic curve. Then there exists a canonical meromorphic section $ $ roentgen \in \Gamma\bigl(\breve{E } \times \breve{E }, p_1^{*}{\mathcal{A}}\otimes p_2^{*}{\mathcal{A}}\bigr),$$ where $ p_1, p_2: \breve{E } \times \breve{E } \rightarrow E$ are canonical projections, satisfying the equality $ $ \bigl[r^{12 }, r^{13}\bigr ] + \bigl[r^{13 }, r^{23}\bigr ] + \bigl[r^{12 }, r^{23}\bigr ] = 0,$$ see Theorem \[T: main2\ ]. It seems that in the font of elliptic curve | T:maln1\]): the diagram $$\label{E:diancomp}
\begin{array}{c}
\rtmatrie{
{\mathcam{A}}\big|_x \af[rr]^-{Y_1} & & \operatorname{\mathfrak{dl}}\vigl(\okvratorname{\mathsf{Hom}}({\maghcal{P}}, {\manhbbm{k}}_x)\bitr) \ac[dd]^{\overline{{\mathsh{j}}}_{x, y}} \\
H^0\blyl({\matggal{A}}(x)\yijr) \ar[u]^{\operatornsme{\mathsf{ras}}_x} \ar[d]_{\operatosnxmz{\mathsf{ev}}_y} & & \\
{\mathcal{A}}\big|_y \ar[rr]^-{Y_2} & & \mwthfrak{lgp}\bigl(\operatornwme{\msehsf{Gom}}({\mathcal{P}}, {\mathbbm{k}}_y)\bigr)
}
\end{array}$$ js commltative, where $Y_1$ amd $Y_2$ are certain canonical antl-isomorphisms of Lle algebras. A vqesion of thir importanu yact has begn stated in [@Polishchuk1 Theorem 4(c)].
Usiny the commujcrivf diagram (\[E:dmagcomi\]), we prove Thcprem A. As a cpnsequence, we pbtein rhe continuity propervy of the solution $r_{(G, (n, d))}$ with rzspect to the Weiersteaß paraketess $g_2$ qnd $g_3$ kf tge curge $X$. This actuzlly leads ro certain unexpecttd wbalytic consesuencef wbout classical $r$–matrices, see Corollary \[C:AhalyticConseq\].
The above xonstruction can be rgphrased ig the following way. Let $E$ be an arbitrary Weierstsaß cuuiz cbvye. Tfwn there exists a canonical meromorphic section $$d \on \Gamma\bigl(\brevc{E} \times \breve{E}, p_1^{*}{\mstjcsj{A}}\otimes p_2^{*}{\matfcal{A}}\bnfr),$$ where $p_1, p_2: \breve{E} \himes \bteve{E} \eightarror E$ sre canonical projections, sqtisfying thv eqyation $$\bigl[r^{12}, r^{13}\bigr] + \bigl[r^{13}, r^{23}\biyr] + \bibl[r^{12}, r^{23}\nigr] = 0,$$ see Theorem \[T:maiu2\]. It ssems that ij the cass of elliptic curxes | T:main1\]): the diagram $$\label{E:diagcomp} \begin{array}{c} \xymatrix{ {\mathcal{A}}\big|_x & {\mathbbm{k}}_x)\bigr) \ar[dd]^{\overline{{\mathsf{m}}}_{x, \\ H^0\bigl({\mathcal{A}}(x)\bigr) \ar[u]^{\operatorname{\mathsf{res}}_x} \ar[rr]^-{Y_2} & \mathfrak{pgl}\bigl(\operatorname{\mathsf{Hom}}({\mathcal{P}}, {\mathbbm{k}}_y)\bigr) \end{array}$$ is commutative, $Y_1$ and $Y_2$ are certain canonical of Lie algebras. A version of this important fact has been stated in Theorem 4(b)]. Using the commutative diagram (\[E:diagcomp\]), we prove Theorem A. As a we the property the solution $r_{(E, (n, d))}$ with respect to the Weierstraß parameters $g_2$ and $g_3$ of the $E$. This actually leads to certain unexpected analytic about classical $r$–matrices, see \[C:AnalyticConseq\]. The above construction can rephrased the following Let be arbitrary Weierstraß cubic Then there exists a canonical meromorphic section $$r \in \Gamma\bigl(\breve{E} \times \breve{E}, p_1^{*}{\mathcal{A}}\otimes p_2^{*}{\mathcal{A}}\bigr),$$ where $p_1, p_2: \times \breve{E} are canonical satisfying equation r^{13}\bigr] + \bigl[r^{13}, \bigl[r^{12}, r^{23}\bigr] = 0,$$ see Theorem that in the case of elliptic curves | T:main1\]): the diagram $$\label{E:diagComp}
\begin{aRray}{c}
\XymAtrIx{
{\MathCal{A}}\Big|_x \ar[rr]^-{Y_1} & & \operaTOrnaMe{\mathfrak{sl}}\bigl(\operatOrnamE{\mAThsf{hOm}}({\MathcAl{P}}, {\mathBBm{K}}_X)\BigR) \aR[dD]^{\ovErLInE{{\mathSf{m}}}_{X, y}} \\
H^0\bigl({\Mathcal{A}}(x)\bIgr) \Ar[U]^{\operatornamE{\MaThsf{res}}_x} \ar[D]_{\opEratorname{\maThsF{ev}}_y} & & \\
{\maThCal{a}}\Big|_y \aR[rr]^-{y_2} & & \mathFrak{pgL}\Bigl(\opEratornamE{\mAThsf{HoM}}({\Mathcal{p}}, {\MAtHbbm{K}}_y)\bigr)
}
\end{array}$$ is cOMmUTative, where $Y_1$ anD $Y_2$ are cErTAiN CAnoNicAl anti-isomOrPhismS Of Lie alGEbRAS. a veRSion of this impOrtant fact hAS beEn statEd In [@POLishchUk1 TheOrEM 4(b)].
USing the commUtatIve diagraM (\[E:diagCOmp\]), we prOVe TheorEm A. As a ConSeqUencE, We ObTaiN tHE coNTiNuiTY prOperty of ThE sOlutiOn $r_{(E, (N, D))}$ WITh reSpeCt to The WeIerstraß paramEteRs $g_2$ aND $g_3$ oF the cUrve $E$. this AcTuallY leads To cerTaIn unexpected anaLytiC consequeNceS aBouT cLassiCAl $r$–matRicEs, sEe CorolLary \[C:AnALytIccONSeQ\].
The above constructIoN CAn Be rephraSed in tHE fOlLOwing way. leT $E$ bE an aRBItrarY WeiERsTraß cubiC curve. tHeN tHere exiStS a canoNiCal MerOmorpHIc seCtion $$r \In \Gamma\bIgl(\brEVe{E} \times \breve{E}, P_1^{*}{\Mathcal{A}}\otimeS P_2^{*}{\mATHcAL{A}}\biGr),$$ wHere $p_1, p_2: \breve{e} \timES \breVe{E} \rIGhTarROw E$ arE canoNiCAl PRojections, satisfyinG tHe equaTion $$\bIgl[r^{12}, r^{13}\bigr] + \bigl[R^{13}, r^{23}\bigr] + \bigl[R^{12}, R^{23}\BIgr] = 0,$$ see ThEoreM \[t:mAIn2\]. It seems that iN the cAse of ellipTIc curves | T:main1\]): the diagram $$ \label{E:d iagco mp}
\b eg in{a rray }{c}
\xymatrix {
{\m athcal{A}}\big|_x \ar[ rr]^- {Y _ 1} & & \ope ratorna m e{ \ m ath fr ak {sl }} \ bi gl(\o per atornam e{\mathsf{ Hom }} ({\mathcal{P } }, {\mathbbm {k} }_x)\bigr) \ ar[ dd]^{\ ov erl i ne{{\ mat hsf{m }}}_{x , y}} \ \
H^0\big l( { \mathc a l{A}}(x ) \ bi gr)\ar[u]^{\operator n am e {\mathsf{res}} _x} \a r[ d ]_ { \ ope rat orname{\ma th sf{ev } }_y} && \ \ {\m a thcal{A}}\big |_y \ar[rr] ^ -{Y _2} &&\ma t hfrak{ pgl}\ bi g l(\ operatornam e{\m athsf{Hom }}({\m a thcal{P } }, {\ma thbbm{ k}} _y) \big r )}\en d{ a rra y }$ $ i s co mmutativ e, w here$Y_1 $ a n d $Y _2$ are cert ain canonical an ti-i s omo rphis ms of Lie a lgebr as. Aversi on of this import antfact hasbee nsta te d in[ @Polis hch uk1 Theore m 4(b)] .
U si n g th e commutative diag ra m (\ [E:diagc omp\]) , w ep rove The or emA. A s a con sequ e nc e, we ob tain t h eco ntinuit yproper ty of th e sol u tion $r_{( E, (n, d ))}$w ith respect to the Weierstra ß p a r am e ters $g _2$ and $g_ 3$ o f the cur v e$E$ . This actu al l yl eads to certain une xp ectedanaly tic consequen ces aboutc l a ssical $ r$–m a tr i ces, see Corol lary\[C:Analyt i cConseq\ ].
T he above construc t i on can b e r eph ras edi n t he followingw a y. L et $E$ be an arbitr ary We ier str aß cubic cu rve. The nth er eexi sts a canonica lmer om orp hic s e ction$$r \ in \ Ga mm a \bi gl(\bre v e{ E } \ti me s\bre ve{ E} , p_1 ^{*} { \ma thcal{A }}\otimes p_ 2 ^{*} {\ ma thcal{A }}\bigr),$$ w he re $p_1, p _2 : \ breve{ E } \times\breve{E} \rightarrow E $ are ca non icalproj ections,sat isfyin g t h e equa tion $ $\big l[ r^{ 1 2 }, r^ { 1 3} \bi gr ] + \bigl[ r ^ {13 }, r^ {2 3}\b igr] +\bigl[r^{12}, r^{2 3 }\b igr] = 0,$$ s eeTheo r e m\[T : ma i n2\ ]. Its e ems that in the case of e ll i pt ic curves | T:main1\]): the_diagram $$\label{E:diagcomp}
\begin{array}{c}
\xymatrix{
{\mathcal{A}}\big|_x_\ar[rr]^-{Y_1} & &_\operatorname{\mathfrak{sl}}\bigl(\operatorname{\mathsf{Hom}}({\mathcal{P}}, {\mathbbm{k}}_x)\bigr)_\ar[dd]^{\overline{{\mathsf{m}}}_{x,_y}} \\
H^0\bigl({\mathcal{A}}(x)\bigr)_\ar[u]^{\operatorname{\mathsf{res}}_x}_\ar[d]_{\operatorname{\mathsf{ev}}_y} & &_ \\
{\mathcal{A}}\big|_y \ar[rr]^-{Y_2}_& & \mathfrak{pgl}\bigl(\operatorname{\mathsf{Hom}}({\mathcal{P}}, {\mathbbm{k}}_y)\bigr)
}
\end{array}$$_is commutative, where_$Y_1$_and $Y_2$ are certain canonical anti-isomorphisms of Lie algebras. A version of this important_fact_has been_stated_in_[@Polishchuk1 Theorem 4(b)].
Using the commutative_diagram (\[E:diagcomp\]), we prove Theorem_A. As_a consequence, we obtain the continuity property of_the_solution $r_{(E, (n,_d))}$ with respect to the Weierstraß parameters $g_2$ and_$g_3$ of the curve $E$. This_actually leads to_certain_unexpected_analytic consequences about classical_$r$–matrices, see Corollary \[C:AnalyticConseq\].
The above construction_can be rephrased in the following_way. Let $E$ be an arbitrary Weierstraß_cubic curve. Then there exists a_canonical meromorphic section $$r \in_\Gamma\bigl(\breve{E} \times_\breve{E}, p_1^{*}{\mathcal{A}}\otimes p_2^{*}{\mathcal{A}}\bigr),$$ where $p_1,_p_2: \breve{E} \times_\breve{E} \rightarrow_E$ are canonical_projections, satisfying the equation $$\bigl[r^{12}, r^{13}\bigr]_+ \bigl[r^{13}, r^{23}\bigr]_+ \bigl[r^{12}, r^{23}\bigr] = 0,$$ see_Theorem_\[T:main2\]. It seems_that_in_the case_of elliptic curves |
( X_{\pm },Y_{\pm }^{\bot }\right) \right) _{\mathfrak{h}}\end{aligned}$$that in $2\times 2$ block matrix form means $$\delta \left( X_{\pm },Y_{\pm }^{\bot }\right) =\left(
\begin{array}{cc}
\delta \left( Y_{\pm }^{\bot }\right) & \tau ^{\ast }\left( X_{\pm }\right)
\\
-\tau ^{\ast }\left( X_{\pm }\right) & 0%
\end{array}%
\right)$$where $\delta :\mathfrak{g}_{\pm }^{\bot }\longrightarrow \mathfrak{g}_{\pm
}^{\bot }\otimes \mathfrak{g}_{\pm }^{\bot }$ such that$$\left( X_{\mp }^{\prime }\otimes X_{\mp }^{\prime \prime },\delta \left(
Y_{\pm }^{\bot }\right) \right) _{\mathfrak{g}}:=\left( \left[ X_{\mp
}^{\prime },X_{\mp }^{\prime \prime }\right],Y_{\pm }^{\bot }\right) _{%
\mathfrak{g}}$$and $\tau ^{\ast }:\mathfrak{g}_{\pm }\longrightarrow \mathfrak{g}_{\pm
}^{\bot }\otimes \mathfrak{g}_{\pm }$ is$$\left( X_{\mp }^{\prime }\otimes Y_{\mp }^{\prime \prime \bot },\tau ^{\ast
}\left( X_{\pm }\right) \right) _{\mathfrak{g}}:=\left( \mathrm{ad}_{X_{\mp
}^{\prime }}^{\tau }Y_{\mp }^{\prime \prime \bot },X_{\pm }\right) _{%
\mathfrak{g}}$$
### **The Poisson-Lie bivector on** $\mathit{H}_{\mathit{+}}$
The Poisson-Lie bivector $\pi _{+}\in \Gamma \left( T^{\otimes
2}H_{+}\right) $, the sections of the vector bundle $T^{\otimes
2}H_{+}\longrightarrow H_{+}$, is defined by the relation $$\begin{aligned}
&&\left\langle \gamma \left( X_{-}^{\prime },Y_{-}^{\prime \bot }\right)
\left( g_{+},Y_{+}^{\bot }\right) ^{-1}\otimes \gamma \left( X_{-}^{\prime
\prime },Y_{- | ( X_{\pm }, Y_{\pm } ^{\bot } \right) \right) _ { \mathfrak{h}}\end{aligned}$$that in $ 2\times 2 $ block matrix form means $ $ \delta \left (X_{\pm }, Y_{\pm } ^{\bot } \right) = \left (
\begin{array}{cc }
\delta \left (Y_{\pm } ^{\bot } \right) & \tau ^{\ast } \left (X_{\pm } \right)
\\
-\tau ^{\ast } \left (X_{\pm } \right) & 0%
\end{array}%
\right)$$where $ \delta: \mathfrak{g}_{\pm } ^{\bot } \longrightarrow \mathfrak{g}_{\pm
} ^{\bot } \otimes \mathfrak{g}_{\pm } ^{\bot } $ such that$$\left (X_{\mp } ^{\prime } \otimes X_{\mp } ^{\prime \prime }, \delta \left (
Y_{\pm } ^{\bot } \right) \right) _ { \mathfrak{g}}:=\left (\left [ X_{\mp
} ^{\prime }, X_{\mp } ^{\prime \prime } \right],Y_{\pm } ^{\bot } \right) _ {%
\mathfrak{g}}$$and $ \tau ^{\ast }: \mathfrak{g}_{\pm } \longrightarrow \mathfrak{g}_{\pm
} ^{\bot } \otimes \mathfrak{g}_{\pm } $ is$$\left (X_{\mp } ^{\prime } \otimes Y_{\mp } ^{\prime \prime \bot }, \tau ^{\ast
} \left (X_{\pm } \right) \right) _ { \mathfrak{g}}:=\left (\mathrm{ad}_{X_{\mp
} ^{\prime } } ^{\tau } Y_{\mp } ^{\prime \prime \bot }, X_{\pm } \right) _ {%
\mathfrak{g}}$$
# # # * * The Poisson - Lie bivector on * * $ \mathit{H}_{\mathit{+}}$
The Poisson - Lie bivector $ \pi _ { + } \in \Gamma \left (T^{\otimes
2}H_{+}\right) $, the sections of the vector package $ T^{\otimes
2}H_{+}\longrightarrow H_{+}$, is define by the relation $ $ \begin{aligned }
& & \left\langle \gamma \left (X_{-}^{\prime }, Y_{-}^{\prime \bot } \right)
\left (g_{+},Y_{+}^{\bot } \right) ^{-1}\otimes \gamma \left (X_{-}^{\prime
\prime }, Y_{- | ( X_{\pl },Y_{\pm }^{\bot }\right) \right) _{\mauhfrak{h}}\end{aligned}$$jhqt in $2\vimes 2$ glock magrix form means $$\delta \left( X_{\'m },Y_{\pm }^{\bou }\right) =\left(
\begin{afray}{cc}
\delna \left( Y_{\pm }^{\bir }\right) & \teh ^{\ast }\lcyt( X_{\pj }\rigkt)
\\
-\tau ^{\ast }\left( W_{\pm }\right) & 0%
\and{array}%
\right)$$wvefe $\delta :\mathfrak{g}_{\pm }^{\bot }\longrightarrjw \mathgrwk{g}_{\pm
}^{\bot }\otimef \mauhfwak{g}_{\lm }^{\bot }$ such that$$\left( X_{\mp }^{\prime }\otjmes X_{\mk }^{\prime \prime },\delts \left(
Y_{\pm }^{\bot }\right) \right) _{\lathvrak{g}}:=\left( \left[ X_{\mp
}^{\orime },X_{\mp }^{\ptjme \prime }\right],Y_{\om }^{\bot }\right) _{%
\mathfrak{g}}$$znd $\tau ^{\ast }:\mathfrak{g}_{\pm }\longrigftarrpw \mathfraj{g}_{\pm
}^{\blj }\otimes \matifrak{g}_{\im }$ is$$\left( X_{\mi }^{\prime }\otimes Y_{\mp }^{\prime \prike \uot },\rau ^{\ast
}\left( X_{\pm }\right) \right) _{\mathfrak{g}}:=\lest( \mathrm{dd}_{R_{\mp
}^{\prime }}^{\tau }Y_{\mp }^{\primw \prime \bot },X_{\pm }\eigft) _{%
\javhfdak{g}}$$
### **Tje 'oisson-Lie givector on** $\mathit{H}_{\mathit{+}}$
The Ppifwon-Lie bivectkr $\pi _{+}\yn \Gamma \left( T^{\otimes
2}H_{+}\right) $, the sections of the vector bundle $T^{\otines
2}H_{+}\longrightarrow H_{+}$, ls defineq by the relation $$\begin{aligned}
&&\left\langle \gamma \lext( X_{-}^{\pcioe },V_{-}^{\ivime \voh }\right)
\left( g_{+},Y_{+}^{\bot }\right) ^{-1}\otimes \gamma \left( X_{-}^{\pryje
\krike },Y_{- | ( X_{\pm },Y_{\pm }^{\bot }\right) \right) _{\mathfrak{h}}\end{aligned}$$that 2$ matrix form $$\delta \left( X_{\pm \delta Y_{\pm }^{\bot }\right) \tau ^{\ast }\left( }\right) \\ -\tau ^{\ast }\left( X_{\pm & 0% \end{array}% \right)$$where $\delta :\mathfrak{g}_{\pm }^{\bot }\longrightarrow \mathfrak{g}_{\pm }^{\bot }\otimes \mathfrak{g}_{\pm }^{\bot such that$$\left( X_{\mp }^{\prime }\otimes X_{\mp }^{\prime \prime },\delta \left( Y_{\pm }^{\bot }\right) _{\mathfrak{g}}:=\left( X_{\mp },X_{\mp \prime }\right],Y_{\pm }^{\bot }\right) _{% \mathfrak{g}}$$and $\tau ^{\ast }:\mathfrak{g}_{\pm }\longrightarrow \mathfrak{g}_{\pm }^{\bot }\otimes \mathfrak{g}_{\pm }$ is$$\left( }^{\prime }\otimes Y_{\mp }^{\prime \prime \bot },\tau ^{\ast X_{\pm }\right) \right) _{\mathfrak{g}}:=\left( }^{\prime }}^{\tau }Y_{\mp }^{\prime \prime },X_{\pm _{% \mathfrak{g}}$$ **The bivector $\mathit{H}_{\mathit{+}}$ The Poisson-Lie $\pi _{+}\in \Gamma \left( T^{\otimes 2}H_{+}\right) $, the sections of the vector bundle $T^{\otimes 2}H_{+}\longrightarrow H_{+}$, is by the &&\left\langle \gamma X_{-}^{\prime \bot \left( g_{+},Y_{+}^{\bot }\right) \left( X_{-}^{\prime \prime },Y_{- | ( X_{\pm },Y_{\pm }^{\bot }\right) \right) _{\mathfrAk{h}}\end{aligNed}$$thAt iN $2\tiMeS 2$ bloCk maTrix form means $$\dELta \lEft( X_{\pm },Y_{\pm }^{\bot }\right) =\left(
\bEgin{aRrAY}{cc}
\dELtA \left( y_{\pm }^{\bot }\rIGhT) & \TAu ^{\aSt }\LeFt( X_{\Pm }\RIgHt)
\\
-\tau ^{\Ast }\Left( X_{\pm }\Right) & 0%
\end{arRay}%
\RiGht)$$where $\deltA :\MaThfrak{g}_{\pm }^{\bOt }\lOngrightarroW \maThfrak{G}_{\pM
}^{\boT }\OtimeS \maThfraK{g}_{\pm }^{\boT }$ Such thAt$$\left( X_{\mp }^{\PrIMe }\otimES X_{\mp }^{\priME \PrIme },\dElta \left(
Y_{\pm }^{\bot }\rigHT) \rIGht) _{\mathfrak{g}}:=\leFt( \left[ x_{\mP
}^{\PrIME },X_{\mP }^{\prIme \prime }\riGhT],Y_{\pm }^{\bOT }\right) _{%
\mAThFRAK{g}}$$aND $\tau ^{\ast }:\mathfrAk{g}_{\pm }\longriGHtaRrow \maThFraK{G}_{\pm
}^{\bot }\OtimeS \mAThfRak{g}_{\pm }$ is$$\lefT( X_{\mp }^{\Prime }\otimEs Y_{\mp }^{\pRIme \primE \Bot },\tau ^{\aSt
}\left( x_{\pm }\RigHt) \riGHt) _{\MaThfRaK{G}}:=\leFT( \mAthRM{ad}_{x_{\mp
}^{\prime }}^{\TaU }Y_{\Mp }^{\priMe \prIME \BOt },X_{\pM }\riGht) _{%
\mAthfrAk{g}}$$
### **The Poisson-lie BiveCTor On** $\matHit{H}_{\mAthiT{+}}$
THe PoiSson-LiE biveCtOr $\pi _{+}\in \Gamma \left( t^{\otiMes
2}H_{+}\right) $, The SeCtiOnS of thE Vector BunDle $t^{\otimes
2}h_{+}\longriGHtaRrOW h_{+}$, Is Defined by the relatiOn $$\BEGiN{aligned}
&&\Left\laNGlE \gAMma \left( X_{-}^{\PrIme },y_{-}^{\priME \Bot }\riGht)
\lEFt( G_{+},Y_{+}^{\bot }\rigHt) ^{-1}\otimES \gAmMa \left( X_{-}^{\PrIme
\priMe },y_{- | ( X_{\pm },Y_{\pm }^{\bot}\right) \ right ) _ {\m at hfra k{h} }\end{aligned} $ $tha t in $2\times 2$ block matr ix form me ans $ $\delta \l e f t(X_ {\ pm}, Y _{ \pm } ^{\ bot }\r ight) =\le ft( \begin{array } {c c}
\delta\le ft( Y_{\pm } ^{\ bot }\ ri ght ) & \t au^{\as t }\le f t( X_{ \pm }\rig ht )
\\
- \ tau ^{\ a s t}\le ft( X_{\pm }\righ t )& 0%
\end{array }%
\ri gh t )$ $ w her e $ \delta :\m at hfrak { g}_{\pm }^ { \ b ot} \longrightarr ow \mathfra k {g} _{\pm}^ {\b o t }\ot imes\m a thf rak{g}_{\pm }^{ \bot }$ s uch th a t$$\lef t ( X_{\m p }^{\ pri me}\ot i me sX_{ \m p }^ { \p rim e \p rime },\ de lt a \le ft(Y _ { \ pm } ^{\ bot}\rig ht) \right) _ {\m athf r ak{ g}}:= \left ( \l ef t[ X_ {\mp
} ^{\pr im e },X_{\mp }^{\ prim e \prime}\r ig ht] ,Y _{\pm }^{\bo t } \ri ght) _{ %
\math f rak {g } } $ $a nd $\tau ^{\ast }: \m a t hf rak{g}_{ \pm }\ l on gr i ghtarrow \ mat hfra k { g}_{\ pm
} ^ {\ bot }\ot imes \ m at hf rak{g}_ {\ pm }$is $$\ lef t( X_ { \mp}^{\pr ime }\ot imesY _{\mp }^{\prim e \prime \bot} ,\ t a u^ {\as t
} \left( X_{\ pm } \ righ t) \ r ig ht) _{\ma thfra k{ g }} : =\left( \mathrm{ad} _{ X_{\mp
}^{\ prime }}^{\ta u }Y_{\mp} ^ { \prime \ prim e \ b ot },X_{\pm }\ right ) _{%
\mat h frak{g}} $$
# ## **The Poisson- L i e bivect oron* * $ \ma t h it {H}_{\mathit{ + } }$
Th e Poiss on- Lie biv ect or$\p i _ {+ }\in \Gam ma \left (T^ {\ ot ime s
2}H _ {+}\righ t) $, t hesecti o ns ofthe v ecto rbu n dle $T^{\o t im e s
2}H _{ +} \lon gri gh tarro w H_ { +}$ , is de fined bythe rela ti on $$\beg in{aligned}
& &\ left\langl e\ga mma \l e f t( X_{-} ^{\prime },Y_{-}^{\prim e \bot } \ri ght)\lef t( g_{+}, Y_{ +}^{\b ot} \right ) ^{-1 }\oti me s \ g a mma \ l e ft ( X _{ -}^{\prime \ pri me }, Y_ {- | ( X_{\pm_},Y_{\pm }^{\bot_}\right) \right) _{\mathfrak{h}}\end{aligned}$$that in_$2\times 2$_block_matrix form_means_$$\delta \left( X_{\pm_},Y_{\pm }^{\bot }\right)_=\left(
\begin{array}{cc}
\delta \left( Y_{\pm_}^{\bot }\right) &_\tau_^{\ast }\left( X_{\pm }\right)
\\
-\tau ^{\ast }\left( X_{\pm }\right) & 0%
\end{array}%
\right)$$where $\delta :\mathfrak{g}_{\pm }^{\bot_}\longrightarrow_\mathfrak{g}_{\pm
}^{\bot }\otimes_\mathfrak{g}_{\pm_}^{\bot_}$ such that$$\left( X_{\mp }^{\prime_}\otimes X_{\mp }^{\prime \prime },\delta_\left(
Y_{\pm }^{\bot_}\right) \right) _{\mathfrak{g}}:=\left( \left[ X_{\mp
}^{\prime },X_{\mp }^{\prime \prime_}\right],Y_{\pm_}^{\bot }\right) _{%
\mathfrak{g}}$$and_$\tau ^{\ast }:\mathfrak{g}_{\pm }\longrightarrow \mathfrak{g}_{\pm
}^{\bot }\otimes \mathfrak{g}_{\pm }$ is$$\left(_X_{\mp }^{\prime }\otimes Y_{\mp }^{\prime \prime_\bot },\tau ^{\ast
}\left(_X_{\pm_}\right)_\right) _{\mathfrak{g}}:=\left( \mathrm{ad}_{X_{\mp
}^{\prime }}^{\tau_}Y_{\mp }^{\prime \prime \bot },X_{\pm }\right)__{%
\mathfrak{g}}$$
### **The Poisson-Lie bivector on** $\mathit{H}_{\mathit{+}}$
The_Poisson-Lie bivector $\pi _{+}\in \Gamma \left( T^{\otimes
2}H_{+}\right)_$, the sections of the vector_bundle $T^{\otimes
2}H_{+}\longrightarrow H_{+}$, is defined_by the_relation $$\begin{aligned}
&&\left\langle \gamma \left( X_{-}^{\prime_},Y_{-}^{\prime \bot }\right)
\left(_g_{+},Y_{+}^{\bot }\right)_^{-1}\otimes \gamma \left(_X_{-}^{\prime
\prime },Y_{- |
2}\alpha-1$ we get $$\begin{aligned}
\|Z_{k+1}\|_{s,r_{k+1}}\leq C(t_k^{s+2-2\alpha}+t_k^{-\frac{3}{2}\alpha-1})\leq Ct_k^{-\frac{3}{2}\alpha-1}\leq C\delta^{1/\alpha}t_{k}^{-\frac{3}{2}\alpha}= C\delta^{1/\alpha}t_{k+1}^{-\alpha},\end{aligned}$$ and by taking $\delta$ even smaller, this finishes the induction.
From (\[EQ\_Z\_nfixed\]) we conclude that for each $n\geq 1$, there is $k_n\geq 0$, such that $$\|Z_{k+1}\|_{n,r_{k+1}}\leq t_k^{s+2}\|Z_{k}\|_{n,r_{k}}, \ \ \forall\ k\geq k_n.$$ By iterating this, we obtain $$t_k^{s+2}\|Z_{k}\|_{n,r_{k}}\leq (t_kt_{k-1}\ldots t_{k_n})^{s+2}\|Z_{k_n}\|_{n,r_{k_n}}.$$ On the other hand we have that $$t_kt_{k-1}\ldots t_{k_n}=t_{k_n}^{1+\frac{3}{2}+\ldots+(\frac{3}{2})^{k-k_n}}\leq t_{k_n}^{2(\frac{3}{2})^{k+1-k_n}}=t_k^3.$$ Therefore, we obtain a bound valid for all $k>k_n$ $$\|Z_{k}\|_{n,r_{k}}\leq t_{k}^{2(s+2)}\|Z_{k_n}\|_{n,r_{k_n}}.$$ Consider now $m>s$ and denote by $n:=4m-3s$. Applying the interpolation inequalities, for $k> k_{n}$, we obtain $$\begin{aligned}
\|Z_k\|_{m,r_k}&\leq& C_m \|Z_k\|_{s,r_k}^{\frac{n-m}{n-s}}\|Z_k\|_{n,r_k}^{\frac{m-s}{n-s | 2}\alpha-1 $ we get $ $ \begin{aligned }
\|Z_{k+1}\|_{s, r_{k+1}}\leq C(t_k^{s+2 - 2\alpha}+t_k^{-\frac{3}{2}\alpha-1})\leq Ct_k^{-\frac{3}{2}\alpha-1}\leq C\delta^{1/\alpha}t_{k}^{-\frac{3}{2}\alpha}= C\delta^{1/\alpha}t_{k+1}^{-\alpha},\end{aligned}$$ and by taking $ \delta$ even smaller, this finishes the evocation.
From (\[EQ\_Z\_nfixed\ ]) we reason that for each $ n\geq 1 $, there is $ k_n\geq 0 $, such that $ $ \|Z_{k+1}\|_{n, r_{k+1}}\leq t_k^{s+2}\|Z_{k}\|_{n, r_{k } }, \ \ \forall\ k\geq k_n.$$ By iterating this, we receive $ $ t_k^{s+2}\|Z_{k}\|_{n, r_{k}}\leq (t_kt_{k-1}\ldots t_{k_n})^{s+2}\|Z_{k_n}\|_{n, r_{k_n}}.$$ On the early hand we have that $ $ t_kt_{k-1}\ldots t_{k_n}=t_{k_n}^{1+\frac{3}{2}+\ldots+(\frac{3}{2})^{k - k_n}}\leq t_{k_n}^{2(\frac{3}{2})^{k+1 - k_n}}=t_k^3.$$ Therefore, we receive a bound valid for all $ k > k_n$ $ $ \|Z_{k}\|_{n, r_{k}}\leq t_{k}^{2(s+2)}\|Z_{k_n}\|_{n, r_{k_n}}.$$ see now $ m > s$ and announce by $ n:=4m-3s$. Applying the interpolation inequalities, for $ k > k_{n}$, we obtain $ $ \begin{aligned }
\|Z_k\|_{m, r_k}&\leq & C_m \|Z_k\|_{s, r_k}^{\frac{n - m}{n - s}}\|Z_k\|_{n, r_k}^{\frac{m - s}{n - mho | 2}\alpja-1$ we get $$\begin{aligned}
\|Z_{k+1}\|_{r,r_{k+1}}\leq C(t_k^{s+2-2\alphc}+r_k^{-\frac{3}{2}\elpha-1})\les Ct_k^{-\fraz{3}{2}\alpha-1}\leq C\delta^{1/\alpha}t_{k}^{-\frac{3}{2}\appya}= C\dtjta^{1/\alpha}t_{k+1}^{-\alpha},\end{xligned}$$ ajd by tajing $\eelta$ even smaller, this rlnishzs the induction.
Nrom (\[EQ\_Z\_nfifed\]) we concluda ghct for each $n\geq 1$, there is $k_n\geq 0$, sush that $$\|Z_{n+1}\|_{n,r_{k+1}}\leq t_k^{s+2}\|Z_{k}\|_{n,t_{k}}, \ \ \gjralm\ k\geq k_n.$$ By iterating this, we obtzin $$t_k^{s+2}\|E_{k}\|_{n,r_{k}}\leq (t_kt_{k-1}\ldots t_{k_n})^{s+2}\|Z_{k_n}\|_{n,r_{k_n}}.$$ On the other jand we have that $$t_kt_{k-1}\pdots t_{k_n}=t_{k_b}^{1+\fras{3}{2}+\odots+(\frac{3}{2})^{k-k_n}}\ueq t_{k_n}^{2(\frab{3}{2})^{n+1-k_n}}=t_k^3.$$ Thergfore, we obtain a bound valid fof all $k>k_n$ $$\|Z_{k}\|_{n,r_{k}}\oew t_{n}^{2(v+2)}\|Z_{k_n}\|_{n,r_{k_n}}.$$ Coisider now $m>s$ and denote bf $n:=4m-3s$. Alplying the inberponatuon inequalities, for $n> k_{n}$, we obtain $$\begyn{aligned}
\|S_k\|_{j,r_k}&\leq& C_m \|Z_k\|_{s,r_k}^{\feax{n-m}{n-s}}\|E_k\|_{n,r_k}^{\xrac{o-w}{n-s | 2}\alpha-1$ we get $$\begin{aligned} \|Z_{k+1}\|_{s,r_{k+1}}\leq C(t_k^{s+2-2\alpha}+t_k^{-\frac{3}{2}\alpha-1})\leq Ct_k^{-\frac{3}{2}\alpha-1}\leq and taking $\delta$ smaller, this finishes conclude for each $n\geq there is $k_n\geq such that $$\|Z_{k+1}\|_{n,r_{k+1}}\leq t_k^{s+2}\|Z_{k}\|_{n,r_{k}}, \ \ k\geq k_n.$$ By iterating this, we obtain $$t_k^{s+2}\|Z_{k}\|_{n,r_{k}}\leq (t_kt_{k-1}\ldots t_{k_n})^{s+2}\|Z_{k_n}\|_{n,r_{k_n}}.$$ On the other we have that $$t_kt_{k-1}\ldots t_{k_n}=t_{k_n}^{1+\frac{3}{2}+\ldots+(\frac{3}{2})^{k-k_n}}\leq t_{k_n}^{2(\frac{3}{2})^{k+1-k_n}}=t_k^3.$$ Therefore, we obtain a bound valid for $k>k_n$ t_{k}^{2(s+2)}\|Z_{k_n}\|_{n,r_{k_n}}.$$ now and denote by $n:=4m-3s$. Applying the interpolation inequalities, for $k> k_{n}$, we obtain $$\begin{aligned} \|Z_k\|_{m,r_k}&\leq& C_m | 2}\alpha-1$ we get $$\begin{aligned}
\|Z_{k+1}\|_{s,R_{k+1}}\leq C(t_k^{s+2-2\aLpha}+t_K^{-\frAc{3}{2}\aLpHa-1})\leQ Ct_k^{-\Frac{3}{2}\alpha-1}\leq C\dELta^{1/\aLpha}t_{k}^{-\frac{3}{2}\alpha}= C\delta^{1/\aLpha}t_{K+1}^{-\aLPha},\eND{aLigneD}$$ and by tAKiNG $\DelTa$ EvEn sMaLLeR, this FinIshes thE induction.
froM (\[Eq\_Z\_nfixed\]) we coNClUde that for EacH $n\geq 1$, there is $K_n\gEq 0$, such ThAt $$\|Z_{K+1}\|_{N,r_{k+1}}\leQ t_k^{S+2}\|Z_{k}\|_{n,r_{K}}, \ \ \foralL\ K\geq k_n.$$ by iteratiNg THis, we oBTain $$t_k^{s+2}\|z_{K}\|_{N,r_{K}}\leq (T_kt_{k-1}\ldots t_{k_n})^{s+2}\|Z_{k_n}\|_{n,R_{K_n}}.$$ oN the other hand wE have tHaT $$T_kT_{K-1}\LdoTs t_{K_n}=t_{k_n}^{1+\frac{3}{2}+\lDoTs+(\fraC{3}{2})^{K-k_n}}\leq t_{K_N}^{2(\fRAC{3}{2})^{K+1-k_n}}=T_K^3.$$ Therefore, we oBtain a bound VAliD for alL $k>K_n$ $$\|Z_{K}\|_{N,r_{k}}\leq T_{k}^{2(s+2)}\|Z_{k_N}\|_{n,R_{K_n}}.$$ COnsider now $m>S$ and Denote by $n:=4M-3s$. ApplYIng the iNTerpolaTion inEquAliTies, FOr $K> k_{N}$, we ObTAin $$\BEgIn{aLIgnEd}
\|Z_k\|_{m,r_k}&\lEq& c_m \|z_k\|_{s,r_k}^{\Frac{N-M}{N-S}}\|z_k\|_{n,r_K}^{\frAc{m-s}{N-s | 2}\alpha-1$ we get $$\begi n{aligned}
\|Z_ {k+ 1}\ |_ {s,r _{k+ 1}}\leq C(t_k^ { s+2- 2\alpha}+t_k^{-\frac{3 }{2}\ al p ha-1 } )\ leq C t_k^{-\ f ra c { 3}{ 2} \a lph a- 1 }\ leq C \de lta^{1/ \alpha}t_{ k}^ {- \frac{3}{2}\ a lp ha}= C\del ta^ {1/\alpha}t_ {k+ 1}^{-\ al pha } ,\end {al igned }$$ an d by ta king $\de lt a $ even smaller , th is f inishes the induc t io n .
From (\[EQ\ _Z\_nf ix e d\ ] ) we co nclude tha tfor e a ch $n\g e q1 $ , th e re is $k_n\ge q 0$, sucht hat $$\|Z _{ k+1 } \|_{n, r_{k+ 1} } \le q t_k^{s+2} \|Z_ {k}\|_{n, r_{k}} , \ \ \f o rall\ k \geq k _n. $$By i t er at ing t h is, we ob t ain $$t_k^{ s+ 2} \|Z_{ k}\| _ { n , r_{k }}\ leq(t_kt _{k-1}\ldotst_{ k_n} ) ^{s +2}\| Z_{k_ n}\| _{ n,r_{ k_n}}. $$ On t he other hand w e ha ve that $ $t_ kt _{k -1 }\ldo t s t_{k _n} =t_ {k_n}^{ 1+\frac { 3}{ 2} + \ l do ts+(\frac{3}{2})^{ k- k _ n} }\leq t_ {k_n}^ { 2( \f r ac{3}{2} )^ {k+ 1-k_ n } }=t_k ^3.$ $ T herefore , we o b ta in a boun dvalidfo r a ll$k>k_ n $ $$ \|Z_{k }\|_{n,r _{k}} \ leq t_{k}^{2(s + 2)}\|Z_{k_n}\ | _{ n , r_ { k_n} }.$ $ Considernow$ m>s$ and de not e by $ n:=4m -3 s $. Applying the interp ol ationinequ alities, for$k> k_{n}$ , w e obtain $$\ b eg i n{aligned}
\|Z _k\|_ {m,r_k}&\l e q& C_m \|Z_k \|_{s,r_ k}^{\frac { n -m}{n-s} }\| Z_k \|_ {n, r _ k} ^{\frac{m-s}{ n - s | 2}\alpha-1$ we_get $$\begin{aligned}
\|Z_{k+1}\|_{s,r_{k+1}}\leq_C(t_k^{s+2-2\alpha}+t_k^{-\frac{3}{2}\alpha-1})\leq Ct_k^{-\frac{3}{2}\alpha-1}\leq C\delta^{1/\alpha}t_{k}^{-\frac{3}{2}\alpha}= C\delta^{1/\alpha}t_{k+1}^{-\alpha},\end{aligned}$$_and by_taking_$\delta$ even_smaller,_this finishes the_induction.
From (\[EQ\_Z\_nfixed\]) we_conclude that for each_$n\geq 1$, there_is_$k_n\geq 0$, such that $$\|Z_{k+1}\|_{n,r_{k+1}}\leq t_k^{s+2}\|Z_{k}\|_{n,r_{k}}, \ \ \forall\ k\geq k_n.$$ By iterating this,_we_obtain $$t_k^{s+2}\|Z_{k}\|_{n,r_{k}}\leq_(t_kt_{k-1}\ldots_t_{k_n})^{s+2}\|Z_{k_n}\|_{n,r_{k_n}}.$$_On the other hand we_have that $$t_kt_{k-1}\ldots t_{k_n}=t_{k_n}^{1+\frac{3}{2}+\ldots+(\frac{3}{2})^{k-k_n}}\leq t_{k_n}^{2(\frac{3}{2})^{k+1-k_n}}=t_k^3.$$_Therefore, we_obtain a bound valid for all $k>k_n$ $$\|Z_{k}\|_{n,r_{k}}\leq_t_{k}^{2(s+2)}\|Z_{k_n}\|_{n,r_{k_n}}.$$_Consider now $m>s$_and denote by $n:=4m-3s$. Applying the interpolation inequalities, for_$k> k_{n}$, we obtain $$\begin{aligned}
\|Z_k\|_{m,r_k}&\leq& C_m_ \|Z_k\|_{s,r_k}^{\frac{n-m}{n-s}}\|Z_k\|_{n,r_k}^{\frac{m-s}{n-s |
end{aligned}\label{eq:deriv-radi-equat:13}$$ which is cast into the final form through the use of the identity $f^\prime(r) = ((rf(r))^\prime - f(r))/r$ $$\begin{aligned}
&(rg(r))^\prime - \frac{rg(r)}{r} - (m_0 - e_0\varphi(r))rf(r) = \Lambda f(r),
\\
&(rf(r))^\prime + \frac{rf(r)}{r} - (m_0 + e_0 \varphi(r) )rg(r) = -\Lambda g(r).
\end{aligned}\label{eq:deriv-radi-equat:14}$$
We introduce the dimensionless variables as follows $$\begin{aligned}
&x = r m_0, \quad \Lambda = m_0 \lambda, \quad e_0 \varphi(r) = m_0\phi(x), \quad \frac{e^2_0}{4 \pi} = \alpha_0,
\\
&u(x) \sqrt{m_0} = r g(r), \quad v(x)\sqrt{m_0} = r f(r),
\end{aligned}\label{eq:deriv-radi-equat:15}$$ which allows one to rewrite the system of equations (\[eq:deriv-radi-equat:14\]) in compact form $$\begin{aligned}
u^\prime(x) - \frac{u(x)}{x} - (1 - \phi(x)) v(x) &= \lambda v(x),
\\
v^\prime(x) + \frac{v(x)}{x} - (1 + \phi(x)) u(x) &= -\lambda u(x).
\end{aligned}\label{eq:deriv-radi-equat:16}$$
In order to complete our derivation we need to add the equation for the potential. For this purpose, we note that the self-consistent potential satisfies the Poisson equation $$\begin{aligned}
\label{eq:deriv-radi-equat:17}
\Delta\varphi({\boldsymbol}x) = -e_{0 | end{aligned}\label{eq: deriv - radi - equat:13}$$ which is cast into the final form through the use of the identity $ f^\prime(r) = (( rf(r))^\prime - f(r))/r$ $ $ \begin{aligned }
& (rg(r))^\prime - \frac{rg(r)}{r } - (m_0 - e_0\varphi(r))rf(r) = \Lambda f(r),
\\
& (rf(r))^\prime + \frac{rf(r)}{r } - (m_0 + e_0 \varphi(r) ) rg(r) = -\Lambda g(r).
\end{aligned}\label{eq: deriv - radi - equat:14}$$
We insert the dimensionless variable star as follows $ $ \begin{aligned }
& x = r m_0, \quad \Lambda = m_0 \lambda, \quad e_0 \varphi(r) = m_0\phi(x), \quad \frac{e^2_0}{4 \pi } = \alpha_0,
\\
& u(x) \sqrt{m_0 } = r g(r), \quad v(x)\sqrt{m_0 } = gas constant f(r),
\end{aligned}\label{eq: deriv - radi - equat:15}$$ which allows one to rewrite the system of equation (\[eq: deriv - radi - equat:14\ ]) in compact form $ $ \begin{aligned }
u^\prime(x) - \frac{u(x)}{x } - (1 - \phi(x) ) v(x) & = \lambda v(x),
\\
v^\prime(x) + \frac{v(x)}{x } - (1 + \phi(x) ) u(x) & = -\lambda u(x).
\end{aligned}\label{eq: deriv - radi - equat:16}$$
In ordering to complete our derivation we want to add the equation for the electric potential. For this purpose, we note that the self - consistent potential satisfies the Poisson equation $ $ \begin{aligned }
\label{eq: deriv - radi - equat:17 }
\Delta\varphi({\boldsymbol}x) = -e_{0 | end{wligned}\label{eq:deriv-radi-eduat:13}$$ which is ccwt intm the rinal fofm through the use of the idxntiry $f^\peime(r) = ((rf(r))^\prime - f(r))/r$ $$\bdgin{alignvd}
&(rg(r))^\primt - \frac{rg(r)}{r} - (m_0 - e_0\tzrphi(r))rn(x) = \Lajnda f(x),
\\
&(rf(r))^\primg + \frac{rf(r)}{r} - (m_0 + e_0 \varphi(r) )sg(f) = -\Lambda g(r).
\end{aligned}\label{eq:deriv-rwdi-equay:14}$$
Wf introduce thg dimtnsyonlsss variables as follows $$\begin{alighed}
&q = r m_0, \quad \Lambds = m_0 \lambda, \quad e_0 \varphi(r) = m_0\pji(x), \quad \frac{e^2_0}{4 \pi} = \alpha_0,
\\
&u(v) \sqrt{m_0} = r g(r), \quad v(x)\sqgc{m_0} = r f(r),
\ehd{aligned}\label{eq:deriv-radi-equat:15}$$ dhich allows ong fo tewrite the wystei of equatiokx (\[eq:desiv-radi-rquat:14\]) in compagt focm $$\bwgin{aligned}
u^\prime(e) - \frac{u(x)}{x} - (1 - \phi(x)) v(v) &= \lambda v(r),
\\
v^\prime(x) + \frax{v(z)}{x} - (1 + \phi(f)) u(x) &= -\laobdz n(x).
\end{allgnxd}\label{eq:dediv-radi-equar:16}$$
In order to compleue jlt derivation we neqd to add the equation for the potential. Xor this purpose, we note tyat the self-consistenj potentiaj satisfies the Poisson equation $$\begin{aligned}
\latel{eq:vefiv-xqdi-eqjqt:17}
\Delta\varphi({\boldsymbol}x) = -e_{0 | end{aligned}\label{eq:deriv-radi-equat:13}$$ which is cast into the final the of the $f^\prime(r) = ((rf(r))^\prime \frac{rg(r)}{r} (m_0 - e_0\varphi(r))rf(r) \Lambda f(r), \\ + \frac{rf(r)}{r} - (m_0 + e_0 )rg(r) = -\Lambda g(r). \end{aligned}\label{eq:deriv-radi-equat:14}$$ We introduce the dimensionless variables as follows $$\begin{aligned} = r m_0, \quad \Lambda = m_0 \lambda, \quad e_0 \varphi(r) = m_0\phi(x), \frac{e^2_0}{4 = \\ \sqrt{m_0} = r g(r), \quad v(x)\sqrt{m_0} = r f(r), \end{aligned}\label{eq:deriv-radi-equat:15}$$ which allows one to rewrite the of equations (\[eq:deriv-radi-equat:14\]) in compact form $$\begin{aligned} u^\prime(x) \frac{u(x)}{x} - (1 - v(x) &= \lambda v(x), \\ + - (1 \phi(x)) &= u(x). \end{aligned}\label{eq:deriv-radi-equat:16}$$ In to complete our derivation we need to add the equation for the potential. For this purpose, we that the satisfies the equation \label{eq:deriv-radi-equat:17} = -e_{0 | end{aligned}\label{eq:deriv-radI-equat:13}$$ whicH is caSt iNto ThE finAl foRm through the usE Of thE identity $f^\prime(r) = ((rf(r))^\priMe - f(r))/r$ $$\BeGIn{alIGnEd}
&(rg(r))^\Prime - \frAC{rG(R)}{R} - (m_0 - e_0\VaRpHi(r))Rf(R) = \laMbda f(R),
\\
&(rf(R))^\prime + \fRac{rf(r)}{r} - (m_0 + e_0 \vArpHi(R) )rg(r) = -\Lambda g(r).
\ENd{Aligned}\labEl{eQ:deriv-radi-eqUat:14}$$
we intrOdUce THe dimEnsIonleSs variABles as Follows $$\beGiN{AligneD}
&X = r m_0, \quad \lAMbDa = m_0 \lAmbda, \quad e_0 \varphi(r) = M_0\PhI(X), \quad \frac{e^2_0}{4 \pi} = \alPha_0,
\\
&u(x) \sQrT{M_0} = r G(R), \QuaD v(x)\Sqrt{m_0} = r f(r),
\enD{aLigneD}\Label{eq:DErIV-RAdi-EQuat:15}$$ which alloWs one to rewrITe tHe systEm Of eQUationS (\[eq:deRiV-RadI-equat:14\]) in comPact Form $$\begin{AligneD}
U^\prime(x) - \FRac{u(x)}{x} - (1 - \pHi(x)) v(x) &= \lAmbDa v(X),
\\
v^\prIMe(X) + \fRac{V(x)}{X} - (1 + \Phi(X)) U(x) &= -\LamBDa u(X).
\end{aligNeD}\lAbel{eQ:derIV-RADi-eqUat:16}$$
in orDer to Complete our deRivAtioN We nEed to Add thE equAtIon foR the poTentiAl. for this purpose, wE notE that the sElf-CoNsiStEnt poTEntial SatIsfIes the POisson eQUatIoN $$\BEGiN{aligned}
\label{eq:derIv-RADi-Equat:17}
\DelTa\varpHI({\bOlDSymbol}x) = -e_{0 | end{aligned}\label{eq:deri v-radi-equ at:13 }$$ wh ic h is cas t into the fin a l fo rm through the use ofthe i de n tity $f ^\pri me(r) = (( r f (r) )^ \p rim e- f (r))/ r$$$\begi n{aligned}
&(rg(r))^\pr i me - \frac{r g(r )}{r} - (m_0 -e_0\va rp hi( r ))rf( r)= \La mbda f ( r),
\\
&( r f(r))^ \ prime + \ fr ac{r f(r)}{r} - (m_0 + e_ 0 \varphi(r) )r g(r) = - \ La m b dag(r ).
\end{ al igned } \label{ e q: d e r iv- r adi-equat:14} $$
We intr o duc e thedi men s ionles s var ia b les as follows $$\ begin{ali gned} &x = r m_0,\quad\La mbd a =m _0 \ lam bd a , \ q ua d e _ 0 \ varphi(r )=m_0\p hi(x ) , \ quad \f rac{ e^2_0 }{4 \pi} = \a lph a_0, \\
&u (x)\s qrt{m _0} =r g(r ), \quad v(x)\sqr t{m_ 0} = r f( r),
\e nd {alig n ed}\la bel {eq :deriv- radi-eq u at: 15 } $ $ w hich allows one to r e w ri te the s ystemo feq u ations ( \[ eq: deri v - radi- equa t :1 4\]) incompac t f or m $$\be gi n{alig ne d} u^\p r ime( x) - \ frac{u(x )}{x} - (1 - \phi(x) ) v(x) &= \lam b da v (x ) ,
\ \
v^\pr ime( x ) +\fra c {v (x) } {x} - (1 + \ p hi ( x)) u(x) &= -\lambd au(x). \en d{aligned}\la bel{eq:der i v - radi-equ at:1 6 }$ $
In order tocompl ete our de r ivationwe ne ed to ad d the equ a t ion forthe po ten tia l . F or this purpo s e , we n ote tha t t he self -co nsi ste ntpo tential s atisfies t he P oi sso n equ a tion $$\ be gin {a lig ned} \labe l{eq: deri v- ra d i-e quat:17 }
\ Delt a\ va rphi ({\ bo ldsym bol} x ) = -e_{0 | end{aligned}\label{eq:deriv-radi-equat:13}$$ which_is cast_into the final form_through the_use_of the_identity_$f^\prime(r) = ((rf(r))^\prime_- f(r))/r$ $$\begin{aligned}
_ &(rg(r))^\prime_- \frac{rg(r)}{r} -_(m_0_- e_0\varphi(r))rf(r) = \Lambda f(r),
\\
&(rf(r))^\prime +_\frac{rf(r)}{r}_- (m_0_+_e_0_\varphi(r) )rg(r) = -\Lambda g(r).
_ \end{aligned}\label{eq:deriv-radi-equat:14}$$
We introduce the dimensionless_variables as_follows $$\begin{aligned}
&x = r_m_0,_\quad \Lambda =_m_0 \lambda, \quad e_0 \varphi(r) = m_0\phi(x), \quad \frac{e^2_0}{4_\pi} = \alpha_0,
_\\
__&u(x)_\sqrt{m_0} = r g(r),_\quad v(x)\sqrt{m_0} = r f(r),
_\end{aligned}\label{eq:deriv-radi-equat:15}$$ which allows one to rewrite_the system of equations (\[eq:deriv-radi-equat:14\]) in compact_form $$\begin{aligned}
u^\prime(x)_- \frac{u(x)}{x} - (1 -_\phi(x)) v(x)_&= \lambda v(x),
_ \\
_ _v^\prime(x) + \frac{v(x)}{x}_- (1 + \phi(x)) u(x) &=_-\lambda u(x).
_\end{aligned}\label{eq:deriv-radi-equat:16}$$
In order to complete our derivation_we_need to add_the_equation_for the_potential. For this_purpose,_we note_that_the self-consistent potential satisfies the Poisson_equation_$$\begin{aligned}
\label{eq:deriv-radi-equat:17}
\Delta\varphi({\boldsymbol}x) = -e_{0 |
set $\{{\mathfrak{m}}' \in \{0,\dots, \mathfrak{M}_p-1\}, \; k_{p,{\mathfrak{m}}'+1} \leq k_{p+1,{\mathfrak{m}}+1} \}$ is non empty and we can thus define the largest integer ${\mathfrak{m}}'$ of $\{0,\dots, \mathfrak{M}_p-1\}$ such that $ k_{p,{\mathfrak{m}}'+1} \leq k_{p+1,{\mathfrak{m}}+1} $. We then have $$k_{p,{\mathfrak{m}}'} = \sup \left\{k < k_{p,{\mathfrak{m}}'+1}, \; \mathfrak{j}_{\text{min}}^{(i,\ell_k)} > p \right\}.$$ Since $k_{p,{\mathfrak{m}}'} < k_{p+1,{\mathfrak{m}}+1}$, $$\begin{aligned}
k_{p,{\mathfrak{m}}'} &=& \sup \left\{k < k_{p+1,{\mathfrak{m}}+1}, \; \mathfrak{j}_{\text{min}}^{(i,\ell_k)} > p \right\} \\
&\geq& \sup \left\{k < k_{p+1,{\mathfrak{m}}+1}, \; \mathfrak{j}_{\text{min}}^{(i,\ell_k)} > p + 1 \right\} \\
&\geq& k_{p+1,{\mathfrak{m}}}.\end{aligned}$$ Hence, $k_{p,{\mathfrak{m}}'+1} \geq k_{p,{\mathfrak{m}}'} + 1 \geq k_{p+1,{\mathfrak{m}}} + 1$. Finally, $k_{p,{\mathfrak{m}}'+1} \in K_{p,{\mathfrak{m}}}$.
\[Proof of Claim \[ClaimPreuveAlgoDim2Kappa0\]\] Let $i \in \{1,\dots,N-1\}$ and $\ell \in \{1,\dots,L_i\}$. Then, $$\begin{aligned}
{\text{diam}}(\Theta_i) &\leq& \sup_{1 \leq j \leq d} { \begingroup
\def\mathaccent#R##2{ \rel@kern{0.8} \overline{\rel@kern{-0.8}\macc@nucleus\ | set $ \{{\mathfrak{m } }' \in \{0,\dots, \mathfrak{M}_p-1\ }, \; k_{p,{\mathfrak{m}}'+1 } \leq k_{p+1,{\mathfrak{m}}+1 } \}$ is non empty and we can thus define the bombastic integer $ { \mathfrak{m}}'$ of $ \{0,\dots, \mathfrak{M}_p-1\}$ such that $ k_{p,{\mathfrak{m}}'+1 } \leq k_{p+1,{\mathfrak{m}}+1 } $. We then take $ $ k_{p,{\mathfrak{m } }' } = \sup \left\{k < k_{p,{\mathfrak{m}}'+1 }, \; \mathfrak{j}_{\text{min}}^{(i,\ell_k) } > p \right\}.$$ Since $ k_{p,{\mathfrak{m } }' } < k_{p+1,{\mathfrak{m}}+1}$, $ $ \begin{aligned }
k_{p,{\mathfrak{m } }' } & = & \sup \left\{k < k_{p+1,{\mathfrak{m}}+1 }, \; \mathfrak{j}_{\text{min}}^{(i,\ell_k) } > p \right\ } \\
& \geq & \sup \left\{k < k_{p+1,{\mathfrak{m}}+1 }, \; \mathfrak{j}_{\text{min}}^{(i,\ell_k) } > p + 1 \right\ } \\
& \geq & k_{p+1,{\mathfrak{m}}}.\end{aligned}$$ Hence, $ k_{p,{\mathfrak{m}}'+1 } \geq k_{p,{\mathfrak{m } }' } + 1 \geq k_{p+1,{\mathfrak{m } } } + 1$. Finally, $ k_{p,{\mathfrak{m}}'+1 } \in K_{p,{\mathfrak{m}}}$.
\[Proof of Claim \[ClaimPreuveAlgoDim2Kappa0\]\ ] Let $ i \in \{1,\dots, N-1\}$ and $ \ell \in \{1,\dots, L_i\}$. Then, $ $ \begin{aligned }
{ \text{diam}}(\Theta_i) & \leq & \sup_{1 \leq j \leq d } { \begingroup
\def\mathaccent#R##2 { \rel@kern{0.8 } \overline{\rel@kern{-0.8}\macc@nucleus\ | seh $\{{\mathfrak{m}}' \in \{0,\dots, \mathnrak{M}_p-1\}, \; k_{p,{\mathftaj{m}}'+1} \leq k_{p+1,{\matgfrak{m}}+1} \}$ ks non empty and we can thus dwfine the largest integer ${\mxthfrak{m}}'$ lf $\{0,\dots, \nathhrak{M}_p-1\}$ such that $ k_{p,{\mathnxak{m}}'+1} \mcq k_{p+1,{\kethfrak{m}}+1} $. We thgn have $$k_{p,{\madhfrak{m}}'} = \sup \lafg\{k < k_{p,{\mathfrak{m}}'+1}, \; \mathfrak{j}_{\text{min}}^{(i,\ell_h)} > p \ribhh\}.$$ Since $k_{p,{\mathsrak{k}}'} < k_{p+1,{\jathfrak{m}}+1}$, $$\begin{aligned}
k_{p,{\mathfrak{m}}'} &=& \sup \leht\{k < k_{p+1,{\mathfral{m}}+1}, \; \mathfrak{j}_{\text{min}}^{(i,\ell_k)} > p \rlght\} \\
&\geq& \sup \left\{n < k_{p+1,{\mathftzk{m}}+1}, \; \mathfrak{j}_{\tebt{min}}^{(i,\ell_k)} > p + 1 \right\} \\
&\geq& k_{p+1,{\mathfrak{m}}}.\end{aligned}$$ Hencd, $k_{p,{\mcthfrak{m}}'+1} \gew j_{p,{\mwjhfrak{m}}'} + 1 \gew k_{p+1,{\mwthfrak{m}}} + 1$. Flmally, $n_{p,{\mathftak{m}}'+1} \in K_{p,{\mathnrak{m}}}$.
\['rood of Claim \[ClaimPreuveElgoDim2Kappa0\]\] Let $i \ig \{1,\dots,N-1\}$ atd $\ell \in \{1,\dots,L_i\}$. Thwn, $$\begit{alicned}
{\gwxt{aiaj}}(\Tietz_i) &\leq& \aup_{1 \leq j \meq d} { \begungroup
\def\mathacctnt#W##2{ \rel@kern{0.8} \overjige{\rel@kern{-0.8}\macc@nucleus\ | set $\{{\mathfrak{m}}' \in \{0,\dots, \mathfrak{M}_p-1\}, \; k_{p,{\mathfrak{m}}'+1} \}$ non empty we can thus of \mathfrak{M}_p-1\}$ such that k_{p,{\mathfrak{m}}'+1} \leq k_{p+1,{\mathfrak{m}}+1} We then have $$k_{p,{\mathfrak{m}}'} = \sup < k_{p,{\mathfrak{m}}'+1}, \; \mathfrak{j}_{\text{min}}^{(i,\ell_k)} > p \right\}.$$ Since $k_{p,{\mathfrak{m}}'} < k_{p+1,{\mathfrak{m}}+1}$, $$\begin{aligned} k_{p,{\mathfrak{m}}'} \sup \left\{k < k_{p+1,{\mathfrak{m}}+1}, \; \mathfrak{j}_{\text{min}}^{(i,\ell_k)} > p \right\} \\ &\geq& \sup \left\{k k_{p+1,{\mathfrak{m}}+1}, \mathfrak{j}_{\text{min}}^{(i,\ell_k)} p 1 \right\} \\ &\geq& k_{p+1,{\mathfrak{m}}}.\end{aligned}$$ Hence, $k_{p,{\mathfrak{m}}'+1} \geq k_{p,{\mathfrak{m}}'} + 1 \geq k_{p+1,{\mathfrak{m}}} + 1$. Finally, \in K_{p,{\mathfrak{m}}}$. \[Proof of Claim \[ClaimPreuveAlgoDim2Kappa0\]\] Let $i \{1,\dots,N-1\}$ and $\ell \in Then, $$\begin{aligned} {\text{diam}}(\Theta_i) &\leq& \sup_{1 j d} { \def\mathaccent#R##2{ \overline{\rel@kern{-0.8}\macc@nucleus\ | set $\{{\mathfrak{m}}' \in \{0,\dots, \mathfraK{M}_p-1\}, \; k_{p,{\mathfRak{m}}'+1} \lEq k_{P+1,{\maThFrak{M}}+1} \}$ is nOn empty and we caN Thus Define the largest integeR ${\mathFrAK{m}}'$ of $\{0,\DOtS, \mathFrak{M}_p-1\}$ sUCh THAt $ k_{P,{\mAtHfrAk{M}}'+1} \LeQ k_{p+1,{\maThfRak{m}}+1} $. We tHen have $$k_{p,{\mAthFrAk{m}}'} = \sup \left\{k < k_{P,{\MaThfrak{m}}'+1}, \; \matHfrAk{j}_{\text{min}}^{(i,\eLl_k)} > P \right\}.$$ siNce $K_{P,{\mathFraK{m}}'} < k_{p+1,{\mAthfraK{M}}+1}$, $$\begin{Aligned}
k_{p,{\MaTHfrak{m}}'} &=& \SUp \left\{k < K_{P+1,{\MaThfrAk{m}}+1}, \; \mathfrak{j}_{\text{mIN}}^{(i,\ELl_k)} > p \right\} \\
&\geq& \suP \left\{k < K_{p+1,{\MAtHFRak{M}}+1}, \; \maThfrak{j}_{\texT{mIn}}^{(i,\elL_K)} > p + 1 \right\} \\
&\GEq& K_{P+1,{\MAthFRak{m}}}.\end{aligneD}$$ Hence, $k_{p,{\matHFraK{m}}'+1} \geq k_{P,{\mAthFRak{m}}'} + 1 \geQ k_{p+1,{\maThFRak{M}}} + 1$. Finally, $k_{p,{\mAthfRak{m}}'+1} \in K_{p,{\mAthfraK{M}}}$.
\[Proof oF claim \[ClAimPreUvealgODim2kApPa0\]\] let $I \iN \{1,\DotS,n-1\}$ aNd $\eLL \in \{1,\Dots,L_i\}$. ThEn, $$\BeGin{alIgneD}
{\TEXT{diaM}}(\ThEta_i) &\Leq& \suP_{1 \leq j \leq d} { \begiNgrOup
\dEF\maThaccEnt#R##2{ \rEl@keRn{0.8} \OverlIne{\rel@Kern{-0.8}\mAcC@nucleus\ | set $\{{\mathfrak{m}}' \i n \{0,\dot s, \m ath fra k{ M}_p -1\} , \; k_{p,{\m a thfr ak{m}}'+1} \leq k_{p+1 ,{\ma th f rak{ m }} +1} \ }$ is n o ne m pty a nd we c a nthusdef ine the largest i nte ge r ${\mathfra k {m }}'$ of $\ {0, \dots, \math fra k{M}_p -1 \}$ suchtha t $ k _{p,{\ m athfra k{m}}'+1} \ l eq k_{ p +1,{\ma t h fr ak{m }}+1} $. We thenh av e $$k_{p,{\math frak{m }} ' }= \su p \ left\{k <k_ {p,{\ m athfrak { m} } ' + 1}, \; \mathfrak{ j}_{\text{m i n}} ^{(i,\ el l_k ) } > p\righ t\ } .$$ Since $k_{ p,{\ mathfrak{ m}}'}< k_{p+1 , {\mathf rak{m} }+1 }$, $$\ b eg in {al ig n ed} k_ {p, { \ma thfrak{m }} '} &=&\sup \ l e ft\{ k < k_ {p+1, {\mathfrak{m} }+1 }, \ ; \m athfr ak{j} _{\t ex t{min }}^{(i ,\ell _k )} > p \right\} \\
&\geq& \ sup \ lef t\ {k <k_{p+1 ,{\ mat hfrak{m }}+1},\ ; \ ma t h f ra k{j}_{\text{min}}^ {( i , \e ll_k)} > p + 1 \r ig h t\} \\&\ geq & k_ { p +1,{\ math f ra k{m}}}.\ end{al i gn ed }$$ Hen ce , $k_{ p, {\m ath frak{ m }}'+ 1} \ge q k_{p,{ \math f rak{m}}'} + 1\ geq k_{p+1,{\ m at h f ra k {m}} } + 1$. Finall y, $ k _{p, {\ma t hf rak { m}}'+ 1} \i nK _{ p ,{\mathfrak{m}}}$.
\ [Proof of C laim \[ClaimP reuveAlgoD i m 2 Kappa0\] \] L e t$ i \in \{1,\dot s,N-1 \}$ and $\ e ll \in \ {1,\d ots,L_i\ }$. Then, $ $\begin{ ali gne d}{\t e x t{ diam}}(\Theta _ i ) &\ le q& \s up_ {1 \leq j\le q d } { \begingro up
\de f\ ma th ac cen t#R## 2 { \re l@ ker n{ 0.8 } \ overli ne{\r el@k er n{ - 0.8 }\macc@ n uc l e us\ | set_$\{{\mathfrak{m}}' \in_\{0,\dots, \mathfrak{M}_p-1\}, \; _k_{p,{\mathfrak{m}}'+1} \leq_k_{p+1,{\mathfrak{m}}+1}_\}$ is_non_empty and we_can thus define_the largest integer ${\mathfrak{m}}'$_of $\{0,\dots, \mathfrak{M}_p-1\}$_such_that $ k_{p,{\mathfrak{m}}'+1} \leq k_{p+1,{\mathfrak{m}}+1} $. We then have $$k_{p,{\mathfrak{m}}'} = \sup \left\{k <_k_{p,{\mathfrak{m}}'+1},_\; \mathfrak{j}_{\text{min}}^{(i,\ell_k)}_>_p_\right\}.$$ Since $k_{p,{\mathfrak{m}}'} < k_{p+1,{\mathfrak{m}}+1}$,_$$\begin{aligned}
k_{p,{\mathfrak{m}}'} &=& \sup \left\{k <_ k_{p+1,{\mathfrak{m}}+1},_\; \mathfrak{j}_{\text{min}}^{(i,\ell_k)} > p \right\} \\
&\geq& \sup_\left\{k_< k_{p+1,{\mathfrak{m}}+1},_\; \mathfrak{j}_{\text{min}}^{(i,\ell_k)} > p + 1 \right\} \\
&\geq&_k_{p+1,{\mathfrak{m}}}.\end{aligned}$$ Hence, $k_{p,{\mathfrak{m}}'+1} \geq k_{p,{\mathfrak{m}}'} +_1 \geq k_{p+1,{\mathfrak{m}}}_+_1$._Finally, $k_{p,{\mathfrak{m}}'+1} \in K_{p,{\mathfrak{m}}}$.
\[Proof_of Claim \[ClaimPreuveAlgoDim2Kappa0\]\] Let $i \in \{1,\dots,N-1\}$_and $\ell \in \{1,\dots,L_i\}$. Then, $$\begin{aligned}
{\text{diam}}(\Theta_i)_&\leq& \sup_{1 \leq j \leq_d} { \begingroup
\def\mathaccent#R##2{_ \rel@kern{0.8} _ _\overline{\rel@kern{-0.8}\macc@nucleus\ |
icheletti, L. Miramonti, D. Mockler, S. Mollerach, F. Montanet, C. Morello, G. Morlino, M. Mostafá, A.L. Müller, M.A. Muller, S. Müller, R. Mussa, M. Muzio, W.M. Namasaka, L. Nellen, M. Niculescu-Oglinzanu, M. Niechciol, D. Nitz, D. Nosek, V. Novotny, L. Nožka, A Nucita, L.A. Núñez, M. Palatka, J. Pallotta, M.P. Panetta, P. Papenbreer, G. Parente, A. Parra, M. Pech, F. Pedreira, J. Pȩkala, R. Pelayo, J. Peña-Rodriguez, L.A.S. Pereira, J. Perez Armand, M. Perlin, L. Perrone, C. Peters, S. Petrera, T. Pierog, M. Pimenta, V. Pirronello, M. Platino, B. Pont, M. Pothast, P. Privitera, M. Prouza, A. Puyleart, S. Querchfeld, J. Rautenberg, D. Ravignani, M. Reininghaus, J. Ridky, F. Riehn, M. Risse, P. Ristori, V. Rizi, W. Rodrigues de Carvalho, J. Rodriguez Rojo, M.J. Roncoroni, M. Roth, E. Roulet, A.C. Rovero, P. Ruehl, S.J. Saffi, A. Saftoiu, F. Salamida, H. Salazar, G. Salina | icheletti, L. Miramonti, D. Mockler, S. Mollerach, F. Montanet, C. Morello, G. Morlino, M. Mostafá, A.L. Müller, M.A. Muller, S. Müller, R. Mussa, M. Muzio, W.M. Namasaka, L. Nellen, M. Niculescu - Oglinzanu, M. Niechciol, D. Nitz, D. Nosek, V. Novotny, L. Nožka, A Nucita, L.A. Núñez, M. Palatka, J. Pallotta, M.P. Panetta, P. Papenbreer, G. Parente, A. Parra, M. Pech, F. Pedreira, J. Pȩkala, R. Pelayo, J. Peña - Rodriguez, L.A.S. Pereira, J. Perez Armand, M. Perlin, L. Perrone, C. Peters, S. Petrera, T. Pierog, M. Pimenta, V. Pirronello, M. Platino, B. Pont, M. Pothast, P. Privitera, M. Prouza, A. Puyleart, S. Querchfeld, J. Rautenberg, D. Ravignani, M. Reininghaus, J. Ridky, F. Riehn, M. Risse, P. Ristori, V. Rizi, W. Rodrigues de Carvalho, J. Rodriguez Rojo, M.J. Roncoroni, M. Roth, E. Roulet, A.C. Rovero, P. Ruehl, S.J. Saffi, A. Saftoiu, F. Salamida, H. Salazar, G. Salina | ichfletti, L. Miramonti, D. Mockltr, S. Mollerach, F. Mouranet, R. Morellk, G. Morlivo, M. Mostafá, A.L. Müller, M.A. Muller, D. Müoler, E. Mussa, M. Muzio, W.M. Namasaya, L. Nellej, M. Niculwscu-Itlinzanu, M. Ijechciol, D. Nita, D. Novxk, V. Novotny, L. Nožla, A Nucitd, L.A. Núñez, M. Palatna, J. 'allotta, M.P. Panetta, P. Papenbreer, G. Paregte, A. Patrw, M. Pech, F. Pedreyra, K. [ȩkalz, R. Pelayo, J. Peña-Rodriguez, L.A.S. Pereira, J. Perez Armand, M. Perlon, L. Perrone, C. Peters, S. Petrega, T. Oierog, M. Pimenta, V. Plrronello, M. Kmatybo, B. Pont, M. Poghast, P. Prinntera, M. Prouea, A. Puyleart, S. Querchfeld, J. Rautencerg, B. Ravignani, N. Rwinltghaus, J. Ridjy, F. Rpehn, M. Risse, P. Vpstori, E. Rizi, W. Todrigues de Csrvelho, J. Rodriguez Rojo, M.J. Roicoroni, M. Roth, E. Roulej, A.C. Rovero, P. Xuehl, S.J. Saffi, A. Saftoiy, D. Salakida, H. Sauqzaf, G. Aakiha | icheletti, L. Miramonti, D. Mockler, S. Mollerach, C. G. Morlino, Mostafá, A.L. Müller, Mussa, Muzio, W.M. Namasaka, Nellen, M. Niculescu-Oglinzanu, Niechciol, D. Nitz, D. Nosek, V. L. Nožka, A Nucita, L.A. Núñez, M. Palatka, J. Pallotta, M.P. Panetta, P. G. Parente, A. Parra, M. Pech, F. Pedreira, J. Pȩkala, R. Pelayo, J. L.A.S. J. Armand, Perlin, L. Perrone, C. Peters, S. Petrera, T. Pierog, M. Pimenta, V. Pirronello, M. Platino, B. M. Pothast, P. Privitera, M. Prouza, A. Puyleart, Querchfeld, J. Rautenberg, D. M. Reininghaus, J. Ridky, F. M. P. Ristori, Rizi, Rodrigues Carvalho, J. Rodriguez M.J. Roncoroni, M. Roth, E. Roulet, A.C. Rovero, P. Ruehl, S.J. Saffi, A. Saftoiu, F. Salamida, H. G. Salina | icheletti, L. Miramonti, D. MocklEr, S. MolleraCh, F. MoNtaNet, c. MOrelLo, G. MOrlino, M. Mostafá, a.l. MülLer, M.A. Muller, S. Müller, R. MusSa, M. MuZiO, w.M. NaMAsAka, L. NEllen, M. NICuLEScu-ogLiNzaNu, m. niEchciOl, D. nitz, D. NoSek, V. NovotnY, L. NOžKa, A Nucita, L.A. NÚÑeZ, M. Palatka, J. palLotta, M.P. PanetTa, P. papenbReEr, G. pArentE, A. PArra, M. pech, F. PEDreira, j. Pȩkala, R. PElAYo, J. PeñA-rodriguEZ, l.A.s. PerEira, J. Perez Armand, M. pErLIn, L. Perrone, C. PetErs, S. PeTrERa, t. pIerOg, M. pimenta, V. PiRrOnellO, m. PlatinO, b. PONT, m. PoTHast, P. PriviterA, M. Prouza, A. PuYLeaRt, S. QueRcHfeLD, J. RautEnberG, D. rAviGnani, M. ReiniNghaUs, J. Ridky, F. riehn, M. rIsse, P. RiSTori, V. RiZi, W. RodRigUes De CaRVaLhO, J. ROdRIguEZ ROjo, m.j. RoNcoroni, M. roTh, e. RoulEt, A.C. rOVERo, P. RUehL, S.J. SAffi, A. saftoiu, F. SalamIda, h. SalAZar, g. SaliNa | icheletti, L. Miramonti, D . Mockler, S. M oll era ch , F. Mon tanet, C. More l lo,G. Morlino, M. Mostafá , A.L .M ülle r ,M.A.Muller, S. M üll er ,R.Mu s sa , M.Muz io, W.M . Namasaka , L .Nellen, M. N i cu lescu-Ogli nza nu, M. Niech cio l, D.Ni tz, D. No sek , V.Novotn y , L. N ožka, A N uc i ta, L. A . Núñez , M. Pal atka, J. Pallotta , M . P. Panetta, P. Papen br e er , G.Par ente, A. P ar ra, M . Pech,F .P e d rei r a, J. Pȩkala, R. Pelayo, J.Peña-R od rig u ez, L. A.S.Pe r eir a, J. Perez Arm and, M. P erlin, L. Perr o ne, C.Peters , S . P etre r a, T . P ie r og, M. Pi m ent a, V. Pi rr on ello, M.P l a t ino, B. Pon t, M. Pothast, P.Pri vite r a,M. Pr ouza, A.Pu ylear t, S.Querc hf eld, J. Rautenb erg, D. Ravig nan i, M. R einin g haus,J.Rid ky, F.Riehn,M . R is s e , P . Ristori, V. Rizi ,W . R odrigues de Ca r va lh o , J. Rod ri gue z Ro j o , M.J . Ro n co roni, M. Roth, E. R oulet,A. C. Rov er o,P.Ruehl , S.J . Saff i, A. Sa ftoiu , F. Salamida,H . Salazar, G. Sa l i na | icheletti, L. Miramonti,_D. Mockler, S. Mollerach,_F. Montanet, C. Morello, G. Morlino, M. Mostafá,_A.L. Müller, M.A. Muller,_S. Müller,_R. Mussa, M. Muzio,_W.M. Namasaka,_L. Nellen, M. Niculescu-Oglinzanu, M. Niechciol,_D. Nitz, D. Nosek, V. Novotny,_L. Nožka, A Nucita, L.A. Núñez,_M. Palatka, J. Pallotta, M.P. Panetta,_P. Papenbreer,_G. Parente, A. Parra, M. Pech, F. Pedreira, J. Pȩkala, R. Pelayo, J. Peña-Rodriguez, L.A.S. Pereira, J. Perez Armand, M. Perlin, L. Perrone, C. Peters, S. Petrera,_T. Pierog,_M. Pimenta, V. Pirronello,_M. Platino,_B. Pont,_M. Pothast, P. Privitera, M. Prouza, A. Puyleart, S. Querchfeld,_J. Rautenberg, D. Ravignani, M. Reininghaus, J. Ridky, F. Riehn,_M. Risse, P. Ristori,_V. Rizi, W. Rodrigues de Carvalho, J. Rodriguez Rojo, M.J. Roncoroni, M. Roth,_E. Roulet,_A.C. Rovero, P. Ruehl, S.J. Saffi,_A. Saftoiu, F. Salamida, H. Salazar, G. Salina |
), discussed above. Moreover the vanishing of the residue for exchanges of the corresponding Goldstone particle might fit the vanishing of the monopole condensate in the limit $a\to 0$, discussed in Sect. 2.2.
Acknowledgments
===============
The work of VIZ is partially supported by grants Leading Scientific Schools NSh-6260.2010.2, RFBR -11-02-01227-à and Federal Special-Purpose Programme ’Cadres’ of the Russian Ministry of Science and Education.
[99]{} A. M. Polyakov, Phys. Lett. [**B59**]{} 82 (1975);\
A. M. Polyakov, Nucl. Phys. [**B120**]{} (1977) 429.
J. Greensite, Prog. Part. Nucl. Phys. [**51**]{} (2003) 1. \[arXiv:hep-lat/0301023\].
A. M. Polyakov, “Gauge fields and Strings,” Harwood Academic Publishers (1987).
V.I. Zakharov, in [*Sense of Beauty in Physics: Miniconference in Honor of Adriano Di Giacomo on his 70th Birthday*]{}, Pisa, Italy, 26-27 Jan 2006, \[arXiv:hep-ph/0602141\].
V.I. Zakharov, Braz. J. Phys. [**37**]{} (2007) 65, \[arXiv:hep-ph/0612342\];\
V.I. Zakharov, AIP Conf.Proc. [**756**]{} 182 (2005), \[arXiv:hep-ph/0501011\].
A. Gorsky, V. Zakharov, [Phys. Rev.]{} [**D77**]{} (2008) 045017, arXiv:0707.1284 \[hep-th\].
A. S. Gorsky, V. I. Zakharov, A. R. Zhitnitsky, [ Phys. Rev.]{} [**D79**]{} (2009) 106003, arXiv:0902.1842 \[hep-ph\].
D. Teaney, Prog. Part. Nucl. Phys. [**62**]{} (2009) 451.
V.G. Bornyakov et al., Phys. Lett. [**B537**]{} (2002) 291, \[arXiv:hep-lat/0103032\];\
V.G. Bornyakov | ), discussed above. Moreover the vanishing of the residue for exchanges of the correspond Goldstone atom might fit the vanishing of the monopole condensate in the terminus ad quem $ a\to 0 $, discussed in Sect. 2.2.
citation
= = = = = = = = = = = = = = =
The work of VIZ is partially confirm by grants Leading Scientific Schools NSh-6260.2010.2, RFBR -11 - 02 - 01227 - à and Federal Special - Purpose Programme ’ Cadres ’ of the Russian Ministry of Science and Education.
[ 99 ] { } A. M. Polyakov, Phys. Lett. [ * * B59 * * ] { } 82 (1975);\
A. M. Polyakov, Nucl. Phys. [ * * B120 * * ] { } (1977) 429.
J. Greensite, Prog. Part. Nucl. Phys. [ * * 51 * * ] { } (2003) 1. \[arXiv: hep - lat/0301023\ ].
A. M. Polyakov, “ Gauge airfield and Strings, ” Harwood Academic Publishers (1987).
V.I. Zakharov, in [ * Sense of Beauty in Physics: Miniconference in Honor of Adriano Di Giacomo on his seventieth Birthday * ] { }, Pisa, Italy, 26 - 27 Jan 2006, \[arXiv: hep - ph/0602141\ ].
V.I. Zakharov, Braz. J. Phys. [ * * 37 * * ] { } (2007) 65, \[arXiv: hep - ph/0612342\];\
V.I. Zakharov, AIP Conf. Proc. [ * * 756 * * ] { } 182 (2005), \[arXiv: hep - ph/0501011\ ].
A. Gorsky, V. Zakharov, [ Phys. Rev. ] { } [ * * D77 * * ] { } (2008) 045017, arXiv:0707.1284 \[hep - th\ ].
A. S. Gorsky, V. I. Zakharov, A. R. Zhitnitsky, [ Phys. Rev. ] { } [ * * D79 * * ] { } (2009) 106003, arXiv:0902.1842 \[hep - ph\ ].
D. Teaney, Prog. Part. Nucl. Phys. [ * * 62 * * ] { } (2009) 451.
V.G. Bornyakov et al. , Phys. Lett. [ * * B537 * * ] { } (2002) 291, \[arXiv: hep - lat/0103032\];\
V.G. Bornyakov | ), didcussed above. Moreover tme vanishing of jhw resivue for exchangds of the corresponding Golddtine pqrticle might fit the xanishing of the nonokole condensate ii the limit $a\tk 0$, divrussed in Sect. 2.2.
Scknowledgkents
===============
The work mf VNZ is partially supported by grants Jeading Sfientific Schojls MFh-6260.2010.2, RRBR -11-02-01227-à and Federal Special-Purpose Prkgramme ’Cadres’ of thr Russian Ministry of Sciejce wnd Education.
[99]{} A. M. Oolyakov, Phia. Lqrt. [**B59**]{} 82 (1975);\
A. M. Pouyakov, Nucl. Phys. [**B120**]{} (1977) 429.
J. Freensite, Prog. Part. Nucl. Phys. [**51**]{} (2003) 1. \[arXnv:hep-lat/0301023\].
A. M. Pilywnov, “Gauge fmelds wnd Strings,” Msrwood Academoc Publishers (1987).
Y.I. Zanhaeov, in [*Sense of Beautb in Physics: Miniconserence it Konor of Adriano Di Guaxomo mn hhs 70tf Bifthsab*]{}, Pjsa, Itwly, 26-27 Jan 2006, \[arXjv:hep-ph/0602141\].
V.I. Zqkharov, Braz. J. Phys. [**37**]{} (2007) 65, \[arXiv:hep-ph/0612342\];\
V.I. Zakhawod, AIP Conf.Proc. [**756**]{} 182 (2005), \[arXiv:hep-ph/0501011\].
A. Gorsky, V. Eakhadov, [Phys. Rev.]{} [**D77**]{} (2008) 045017, arXiv:0707.1284 \[yep-th\].
A. S. Gorsky, V. I. Zwkharov, A. R. Zhitnitsky, [ Phys. Rev.]{} [**D79**]{} (2009) 106003, arXiv:0902.1842 \[hep-ph\].
D. Teaney, Psog. Perg. Nbgj. Phjs. [**62**]{} (2009) 451.
V.G. Bornyakov et al., Phys. Lett. [**B537**]{} (2002) 291, \[arXiv:he[-mau/0103032\];\
V.G. Bornyakov | ), discussed above. Moreover the vanishing of for of the Goldstone particle might monopole in the limit 0$, discussed in 2.2. Acknowledgments =============== The work of is partially supported by grants Leading Scientific Schools NSh-6260.2010.2, RFBR -11-02-01227-à and Federal Programme ’Cadres’ of the Russian Ministry of Science and Education. [99]{} A. M. Phys. [**B59**]{} (1975);\ M. Polyakov, Nucl. Phys. [**B120**]{} (1977) 429. J. Greensite, Prog. Part. Nucl. Phys. [**51**]{} (2003) 1. A. M. Polyakov, “Gauge fields and Strings,” Harwood Publishers (1987). V.I. Zakharov, [*Sense of Beauty in Physics: in of Adriano Giacomo his Birthday*]{}, Pisa, Italy, Jan 2006, \[arXiv:hep-ph/0602141\]. V.I. Zakharov, Braz. J. Phys. [**37**]{} (2007) 65, \[arXiv:hep-ph/0612342\];\ V.I. Zakharov, AIP Conf.Proc. [**756**]{} (2005), \[arXiv:hep-ph/0501011\]. V. Zakharov, Rev.]{} (2008) arXiv:0707.1284 \[hep-th\]. A. V. I. Zakharov, A. R. Zhitnitsky, [**D79**]{} (2009) 106003, arXiv:0902.1842 \[hep-ph\]. D. Teaney, Prog. Nucl. Phys. (2009) 451. V.G. Bornyakov et al., Lett. [**B537**]{} (2002) 291, \[arXiv:hep-lat/0103032\];\ V.G. Bornyakov | ), discussed above. Moreover the Vanishing oF the rEsiDue FoR excHangEs of the correspONdinG Goldstone particle mighT fit tHe VAnisHInG of thE monopoLE cONDenSaTe In tHe LImIt $a\to 0$, DisCussed iN Sect. 2.2.
AcknoWleDgMents
===============
The work OF ViZ is partiaLly Supported by gRanTs LeadInG ScIEntifIc SChoolS NSh-6260.2010.2, RFbr -11-02-01227-à and FEderal SpeCiAL-PurpoSE PrograMME ’CAdreS’ of the Russian MiniSTrY Of Science and EdUcatioN.
[99]{} A. m. poLYAkoV, PhYs. Lett. [**B59**]{} 82 (1975);\
A. M. POlYakov, nUcl. Phys. [**b120**]{} (1977) 429.
j. GREENsiTE, Prog. Part. Nucl. phys. [**51**]{} (2003) 1. \[arXiv:heP-Lat/0301023\].
a. M. PolyAkOv, “GAUge fieLds anD STRinGs,” Harwood AcAdemIc PublishErs (1987).
V.I. ZAKharov, iN [*sense of beauty In PHysIcs: MINiCoNfeReNCe iN hoNor OF AdRiano Di GIaCoMo on hIs 70th bIRTHday*]{}, pisA, ItaLy, 26-27 Jan 2006, \[ArXiv:hep-ph/0602141\].
V.I. ZAkhArov, bRaz. j. Phys. [**37**]{} (2007) 65, \[ArXiv:Hep-pH/0612342\];\
V.i. ZakhArov, AIp Conf.prOc. [**756**]{} 182 (2005), \[arXiv:hep-ph/0501011\].
A. GoRsky, v. Zakharov, [phyS. REv.]{} [**D77**]{} (2008) 045017, Arxiv:0707.1284 \[heP-Th\].
A. S. GoRskY, V. I. zakharoV, A. R. ZhitNItsKy, [ pHYS. REv.]{} [**D79**]{} (2009) 106003, arXiv:0902.1842 \[hep-ph\].
D. TeanEy, pROg. part. Nucl. phys. [**62**]{} (2009) 451.
V.G. bOrNyAKov et al., PHyS. LeTt. [**B537**]{} (2002) 291, \[aRxIv:hep-Lat/0103032\];\
V.g. boRnyakov | ), discussed above. Moreov er the van ishin g o f t he res idue for exchanges of t he corresponding Golds tonepa r ticl e m ightfit the va n i shi ng o f t he mo nopol e c ondensa te in thelim it $a\to 0$, d i sc ussed in S ect . 2.2.
Ackn owl edgmen ts
== = ===== === ====
The w o rk ofVIZ is pa rt i ally s u pported b ygran ts Leading Scient i fi c Schools NSh-6 260.20 10 . 2, R FBR -1 1-02-01227 -à andF ederalS pe c i a l-P u rpose Program me ’Cadres’ ofthe Ru ss ian Minist ry of S c ien ce and Educ atio n.
[99]{ } A. M . Polyak o v, Phys . Lett . [ **B 59** ] {} 8 2 ( 19 7 5); \
A . M . Po lyakov,Nu cl . Phy s. [ * * B 1 20** ]{} (19 77) 4 29.
J. Green sit e, P r og. Part . Nuc l. P hy s. [* *51**] {} (2 00 3) 1. \[arXiv:h ep-l at/030102 3\] .
A. M . Pol y akov,“Ga uge fields and St r ing s, ” H ar wood Academic Publ is h e rs (1987).
V.I. Za kh a rov, in[* Sen se o f Beaut y in Ph ysics: M inicon f er en ce in H on or ofAd ria noDi Gi a como on hi s 70th B irthd a y*]{}, Pisa, I t aly, 26-27 Ja n 2 0 0 6, \[ar Xiv :hep-ph/060 2141 \ ].
V.I. Za kha r ov, B raz.J. Ph y s. [**37**]{} (2007 )65, \[ arXiv :hep-ph/06123 42\];\
V.I . Z akharov, AIP Co n f.Proc. [**756 **]{} 182 (2005 ) , \[arXi v:hep -ph/0501 011\].
A . Gorsky,V.Zak har ov, [ Ph ys. Rev.]{} [ * * D77* *] {} (200 8)045017, ar Xiv :07 07. 12 84 \[hep- th\].
A .S. G or sky , V.I . Zakhar ov , A .R.Zhitn i tsky,[ Phy s. R ev .] { } [ **D79** ] {} ( 2009 )10 6003 , a rX iv:09 02.1 8 42\[hep-p h\].
D.Tea n ey,Pr og . Part. Nucl. Phys.[* *62**]{} ( 20 09) 451.V .G. Born yakov et al., Phys. Let t . [**B5 37* *]{}(200 2) 291, \ [ar Xiv:he p-l a t/0103 032\]; \
V.G .Bor n y akov | ), discussed_above. Moreover_the vanishing of the_residue for_exchanges_of the_corresponding_Goldstone particle might_fit the vanishing_of the monopole condensate_in the limit_$a\to_0$, discussed in Sect. 2.2.
Acknowledgments
===============
The work of VIZ is partially supported by grants Leading_Scientific_Schools NSh-6260.2010.2,_RFBR_-11-02-01227-à_and Federal Special-Purpose Programme ’Cadres’_of the Russian Ministry of_Science and_Education.
[99]{} A. M. Polyakov, Phys. Lett. [**B59**]{} 82_(1975);\
A._M. Polyakov, Nucl._Phys. [**B120**]{} (1977) 429.
J. Greensite, Prog. Part. Nucl. Phys._[**51**]{} (2003) 1. \[arXiv:hep-lat/0301023\].
A. M. Polyakov,_“Gauge fields and_Strings,”_Harwood_Academic Publishers (1987).
V.I. Zakharov,_in [*Sense of Beauty in Physics:_Miniconference in Honor of Adriano Di_Giacomo on his 70th Birthday*]{}, Pisa, Italy,_26-27 Jan 2006, \[arXiv:hep-ph/0602141\].
V.I. Zakharov, Braz._J. Phys. [**37**]{} (2007) 65,_\[arXiv:hep-ph/0612342\];\
V.I. Zakharov,_AIP Conf.Proc. [**756**]{} 182 (2005),_\[arXiv:hep-ph/0501011\].
A. Gorsky, V._Zakharov, [Phys._Rev.]{} [**D77**]{} (2008)_045017, arXiv:0707.1284 \[hep-th\].
A. S. Gorsky, V._I. Zakharov, A._R. Zhitnitsky, [ Phys. Rev.]{} [**D79**]{}_(2009)_106003, arXiv:0902.1842 \[hep-ph\].
D._Teaney,_Prog._Part. Nucl._Phys. [**62**]{} (2009)_451.
V.G._Bornyakov et_al.,_Phys. Lett. [**B537**]{} (2002) 291, \[arXiv:hep-lat/0103032\];\
V.G._Bornyakov |
gtrsim \rm{O}(100) \rm{AU}$) [e.g. @Boley_CA_and_GI] though a hybrid scenario of forming gas giants in the same system by both core accretion and gravitational instability has yet to be modelled.
There are two quantities that have historically been used to determine whether a disc is likely to fragment. The first is the stability parameter [@Toomre_stability1964],
$$\label{eq:Toomre}
Q=\frac{c_{\rm s}\kappa_{\rm ep}}{\pi\Sigma G},$$
where $c_{\rm s}$ is the sound speed in the disc, $\kappa_{\rm ep}$ is the epicyclic frequency, which for Keplerian discs is approximately equal to the angular frequency, $\Omega$, $\Sigma$ is the surface mass density and $G$ is the gravitational constant. Therefore, once the surface mass density and the rotation of the disc have been established, the stability is purely dependent on the disc temperature. [@Toomre_stability1964] showed that for an infinitesimally thin disc to fragment, the stability parameter must be less than a critical value, $Q_{\rm crit} \approx 1$.
[@Gammie_betacool] showed that in addition to the stability criterion above, the disc must cool at a fast enough rate. Using shearing sheet simulations, he showed that if the cooling timescale can be parametrized as
$$\label{eq:beta}
\beta = t_{\rm cool}\Omega,$$
where
$$\label{eq:tcool}
t_{\rm cool} = u \big(\frac{du_{\rm cool}}{dt}\big)^{-1},$$
$u$ is the internal energy and $du_{\rm cool}/dt$ is the total cooling rate, then for fragmentation we require $\beta \lesssim 3$, for a ratio of specific heats $\gamma = 2$ (in two dimensions). [-@Rice_beta_condition] carried out three-dimensional simulations using a smoothed particle hydrodynamics (SPH) code and showed that this cooling parameter is dependent on the equation of state so that fragmentation can occur for discs with $\gamma = 5/3$ and $7/5$ if $\beta \lesssim 7$ and $13$, respectively. [-@Libby_MSci] showed that the critical value of $\beta$ (below which fragmentation will occur if the stability | gtrsim \rm{O}(100) \rm{AU}$) [ e.g. @Boley_CA_and_GI ] though a hybrid scenario of forming gas giants in the like arrangement by both core accretion and gravitational instability experience even to be modelled.
There are two measure that have historically been used to determine whether a phonograph record is likely to fragment. The beginning is the stability parameter [ @Toomre_stability1964 ],
$ $ \label{eq: Toomre }
Q=\frac{c_{\rm s}\kappa_{\rm ep}}{\pi\Sigma G},$$
where $ c_{\rm s}$ is the reasoned speed in the disc, $ \kappa_{\rm ep}$ is the epicyclic frequency, which for Keplerian discs is approximately equal to the angular frequency, $ \Omega$, $ \Sigma$ is the surface mass concentration and $ G$ is the gravitational constant. Therefore, once the surface mass concentration and the rotation of the disc have been established, the stability is purely dependent on the phonograph record temperature. [ @Toomre_stability1964 ] showed that for an infinitesimally thin disc to fragment, the stability argument must be less than a critical value, $ Q_{\rm crit } \approx 1$.
[ @Gammie_betacool ] showed that in addition to the stability criterion above, the disc must cool at a fast adequate pace. practice shearing sheet simulations, he showed that if the cooling timescale can be parametrized as
$ $ \label{eq: beta }
\beta = t_{\rm cool}\Omega,$$
where
$ $ \label{eq: tcool }
t_{\rm cool } = u \big(\frac{du_{\rm cool}}{dt}\big)^{-1},$$
$ u$ is the internal energy and $ du_{\rm cool}/dt$ is the entire cooling system pace, then for fragmentation we require $ \beta \lesssim 3 $, for a ratio of specific heat $ \gamma = 2 $ (in two dimensions). [ -@Rice_beta_condition ] carried out three - dimensional simulations use a smoothed atom hydrodynamics (SPH) code and showed that this cooling parameter is pendent on the equation of state so that fragmentation can occur for discs with $ \gamma = 5/3 $ and $ 7/5 $ if $ \beta \lesssim 7 $ and $ 13 $, respectively. [ -@Libby_MSci ] showed that the critical value of $ \beta$ (below which fragmentation will occur if the stability | gtrdim \rm{O}(100) \rm{AU}$) [e.g. @Boley_CA_akd_GI] though a hyyeid scxnario kf formivg gas giants in the same sydtwm by both core accretion avd gravitwtional unstebility has yet vk be modelled.
Fmere crx two quantitiex that hava historically bdeu used to determine whether a disc if likelu ho fragment. Thg firxe is nht stability parameter [@Toomre_stabimity1964],
$$\labtl{eq:Toomre}
Q=\frac{c_{\tm s}\kappa_{\rm ep}}{\pi\Sigma G},$$
whege $c_{\gm s}$ is the sound dpeed in thg difx, $\kappa_{\rm ep}$ is the eppeyclic freqoency, which for Keplerian discs ks ap'roximately ewuap to the angnlar fgequency, $\Omega$, $\Sigma$ hs the xurface mass dcnsitb ane $G$ is the gravitatioial constant. Therefote, once tha aurface mass densutt and the rotxriov or vhe disc javx been estaglished, the stability is purelu qvlendent on tge diss eemperature. [@Toomre_stability1964] showed that fod an infinitesimally thun disc to fragment, tje stabilyty parameter must be less than a critical value, $X_{\rm ccig} \akpvjb 1$.
[@Gwmmie_betacool] showed that in addition to the ffanikity criterion above, the cidc iust cool at x fast enkugh rate. Using shfaring fheet simulatijns, ne showed that if the coolibg timescale xan be parametrizeb as
$$\label{eq:bzta}
\bgta = t_{\tm cool}\Omega,$$
where
$$\label{ed:tcokl}
t_{\rm coop} = u \big(\fdxc{du_{\rm cool}}{dt}\big)^{-1},$$
$j$ ix dhe internal energy and $du_{\wm cool}/dt$ is che totau copling wate, then vor fvdgmentation we reqkire $\yeta \nesssim 3$, flr a ratio of specific heats $\gamma = 2$ (in two domanspons). [-@Rice_yeta_cokdition] carried out three-dimeusional fimulxtions usihg a smmothed partycle hydrodyndlics (SPH) codx and shored rhat this cuuling parametet is depeubent on tye equation of stabe so fhat fragmentatnub can occur for dircs wpth $\gwkma = 5/3$ and $7/5$ ix $\bega \ursssio 7$ and $13$, resiecgiveky. [-@Libby_MSci] showed dhat the critical valur jf $\beta$ (velow whych fragmentayion will occur if the vtauility | gtrsim \rm{O}(100) \rm{AU}$) [e.g. @Boley_CA_and_GI] though a of gas giants the same system gravitational has yet to modelled. There are quantities that have historically been used determine whether a disc is likely to fragment. The first is the stability [@Toomre_stability1964], $$\label{eq:Toomre} Q=\frac{c_{\rm s}\kappa_{\rm ep}}{\pi\Sigma G},$$ where $c_{\rm s}$ is the sound speed the $\kappa_{\rm is epicyclic frequency, which for Keplerian discs is approximately equal to the angular frequency, $\Omega$, $\Sigma$ is surface mass density and $G$ is the gravitational Therefore, once the surface density and the rotation of disc been established, stability purely on the disc [@Toomre_stability1964] showed that for an infinitesimally thin disc to fragment, the stability parameter must be less than critical value, \approx 1$. showed in to the stability the disc must cool at a Using shearing sheet simulations, he showed that if cooling timescale be parametrized as $$\label{eq:beta} \beta = cool}\Omega,$$ where $$\label{eq:tcool} t_{\rm cool} = u \big(\frac{du_{\rm $u$ is the internal energy and $du_{\rm cool}/dt$ is the total cooling rate, then for require $\beta \lesssim 3$, a ratio of heats = (in dimensions). [-@Rice_beta_condition] out three-dimensional simulations using a smoothed particle hydrodynamics (SPH) code and that this cooling parameter is dependent on the equation of that can occur for with $\gamma = 5/3$ $7/5$ $\beta \lesssim 7$ and [-@Libby_MSci] that of (below fragmentation will occur if stability | gtrsim \rm{O}(100) \rm{AU}$) [e.g. @Boley_CA_and_gI] though a hYbrid SceNarIo Of foRminG gas giants in thE Same System by both core accretIon anD gRAvitATiOnal iNstabilITy HAS yeT tO bE moDeLLeD.
TherE arE two quaNtities thaT haVe Historically BEeN used to detErmIne whether a dIsc Is likeLy To fRAgmenT. ThE firsT is the STabiliTy parametEr [@tOomre_sTAbility1964],
$$\LABeL{eq:TOomre}
Q=\frac{c_{\rm s}\kapPA_{\rM Ep}}{\pi\Sigma G},$$
wherE $c_{\rm s}$ iS tHE sOUNd sPeeD in the disc, $\KaPpa_{\rm EP}$ is the ePIcYCLIc fREquency, which fOr Keplerian DIscS is appRoXimATely eqUal to ThE AngUlar frequenCy, $\OmEga$, $\Sigma$ iS the suRFace masS Density And $G$ is The GraVitaTIoNaL coNsTAnt. tHeRefORe, oNce the suRfAcE mass DensITY ANd thE roTatiOn of tHe disc have beeN esTablISheD, the sTabilIty iS pUrely DependEnt on ThE disc temperaturE. [@TooMre_stabilIty1964] ShOweD tHat foR An infiNitEsiMally thIn disc tO FraGmENT, ThE stability parameteR mUST bE less thaN a critICaL vALue, $Q_{\rm crIt} \AppRox 1$.
[@GAMMie_beTacoOL] sHowed thaT in addITiOn To the stAbIlity cRiTerIon Above, THe diSc must Cool at a fAst enOUgh rate. Using shEAring sheet simULaTIOnS, He shOweD that if the cOoliNG timEscaLE cAn bE ParamEtrizEd AS
$$\lABel{eq:beta}
\beta = t_{\rm cooL}\OMega,$$
whEre
$$\laBel{eq:tcool}
t_{\rm Cool} = u \big(\frAC{DU_{\rm cool}}{dT}\big)^{-1},$$
$U$ Is THe internal enerGy and $Du_{\rm cool}/dt$ IS the totaL coolIng rate, tHen for fraGMEntation We rEquIre $\BetA \LEsSsim 3$, for a ratio OF SpecIfIc heats $\GamMa = 2$ (in two DimEnsIonS). [-@RiCe_Beta_condiTion] carrIeD oUt ThRee-DimenSIonal simUlAtiOnS usIng a sMOothed PartiCle hYdRoDYnaMics (SPH) COdE ANd shOwEd That ThiS cOolinG parAMetEr is depEndent on tHe eQUatiOn Of State so That fragmentaTiOn can occur FoR diScs witH $\GAmma = 5/3$ and $7/5$ iF $\beta \lesssim 7$ and $13$, respectiVEly. [-@LibbY_MSCi] shoWed tHat the criTicAl valuE of $\BEta$ (belOw whicH fragMeNtaTIOn wilL OCcUr iF tHe stabilitY | gtrsim \rm{O}(100) \rm{AU} $) [e.g. @ Boley _CA _an d_ GI]thou gh a hybrid sc e nari o of forming gas giant s inth e sam e s ystem by bot h c o r e a cc re tio na nd grav ita tionalinstabilit y h as yet to be m o de lled.
The reare two quan tit ies th at ha v e his tor icall y been used t o determi ne whethe r a disc i slike ly to fragment. T h ef irst is the st abilit yp ar a m ete r [ @Toomre_st ab ility 1 964],
$ $\ l a b el{ e q:Toomre}
Q =\frac{c_{\ r m s }\kapp a_ {\r m ep}}{ \pi\S ig m a G },$$
where $c_ {\rm s}$is the sound s p eed inthe di sc, $\ kapp a _{ \r m e p} $ is th e e p icy clic fre qu en cy, w hich f o r Kep ler iandiscs is approxima tel y eq u alto th e ang ular f reque ncy, $ \Omeg a$ , $\Sigma$ is t he s urface ma ssde nsi ty and$ G$ isthe gr avitati onal co n sta nt . T he refore, once the s ur f a ce mass de nsitya nd t h e rotati on of the d isc h aveb ee n establ ished, th estabili ty is pu re lydep enden t onthe di sc tempe ratur e . [@Toomre_sta b ility1964] sh o we d th a t fo r a n infinites imal l y th in d i sc to fragm ent,th e s t ability parameter m us t be l ess t han a critica l value, $ Q _ { \rm crit } \a p pr o x 1$.
[@Gammi e_bet acool] sho w ed thatin ad dition t o the sta b i lity cri ter ion ab ove , th e disc must c o o l at a fast e nou gh rate . U sin g s hea ri ng sheetsimulati on s, h esho wed t h at if th ecoo li ngtimes c ale ca n bepara me tr i zed as
$$ \ la b e l{eq :b et a}
\b et a = t _{\r m co ol}\Ome ga,$$
wh ere
$$\ la be l{eq:tc ool}
t_{\rm c ool} = u \ bi g(\ frac{d u _ {\rm coo l}}{dt}\big)^{-1},$$
$ u $ is th e i ntern al e nergy and $d u_{\rm co o l}/dt$ is th e tot al co o l ing r a t e, th en for fragm e n tat ion w erequ ire $\b eta \lesssim 3$, f o r a ratio of spe cif ic h e a ts $\ g am m a = 2 $ (i n two dimensions) . [-@Rice_ be t a_ condition] car ri ed outthree-d imens i onal si mulations using asm ooth e d pa rticle hyd rodynami cs (SPH)c ode a n dshowe d t hat th is co oling param e ter is d epende nt on th e equ at ion of s tate so that fragmentat ion ca n occ urfor discs wi t h $ \gamma =5/3$ and $7/5$ if $\ beta\le s ssim7$ a n d$13 $ , res pect i vely. [-@ L ib by_ M S ci ] showed th a t the crit ica l value of$\beta$ (below wh i ch fragmentati on w i l l o ccu r ifth e stability | gtrsim \rm{O}(100)_\rm{AU}$) [e.g._@Boley_CA_and_GI] though a hybrid_scenario of_forming_gas giants_in_the same system_by both core_accretion and gravitational instability_has yet to_be_modelled.
There are two quantities that have historically been used to determine whether a disc_is_likely to_fragment._The_first is the stability parameter_[@Toomre_stability1964],
$$\label{eq:Toomre}
Q=\frac{c_{\rm s}\kappa_{\rm ep}}{\pi\Sigma_G},$$
where $c_{\rm_s}$ is the sound speed in the disc,_$\kappa_{\rm_ep}$ is the_epicyclic frequency, which for Keplerian discs is approximately equal_to the angular frequency, $\Omega$, $\Sigma$_is the surface_mass_density_and $G$ is the_gravitational constant. Therefore, once the surface_mass density and the rotation of_the disc have been established, the stability_is purely dependent on the disc_temperature. [@Toomre_stability1964] showed that for_an infinitesimally_thin disc to fragment, the_stability parameter must_be less_than a critical_value, $Q_{\rm crit} \approx 1$.
[@Gammie_betacool] showed_that in addition_to the stability criterion above, the_disc_must cool at_a_fast_enough rate._Using shearing sheet_simulations,_he showed_that_if the cooling timescale can be_parametrized_as
$$\label{eq:beta}
\beta = t_{\rm cool}\Omega,$$
where
$$\label{eq:tcool}
_t_{\rm cool} = u_\big(\frac{du_{\rm_cool}}{dt}\big)^{-1},$$
$u$ is the internal_energy and $du_{\rm cool}/dt$ is_the total cooling rate, then for_fragmentation we_require $\beta_\lesssim 3$, for a ratio of specific heats $\gamma = 2$_(in two dimensions). [-@Rice_beta_condition] carried out_three-dimensional simulations using a_smoothed particle_hydrodynamics_(SPH) code and_showed_that this_cooling parameter is dependent on the equation_of state_so that fragmentation can occur for_discs with $\gamma =_5/3$_and $7/5$ if $\beta \lesssim 7$_and $13$, respectively. [-@Libby_MSci] showed that_the critical value of $\beta$_(below_which_fragmentation will occur if the_stability |
s$ and $\xi$. These roots are associated with the same $m,s$, and $\xi$ but with different $\ell$. In order to find the root for a given $\ell$, it is useful to choose $n=n_{\ell}\equiv \ell-({\alpha}+{\beta})/2$ in Eq. (\[eq:determine\_elm\]) since in the limit $\xi\rightarrow 0$ all the terms in Eq. (\[eq:determine\_elm\]) become $O(\xi^2)$. This means that the choice naturally gives the leading term of a series expansion of the eigenvalue ${}_{s}E_{\ell m}(\xi)$ in terms of $\xi$ as $\ell(\ell+1)$.
When $\mid \xi\mid$ is not large, we can obtain the analytic expression of $_{s}E_{\ell m}(\xi)$ in a series of $\xi$ as $$\begin{aligned}
_{s}E_{\ell m}(\xi) = \ell (\ell +1) -\frac{2 s^2 m}{\ell (\ell +1)} \xi
+ \left[H(\ell+1)-H(\ell)-1\right]\xi^2 +O(\xi^3),\end{aligned}$$ where $$\begin{aligned}
H(\ell)=\frac{2(\ell^2-m^2)(\ell^2-s^2)^2}{(2\ell-1)\ell^3(2\ell+1)}.\end{aligned}$$ For the numerical calculation to determine $_{s}E_{\ell m}(\xi)$, one can use the analytic expression of $_{s}E_{\ell m}(\xi)$ above as an initial value to find the root in Eq. (\[eq:determine\_elm\]).
Once we obtain the eigenvalue, using Eqs. (\[eq:RncontElm\]) and (\[eq:LncontElm\]) we can compute all the coefficients ${}_{s}A_{\ell m}^{(n)}(\xi)$ from ${}_{s}A_{\ell m}^{(\tilde{n})}(\xi)$ for a given $\tilde{n}$. The coefficient for $n=n_\ell=l-(\alpha+\beta)/2$ is usually the largest term. The ratio of other terms to the dominant term, i.e. ${}_{s}A_{\ell m}^{(n)}(\xi)/{}_{s}A_{\ell m}^{( | s$ and $ \xi$. These roots are associated with the same $ m, s$, and $ \xi$ but with different $ \ell$. In decree to receive the root for a give $ \ell$, it is utilitarian to choose $ n = n_{\ell}\equiv \ell-({\alpha}+{\beta})/2 $ in Eq. (\[eq: determine\_elm\ ]) since in the limit $ \xi\rightarrow 0 $ all the terms in Eq. (\[eq: determine\_elm\ ]) become $ O(\xi^2)$. This mean that the choice naturally gives the contribute term of a series expansion of the eigenvalue $ { } _ { s}E_{\ell m}(\xi)$ in term of $ \xi$ as $ \ell(\ell+1)$.
When $ \mid \xi\mid$ is not large, we can obtain the analytic expression of $ _ { s}E_{\ell m}(\xi)$ in a series of $ \xi$ as $ $ \begin{aligned }
_ { s}E_{\ell m}(\xi) = \ell (\ell +1) -\frac{2 s^2 m}{\ell (\ell +1) } \xi
+ \left[H(\ell+1)-H(\ell)-1\right]\xi^2 + O(\xi^3),\end{aligned}$$ where $ $ \begin{aligned }
H(\ell)=\frac{2(\ell^2 - m^2)(\ell^2 - s^2)^2}{(2\ell-1)\ell^3(2\ell+1)}.\end{aligned}$$ For the numerical calculation to determine $ _ { s}E_{\ell m}(\xi)$, one can practice the analytic expression of $ _ { s}E_{\ell m}(\xi)$ above as an initial value to find the root in Eq. (\[eq: determine\_elm\ ]).
Once we prevail the eigenvalue, using Eqs. (\[eq: RncontElm\ ]) and (\[eq: LncontElm\ ]) we can compute all the coefficient $ { } _ { s}A_{\ell m}^{(n)}(\xi)$ from $ { } _ { s}A_{\ell m}^{(\tilde{n})}(\xi)$ for a given $ \tilde{n}$. The coefficient for $ north = n_\ell = l-(\alpha+\beta)/2 $ is usually the big term. The ratio of other price to the dominant term, i.e. $ { } _ { s}A_{\ell m}^{(n)}(\xi)/{}_{s}A_{\ell m}^ { ( | s$ ajd $\xi$. These roots are asrociated with tkw same $m,s$, ans $\xi$ but with different $\ell$. In order ti fine the root for a given $\ell$, it id useful to rhoose $n=n_{\ell}\equit \ell-({\alpmc}+{\beta})/2$ ln Eq. (\[zq:vetermine\_elm\]) sikce in the nimit $\xi\rightasruw 0$ all the terms in Eq. (\[eq:determine\_elm\]) become $O(\di^2)$. This means jhat uhe chojbe naturally gives the leading ferm of a series expsnsion of the eigenvalue ${}_{s}F_{\ell m}(\xi)$ in terms of $\xl$ as $\ell(\ell+1)$.
Qhen $\nid \xi\mid$ is not large, we can obtajn the analytic expression of $_{s}D_{\ell k}(\xi)$ in a sgxues mf $\xi$ as $$\bejin{alidned}
_{s}E_{\ell m}(\xl) = \ell (\all +1) -\frsc{2 s^2 m}{\ell (\ell +1)} \xi
+ \left[H(\ell+1)-H(\ell)-1\right]\xm^2 +O(\xi^3),\end{aligned}$$ wherg $$\begin{alicnzd}
H(\ell)=\frac{2(\ell^2-m^2)(\ell^2-s^2)^2}{(2\ell-1)\wlo^3(2\ell+1)}.\etd{alhgnea}$$ Fof tge nhmericwl ralculation to determibe $_{s}E_{\ell m}(\xi)$, one cam twe the analytjc expwefsion of $_{s}E_{\ell m}(\xi)$ above as an initial nalus to find the root in Ew. (\[eq:determine\_elm\]).
Once wg obtain tre eigenvalue, using Eqs. (\[eq:RncontElm\]) and (\[eq:LncontElk\]) we rav cimiute qlp the coefficients ${}_{s}A_{\ell m}^{(n)}(\xi)$ from ${}_{s}A_{\ell m}^{(\tilqs{n})}(\ci)$ for a given $\tllde{n}$. The coefficirnh gjr $n=n_\ell=l-(\alphx+\beta)/2$ na hsually the largesh term. Jhe rario of otrer yerms to the dominant term, u.e. ${}_{s}A_{\ell m}^{(n)}(\xp)/{}_{s}A_{\eol m}^{( | s$ and $\xi$. These roots are associated same and $\xi$ with different $\ell$. root a given $\ell$, is useful to $n=n_{\ell}\equiv \ell-({\alpha}+{\beta})/2$ in Eq. (\[eq:determine\_elm\]) since the limit $\xi\rightarrow 0$ all the terms in Eq. (\[eq:determine\_elm\]) become $O(\xi^2)$. This that the choice naturally gives the leading term of a series expansion of eigenvalue m}(\xi)$ terms $\xi$ as $\ell(\ell+1)$. When $\mid \xi\mid$ is not large, we can obtain the analytic expression of m}(\xi)$ in a series of $\xi$ as $$\begin{aligned} m}(\xi) = \ell (\ell -\frac{2 s^2 m}{\ell (\ell +1)} + +O(\xi^3),\end{aligned}$$ where H(\ell)=\frac{2(\ell^2-m^2)(\ell^2-s^2)^2}{(2\ell-1)\ell^3(2\ell+1)}.\end{aligned}$$ the calculation to determine m}(\xi)$, one can use the analytic expression of $_{s}E_{\ell m}(\xi)$ above as an initial value to find root in Once we the using (\[eq:RncontElm\]) and (\[eq:LncontElm\]) compute all the coefficients ${}_{s}A_{\ell m}^{(n)}(\xi)$ for a given $\tilde{n}$. The coefficient for $n=n_\ell=l-(\alpha+\beta)/2$ usually the term. The ratio of other terms the dominant term, i.e. ${}_{s}A_{\ell m}^{(n)}(\xi)/{}_{s}A_{\ell m}^{( | s$ and $\xi$. These roots are associAted with thE same $M,s$, aNd $\xI$ bUt wiTh diFferent $\ell$. In orDEr to Find the root for a given $\elL$, it is UsEFul tO ChOose $n=N_{\ell}\equIV \eLL-({\AlpHa}+{\BeTa})/2$ iN EQ. (\[Eq:DeterMinE\_elm\]) sinCe in the limIt $\xI\rIghtarrow 0$ all THe Terms in Eq. (\[eQ:deTermine\_elm\]) beComE $O(\xi^2)$. ThIs MeaNS that The ChoicE naturALly givEs the leadInG Term of A Series eXPAnSion Of the eigenvalue ${}_{s}E_{\ELl M}(\Xi)$ in terms of $\xi$ aS $\ell(\elL+1)$.
WHEn $\MID \xi\Mid$ Is not large, We Can obTAin the aNAlYTIC exPRession of $_{s}E_{\elL m}(\xi)$ in a seriES of $\Xi$ as $$\beGiN{alIGned}
_{s}E_{\Ell m}(\xI) = \eLL (\elL +1) -\frac{2 s^2 m}{\ell (\eLl +1)} \xi
+ \Left[H(\ell+1)-H(\Ell)-1\rigHT]\xi^2 +O(\xi^3),\eND{aligneD}$$ where $$\BegIn{aLignED}
H(\ElL)=\frAc{2(\ELl^2-m^2)(\ELl^2-S^2)^2}{(2\elL-1)\Ell^3(2\Ell+1)}.\end{alIgNeD}$$ For tHe nuMERICal cAlcUlatIon to Determine $_{s}E_{\elL m}(\xI)$, one CAn uSe the AnalyTic eXpRessiOn of $_{s}E_{\Ell m}(\xI)$ aBove as an initial ValuE to find thE roOt In EQ. (\[eQ:deteRMine\_elM\]).
OnCe wE obtain The eigeNValUe, USINg eqs. (\[eq:RncontElm\]) and (\[eQ:LNCOnTElm\]) we caN compuTE aLl THe coeffiCiEntS ${}_{s}A_{\eLL M}^{(n)}(\xi)$ fRom ${}_{s}a_{\ElL m}^{(\tilde{n})}(\Xi)$ for a GIvEn $\Tilde{n}$. THe CoeffiCiEnt For $N=n_\ell=L-(\AlphA+\beta)/2$ iS usually The laRGest term. The ratIO of other terms TO tHE DoMInanT teRm, i.e. ${}_{s}A_{\ell m}^{(n)}(\Xi)/{}_{s}A_{\ELl m}^{( | s$ and $\xi$. These rootsare associ atedwit h t he sam e $m ,s$, and $\xi$ butwith different $\ell$. In o rd e r to fi nd th e rootf or a gi ve n$\e ll $ ,it is us eful to choose $n =n_ {\ ell}\equiv \ e ll -({\alpha} +{\ beta})/2$ in Eq . (\[e q: det e rmine \_e lm\]) since in the limit $\ xi \ righta r row 0$a l ltheterms in Eq. (\[e q :d e termine\_elm\] ) beco me $O ( \ xi^ 2)$ . This mea ns that the cho i ce n a tur a lly gives the leading te r m o f a se ri ese xpansi on of t h e e igenvalue $ {}_{ s}E_{\ell m}(\x i )$ in t e rms of$\xi$as$\e ll(\ e ll +1 )$.
W hen $\ mid \xi \mid$ is n ot larg e, w e c a n ob tai n th e ana lytic express ion of$ _{s }E_{\ ell m }(\x i) $ ina seri es of $ \xi$ as $$\begi n{al igned}
_{ s}E _{ \el lm}(\x i ) = \e ll(\e ll +1)-\frac{ 2 s^ 2m } { \e ll (\ell +1)} \xi
+ \left[H (\ell+ 1 )- H( \ ell)-1\r ig ht] \xi^ 2 +O(\x i^3) , \e nd{align ed}$$w he re $$\beg in {align ed }
H (\e ll)=\ f rac{ 2(\ell ^2-m^2)( \ell^ 2 -s^2)^2}{(2\el l -1)\ell^3(2\e l l+ 1 ) }. \ end{ ali gned}$$ For the nume rica l c alc u latio n tode t er m ine $_{s}E_{\ell m} (\ xi)$,one c an use the an alytic exp r e s sion of$_{s } E_ { \ell m}(\xi)$above as an ini t ial valu e tofind the root inE q . (\[eq: det erm ine \_e l m \] ).
Once we o b t ainth e eigen val ue, usi ngEqs . ( \[e q: RncontElm \]) and(\ [e q: Ln con tElm\ ] ) we can c omp ut e a ll th e coeff icien ts $ {} _{ s }A_ {\ell m } ^{ ( n )}(\ xi )$ fro m $ {} _{s}A _{\e l l m }^{(\ti lde{n})}( \xi ) $ fo ragiven $ \tilde{n}$. T he coefficie nt fo r $n=n _ \ ell=l-(\ alpha+\beta)/2$ is usua l ly thelar gestterm . The rat ioof oth ert erms t o thedomin an t t e r m, i. e . $ {}_ {s }A_{\ell m } ^ {(n )}(\x i) /{}_ {s}A_{\ ell m}^{( | s$ and_$\xi$. These_roots are associated with_the same_$m,s$,_and $\xi$_but_with different $\ell$._In order to_find the root for_a given $\ell$,_it_is useful to choose $n=n_{\ell}\equiv \ell-({\alpha}+{\beta})/2$ in Eq. (\[eq:determine\_elm\]) since in the limit $\xi\rightarrow 0$_all_the terms_in_Eq. (\[eq:determine\_elm\])_become $O(\xi^2)$. This means that_the choice naturally gives the_leading term_of a series expansion of the eigenvalue ${}_{s}E_{\ell_m}(\xi)$_in terms of_$\xi$ as $\ell(\ell+1)$.
When $\mid \xi\mid$ is not large, we_can obtain the analytic expression of_$_{s}E_{\ell m}(\xi)$ in_a_series_of $\xi$ as $$\begin{aligned}
_{s}E_{\ell_m}(\xi) = \ell (\ell +1) -\frac{2_s^2 m}{\ell (\ell +1)} \xi
_ + \left[H(\ell+1)-H(\ell)-1\right]\xi^2 +O(\xi^3),\end{aligned}$$ where_$$\begin{aligned}
H(\ell)=\frac{2(\ell^2-m^2)(\ell^2-s^2)^2}{(2\ell-1)\ell^3(2\ell+1)}.\end{aligned}$$ For the numerical calculation to_determine $_{s}E_{\ell m}(\xi)$, one can_use the_analytic expression of $_{s}E_{\ell m}(\xi)$_above as an_initial value_to find the_root in Eq. (\[eq:determine\_elm\]).
Once we obtain the_eigenvalue, using Eqs. (\[eq:RncontElm\])_and (\[eq:LncontElm\]) we can compute all_the_coefficients ${}_{s}A_{\ell m}^{(n)}(\xi)$_from_${}_{s}A_{\ell_m}^{(\tilde{n})}(\xi)$ for_a given $\tilde{n}$._The_coefficient for_$n=n_\ell=l-(\alpha+\beta)/2$_is usually the largest term. The_ratio_of other terms to the dominant term,_i.e. ${}_{s}A_{\ell m}^{(n)}(\xi)/{}_{s}A_{\ell m}^{( |
18.39 1315 9.38 0.010 $1.7 \times 10^3 $ 4.9
SDSS J114749.70+163106.7 11 47 49.7 +16 31 06.8 0.554 18.78 1449 8.22 0.005 $3.5 \times 10^5 $ 0.1
SDSS J114857.33+160023.1 11 48 57.4 +16 00 22.7 1.224 18.14 1851 9.90 0.017 $4.1 \times 10^2 $ 37.6
SDSS J115141.81+142156.6 11 51 41.8 +14 21 57.0 1.002 18.11 1492 9.11 0.009 $6.3 \times 10^3 $ 1.8
SDSS J115346.39+241829.4 11 53 46.4 +24 18 29.8 0.753 18.41 1666 8.97 0.009 $2.1 \times 10^4 $ 1.3
SDSS J121018.34+015405.9 12 10 18.3 +01 54 06.2 0.216 16.90 1612 8.54 0.008 $2.6 \times 10^5 $ 0.6
SDSS J121018.66+185726.0 12 10 18.7 +18 57 27.0 1.516 17.42 1754 9.53 0.011 $1.1 \times 10^3 $ 8.4
SDSS J121056.83+231912.5 12 10 56.8 +23 19 13.0 1.260 18.47 1785 8.78 0.007 $2.7 \times 10^4 $ 0.5
-------------------------- ------------ --------------- ------- -------------- -------- ----------------------------------------------- ------- ----------------------- -----------------
-------------------------- ------------ ------------- ------- -------------- -------- ----------------------------------------------- ------- -------------------- -----------------
Id RA Dec $z$ $V_{ | 18.39 1315 9.38 0.010 $ 1.7 \times 10 ^ 3 $ 4.9
SDSS J114749.70 + 163106.7 11 47 49.7 +16 31 06.8 0.554 18.78 1449 8.22 0.005 $ 3.5 \times 10 ^ 5 $ 0.1
SDSS J114857.33 + 160023.1 11 48 57.4 +16 00 22.7 1.224 18.14 1851 9.90 0.017 $ 4.1 \times 10 ^ 2 $ 37.6
SDSS J115141.81 + 142156.6 11 51 41.8 +14 21 57.0 1.002 18.11 1492 9.11 0.009 $ 6.3 \times 10 ^ 3 $ 1.8
SDSS J115346.39 + 241829.4 11 53 46.4 +24 18 29.8 0.753 18.41 1666 8.97 0.009 $ 2.1 \times 10 ^ 4 $ 1.3
SDSS J121018.34 + 015405.9 12 10 18.3 +01 54 06.2 0.216 16.90 1612 8.54 0.008 $ 2.6 \times 10 ^ 5 $ 0.6
SDSS J121018.66 + 185726.0 12 10 18.7 +18 57 27.0 1.516 17.42 1754 9.53 0.011 $ 1.1 \times 10 ^ 3 $ 8.4
SDSS J121056.83 + 231912.5 12 10 56.8 +23 19 13.0 1.260 18.47 1785 8.78 0.007 $ 2.7 \times 10 ^ 4 $ 0.5
-------------------------- ------------ --------------- ------- -------------- -------- ----------------------------------------------- ------- ----------------------- -----------------
-------------------------- ------------ ------------- ------- -------------- -------- ----------------------------------------------- ------- -------------------- -----------------
I d RA Dec $ z$ $ V _ { | 18.39 1315 9.38 0.010 $1.7 \gimes 10^3 $ 4.9
SDSS J114749.70+163106.7 11 47 49.7 +16 31 06.8 0.554 18.78 1449 8.22 0.005 $3.5 \tjles 10^5 $ 0.1
SDSS J114857.33+160023.1 11 48 57.4 +16 00 22.7 1.224 18.14 1851 9.90 0.017 $4.1 \times 10^2 $ 37.6
SDSS J115141.81+142156.6 11 51 41.8 +14 21 57.0 1.002 18.11 1492 9.11 0.009 $6.3 \times 10^3 $ 1.8
SDSS J115346.39+241829.4 11 53 46.4 +24 18 29.8 0.753 18.41 1666 8.97 0.009 $2.1 \times 10^4 $ 1.3
SDSS J121018.34+015405.9 12 10 18.3 +01 54 06.2 0.216 16.90 1612 8.54 0.008 $2.6 \times 10^5 $ 0.6
SDSS J121018.66+185726.0 12 10 18.7 +18 57 27.0 1.516 17.42 1754 9.53 0.011 $1.1 \times 10^3 $ 8.4
XDFW J121056.83+231912.5 12 10 56.8 +23 19 13.0 1.260 18.47 1785 8.78 0.007 $2.7 \tines 10^4 $ 0.5
-------------------------- ------------ --------------- ------- -------------- -------- ----------------------------------------------- ------- ----------------------- -----------------
-------------------------- ------------ ------------- ------- -------------- -------- ----------------------------------------------- ------- -------------------- -----------------
Id RA Dec $z$ $V_{ | 18.39 1315 9.38 0.010 $1.7 \times 10^3 SDSS 11 47 +16 31 06.8 $3.5 10^5 $ 0.1 J114857.33+160023.1 11 48 +16 00 22.7 1.224 18.14 1851 0.017 $4.1 \times 10^2 $ 37.6 SDSS J115141.81+142156.6 11 51 41.8 +14 21 1.002 18.11 1492 9.11 0.009 $6.3 \times 10^3 $ 1.8 SDSS J115346.39+241829.4 11 46.4 18 0.753 1666 8.97 0.009 $2.1 \times 10^4 $ 1.3 SDSS J121018.34+015405.9 12 10 18.3 +01 54 06.2 16.90 1612 8.54 0.008 $2.6 \times 10^5 $ SDSS J121018.66+185726.0 12 10 +18 57 27.0 1.516 17.42 9.53 $1.1 \times $ SDSS 12 10 56.8 19 13.0 1.260 18.47 1785 8.78 0.007 $2.7 \times 10^4 $ 0.5 -------------------------- ------------ --------------- ------- -------------- ----------------------------------------------- ------- -------------------------- ------------ ------- -------- ------- -------------------- ----------------- Dec $z$ $V_{ | 18.39 1315 9.38 0.010 $1.7 \times 10^3 $ 4.9
SDSS J114749.70+163106.7 11 47 49.7 +16 31 06.8 0.554 18.78 1449 8.22 0.005 $3.5 \times 10^5 $ 0.1
SDSS J114857.33+160023.1 11 48 57.4 +16 00 22.7 1.224 18.14 1851 9.90 0.017 $4.1 \times 10^2 $ 37.6
sDSS J115141.81+142156.6 11 51 41.8 +14 21 57.0 1.002 18.11 1492 9.11 0.009 $6.3 \times 10^3 $ 1.8
sDSS J115346.39+241829.4 11 53 46.4 +24 18 29.8 0.753 18.41 1666 8.97 0.009 $2.1 \TimEs 10^4 $ 1.3
SdSs J121018.34+015405.9 12 10 18.3 +01 54 06.2 0.216 16.90 1612 8.54 0.008 $2.6 \tiMes 10^5 $ 0.6
SdSS J121018.66+185726.0 12 10 18.7 +18 57 27.0 1.516 17.42 1754 9.53 0.011 $1.1 \times 10^3 $ 8.4
SDSS J121056.83+231912.5 12 10 56.8 +23 19 13.0 1.260 18.47 1785 8.78 0.007 $2.7 \TImes 10^4 $ 0.5
-------------------------- ------------ --------------- ------- -------------- -------- ----------------------------------------------- ------- ----------------------- -----------------
-------------------------- ------------ ------------- ------- -------------- -------- ----------------------------------------------- ------- -------------------- -----------------
id RA Dec $z$ $V_{ | 18.39 1315 9. 38 0.01 0 $1.7 \times 10^3 $ 4 . 9
S DS S J11 4749.70 + 16 3 1 06. 7 1147 49 .7 +16 31 06. 8 0.55 4 1 8.78 1 449 8.22 0.0 0 5 $3 .5 \times 1 0 ^5 $ 0.1SD SS J 114857.33+160023. 1 11 48 57.4 + 16 0022 . 7 1 .22 4 18.14 1851 9.90 0.017 $4 .1 \ti mes 10^2 $ 37.6
S DSS J1 1 5141.81 + 142156. 6 11 51 41 .8 +1 42157 . 0 1.0 0 2 18.11 1 492 9.11 0. 009 $6.3\times 10^3 $ 1.8
SDS S J1 15346.39+ 241 82 9.4 11 5 3 46.4 + 2418 29.8 0. 7 53 1 8 . 41 1666 8 .9 7 0 .009 $2.1 \ti mes 10 ^ 4$ 1. 3 SDSS J 121 018 .34+0 1 5405 .9 1 2 10 18. 3 + 0 1 54 06.2 0 .216 16.90 161 2 8 .5 4 0.008 $2.6 \t im es 10^ 5 $ 0.6
SDS S J121018. 6 6 + 185726.0 1 2 1 0 18.7 +18 57 27.0 1.516 17.42 1754 9. 53 0 .0 11 $1 .1\times10^ 3 $ 8. 4
SDSSJ121056. 83 +2 31 91 2.5 12 10 56.8 +23 1 9 1 3.0 1.26 0 1 8.47 1785 8. 78 0. 0 07 $ 2. 7 \time s 10^4 $ 0 .5
----- -- --- ------ - - -------- ------------ --------- - ----- - --- --- - ---- --------- -- ------ -- - ------ ------ ----- -- --- - - ----- - - -- --- -- ------ --- - - ------- -- ---- ------- ----- ------------ - --- -
-------- --- ---- - - -- --- - -- - -- -- - --- - - -- ------------ - --------- - -- ---------- --- -- -- ---- ------- ----- - ------- --------- --------- -- ---- - --- -- ------- -------- ----- --- - ----- - -- -----
Id RA Dec $z$ $V_{ | 18.39 _ _ _ _1315_ __ _ _ _ __ 9.38 __ ___ _ _ _0.010 $1.7 \times 10^3 $ __ _4.9
SDSS J114749.70+163106.7 11 47 49.7_ +16 31 06.8 _ _0.554__ 18.78 _ _ 1449 _ _ _ _ _ 8.22 _ _ _ _ _ _ 0.005 $3.5_\times_10^5 $ ___ _0.1
SDSS_J114857.33+160023.1_ _11_48 57.4 +16 00_22.7_ 1.224 _18.14 __ _ 1851 _ _ _ _ _ 9.90 _ _ __ __ _ 0.017 _ $4.1_\times 10^2 $ _ 37.6
_SDSS_J115141.81+142156.6 11 51 41.8_ +14 21 57.0 _ 1.002 __18.11_ _ 1492 _ _ __ __ _9.11__ _ _ _ 0.009 _ $6.3 \times 10^3 $___ 1.8
SDSS J115346.39+241829.4 11_53 46.4_ +24_18 29.8 0.753 18.41_ _ 1666 _ _ _ 8.97 __ ____ _ _ _ 0.009_ $2.1_\times 10^4_$ __ 1.3
SDSS J121018.34+015405.9 12 10 18.3 +01 54 06.2 _ 0.216 16.90_ __ _1612 _ _ __ _ 8.54 _ _ _ _ _ _ _ __0.008 $2.6 \times 10^5 $ ____0.6
SDSS_J121018.66+185726.0 _12 10 18.7_ +18_57_27.0 1.516_ 17.42_ _ 1754 _ _ _ 9.53 _ __ _ _ __ 0.011 $1.1 \times 10^3 $ __8.4
_ SDSS J121056.83+231912.5__ 12 10 56.8 _+23_19_13.0__ 1.260 _18.47 __ 1785 _ __ _ _ __8.78 ____ _ 0.007__ $2.7 \times 10^4 $_ 0.5
__-------------------------- ------------ --------------- -------_-------------- -------- ----------------------------------------------- ------- ----------------------- -----------------
--------------------------_------------ ------------- ------- -------------- --------_----------------------------------------------- ------- -------------------- -----------------
_ Id _ _RA _ __ Dec_ _ $z$ _ $V_{ |
a value of $\approx 44$. Self-gravity is therefore not expected to lead to instabilities. As in M09, we assume that the initial mass of the disc is 1% of the mass of the star and we set the outer radius of the disc to 100 AU and the inner radius to 10 AU. Realistically, the disc should probably extend inwards towards the radius of the star, but it is computationally very expensive to simulate the disc on the orbital time scales at these small radii. During the simulation, particles in the disc will migrate inwards due to viscous evolution of the disc and are accreted onto the star, resulting in a disc that extends further inwards. This set-up allows us to postpone the expensive calculations towards later times in our simulations, thereby significantly decreasing the duration of our simulations.
We position the disc perpendicular to the flow (see Fig. \[fig:schematics\_eda1\]). This perfect alignment may not occur frequently in nature, but allows to discern the relevant physical processes more clearly. In future work we will investigate the influence of giving the disc an inclination with respect to the flow direction.
### Star {#sec:sink_eda1}
We assume that the mass of the star is concentrated in a single point, which has a mass of $0.4\,{\mbox{$\mathrm{M}_{\odot}$}}$. This corresponds to the typical mass expected in the early disc accretion scenario. The point mass is added as a collisionless particle to the SPH code and we treat it as a sink particle, i.e. it can accrete gas particles that fall within a certain (fixed) radius. The sink particle absorbs the mass and momentum carried by the gas particles it accretes.
We assume the radius of the sink particle is 5% of the Bondi-Hoyle accretion radius, $R_{\rm A}$, defined as [see also @Bondi52; @shima85]:
$$\label{eq:ra_eda1}
R_{\rm A} = \frac{2GM_{*}}{c_s^2 +v_{\rm ISM}^2}\\$$
in which $M_{*}$ is the mass of the star, $c_s$ is the sound speed and $v_{\rm ISM}$ the velocity of the ISM (with respect to the star). Using the values mentioned above, gives us an accretion radius of 1.8 AU, which is significantly smaller than in M09 | a value of $ \approx 44$. Self - gravity is therefore not expected to lead to imbalance. As in M09, we wear that the initial mass of the disc is 1% of the mass of the headliner and we set the outer spoke of the disc to 100 AU and the inner radius to 10 AU. Realistically, the phonograph record should probably extend inwards towards the spoke of the star, but it is computationally very expensive to simulate the disc on the orbital time scale at these small radii. During the simulation, particles in the disc will migrate inwards due to syrupy evolution of the disc and are accreted onto the star topology, resulting in a disc that extends further inwards. This set - up let us to postpone the expensive calculations towards later times in our simulations, thereby significantly decrease the duration of our simulations.
We position the disc perpendicular to the flow (see Fig. \[fig: schematics\_eda1\ ]). This perfect alignment may not occur frequently in nature, but allows to discern the relevant physical processes more clearly. In future work we will investigate the influence of giving the disc an dip with respect to the flow steering.
# # # Star { # sec: sink_eda1 }
We wear that the mass of the star is concentrated in a single point, which has a mass of $ 0.4\,{\mbox{$\mathrm{M}_{\odot}$}}$. This correspond to the typical mass expected in the early disc accretion scenario. The point mass is add as a collisionless particle to the SPH code and we treat it as a sink particle, i.e. it can accrete accelerator particles that hang within a certain (fixed) radius. The sink particle steep the mass and momentum carried by the gas particle it accretes.
We assume the radius of the sink atom is 5% of the Bondi - Hoyle accretion radius, $ R_{\rm A}$, defined as [ see also @Bondi52; @shima85 ]:
$ $ \label{eq: ra_eda1 }
R_{\rm A } = \frac{2GM_{*}}{c_s^2 + v_{\rm ISM}^2}\\$$
in which $ M_{*}$ is the mass of the ace, $ c_s$ is the sound focal ratio and $ v_{\rm ISM}$ the velocity of the ISM (with deference to the star). use the values mentioned above, give us an accretion radius of 1.8 AU, which is importantly smaller than in M09 | a galue of $\approx 44$. Self-grayity is thereforg bot ex'ected fo lead go instabilities. As in M09, we essune thqt the initial mass of the disc is 1% of rhe nqss of the star and we ssb the iuter radius on the disc do 100 AU and the ivnzr radius to 10 AU. Realistically, the dysc shoilf probably extgnd imrarda towards the radius of the star, bht it iv computationslly very expensive to simklatf the disc on the lrbital timg scwoes at these small radpn. During thg simulation, particles in the dirc wikl migrate ibwaggs due to vmscous evolution on the dhsc and are accreted pntm tye star, resulting in e disc that extends surther itwcrds. This set-up alloww ys to posdpond thd espxnsjve capcumations tosards later times in our simulstyins, thereby sjgnifisagtly decreasing the duration of our simllatjons.
We position the disx perpendicular to thg flow (see Fig. \[fig:schematics\_eda1\]). This perfect alignment may tot orcjr yvcquevrlj in nature, but allows to discern the relevane pnyxical processef more cleatlj. Og future work we will jnvestigate the invluence of guving the disv an inclination with respext to the fljq direction.
### Star {#szc:sink_eda1}
We cssume that the mass of the star ir cohcentrated ln a singmd point, which har a mdss of $0.4\,{\mbox{$\mathrm{M}_{\odot}$}}$. Thif correspinds to the typocal mwss expectfd in the early disc accgetiou scetario. The ooint mass is added as a collismpnless partivla tm the SPK code and we treat it as a sink karticle, n.e. it zan accretv gas parvicles that sall within a fertain (fixev) radius. Ehe wink particud absorbs the kass and momentum cqrried by the gas iartizmes it accretes.
Ct awsume the radiux ow tre smnk pwsticle is 5% ox thd Bumdi-Hohle accretiiv racius, $R_{\rm A}$, defined av [ses also @Bondi52; @shima85]:
$$\kanel{eq:ra_edq1}
R_{\rm A} = \srac{2GM_{*}}{c_s^2 +v_{\rm OSM}^2}\\$$
in which $M_{*}$ is tje mavs uf thr sjar, $c_s$ is the sound speed and $v_{\dm ISM}$ thf vclocity of thq ISN (with respeet to the star). Using the values mentionev above, gives us an acceetion radius of 1.8 AO, wmich is signmficanely smallar than in M09 | a value of $\approx 44$. Self-gravity is expected lead to As in M09, mass the disc is of the mass the star and we set the radius of the disc to 100 AU and the inner radius to 10 Realistically, the disc should probably extend inwards towards the radius of the star, it computationally expensive simulate the disc on the orbital time scales at these small radii. During the simulation, particles the disc will migrate inwards due to viscous of the disc and accreted onto the star, resulting a that extends inwards. set-up us to postpone expensive calculations towards later times in our simulations, thereby significantly decreasing the duration of our simulations. We the disc the flow Fig. This alignment may not in nature, but allows to discern processes more clearly. In future work we will the influence giving the disc an inclination with to the flow direction. ### Star {#sec:sink_eda1} We that the mass of the star is concentrated in a single point, which has a $0.4\,{\mbox{$\mathrm{M}_{\odot}$}}$. This corresponds to typical mass expected the disc scenario. point mass added as a collisionless particle to the SPH code and we it as a sink particle, i.e. it can accrete gas fall a certain (fixed) The sink particle absorbs mass momentum carried by the it We of sink is 5% of the accretion radius, $R_{\rm A}$, defined [see also @Bondi52; @shima85]: +v_{\rm ISM}^2}\\$$ in which $M_{*}$ is the mass the star, $c_s$ is the sound speed $v_{\rm ISM}$ the velocity of the ISM (with respect to the star). the values gives us an accretion radius of 1.8 AU, is significantly smaller than M09 | a value of $\approx 44$. Self-gravity Is thereforE not eXpeCteD tO leaD to iNstabilities. As IN M09, we Assume that the initial maSs of tHe DIsc iS 1% Of The maSs of the STaR ANd wE sEt The OuTEr RadiuS of The disc To 100 AU and the InnEr Radius to 10 AU. ReALiStically, thE diSc should probAblY extenD iNwaRDs towArdS the rAdius oF The staR, but it is cOmPUtatioNAlly verY EXpEnsiVe to simulate the diSC oN The orbital time Scales At THeSE SmaLl rAdii. During ThE simuLAtion, paRTiCLES in THe disc will migRate inwards DUe tO viscoUs EvoLUtion oF the dIsC And Are accreted Onto The star, reSultinG In a disc THat exteNds furTheR inWardS. thIs Set-Up ALloWS uS to POstPone the eXpEnSive cAlcuLATIOns tOwaRds lAter tImes in our simuLatIons, THerEby siGnifiCantLy DecreAsing tHe durAtIon of our simulatIons.
we positioN thE dIsc PeRpendICular tO thE flOw (see FiG. \[fig:schEMatIcS\_EDA1\]). THis perfect alignmenT mAY NoT occur frEquentLY iN nATure, but aLlOws To diSCErn thE relEVaNt physicAl procESsEs More cleArLy. In fuTuRe wOrk We wilL InveStigatE the inflUence OF giving the disc AN inclination wITh RESpECt to The Flow directiOn.
### StAR {#sec:Sink_EDa1}
we aSSume tHat thE mASs OF the star is concentraTeD in a siNgle pOint, which has a Mass of $0.4\,{\mbox{$\MATHrm{M}_{\odot}$}}$. this COrREsponds to the tyPical Mass expectED in the eaRly diSc accretIon scenarIO. the point MasS is AddEd aS A CoLlisionless paRTIcle To The SPH cOde And we trEat It aS a sInk PaRticle, i.e. iT can accrEtE gAs PaRtiCles tHAt fall wiThIn a CeRtaIn (fixED) radiuS. The sInk pArTiCLe aBsorbs tHE mASS and MoMeNtum CarRiEd by tHe gaS ParTicles iT accretes.
we aSSume ThE rAdius of The sink particLe Is 5% of the BonDi-hoyLe accrETIon radiuS, $R_{\rm A}$, defined as [see also @BoNDi52; @shima85]:
$$\LabEl{eq:rA_eda1}
r_{\rm A} = \frac{2Gm_{*}}{c_s^2 +V_{\rm ISM}^2}\\$$
In wHIch $M_{*}$ is The masS of thE sTar, $C_S$ Is the SOUnD spEeD and $v_{\rm ISM}$ THE veLocitY oF the iSM (with Respect to the star). UsINg tHe values mentiOneD aboVE, GiVes US aN AccReTIon RADius of 1.8 AU, which is SignificanTlY SmAller than iN m09 | a value of $\approx 44$.Self-gravi ty is th ere fo re n ot e xpected to lea d toinstabilities. As in M 09, w ea ssum e t hat t he init i al m ass o fthe d i sc is 1 % o f the m ass of the st ar and we sett he outer rad ius of the disc to 100 A Uand the i nne r rad ius to 10 AU. Realisti ca l ly, th e disc s h o ul d pr obably extend inw a rd s towards the r adiusof th e sta r,but it isco mputa t ionally ve r y exp e nsive to simu late the di s c o n theor bit a l time scal es atthese small rad ii. Durin g thes imulati o n, part iclesinthe dis c w il l m ig r ate in war d s d ue to vi sc ou s evo luti o n o f th e d iscand a re accreted o nto the sta r, re sulti ng i na dis c that exte nd s further inwar ds.This set- upal low sus to postpo nethe expens ive cal c ula ti o n s t owards later times i n ou r simula tions, th er e by signi fi can tlyd e creas ingt he duratio n of o u rsi mulatio ns .
Wepo sit ion thed iscperpen dicularto th e flow (see Fig . \[fig:schema t ic s \ _e d a1\] ).This perfec t al i gnme nt m a ynot occur freq ue n tl y in nature, but all ow s to d iscer n the relevan t physical p r ocessesmore cl e arly. In futur e wor k we willi nvestiga te th e influe nce of gi v i ng the d isc an in cli n a ti on with respe c t toth e flowdir ection.
# ##Sta r { #s ec:sink_e da1}
We a ss um etha t the mass ofth e s ta r i s con c entrat ed in a s in gl e po int, wh i ch h as a m as s of $0 .4 \,{\m box{ $ \ma thrm{M} _{\odot}$ }}$ . Thi sco rrespon ds to the typ ic al mass ex pe cte d in t h e early d isc accretion scenario. The poi ntmassis a dded as a co llisio nle s s part icle t o the S PHc o de an d we tr ea t it as as i nkparti cl e, i .e. itcan accrete gas pa r tic les that fall wi thin a c ert a in (fi xe d ) r a d ius. The sink p article ab so r bs the massa ndmo mentumcarried by t h e gas p articlesit accret es .
W e ass ume the ra dius ofthe sinkp artic l eis 5% of the B on di- Hoyle accre t ion radi us, $R _{ \rm A} $, de fi ned as [ see also @Bondi52; @shi ma85]:
$$\ lab el{eq:ra_ eda 1 }
R _{\rm A}= \f rac{2GM_{* }}{ c_s ^2 +v _{\ r m ISM }^2} \ \$ $
i n whi ch $ M _{*}$ ist he ma s s o f the star, $ c _s$ is t hes ound s peed and $v_{\rm ISM} $ the velocityof t h e IS M ( w ithre spect to the s tar ). U sing the v alues menti oned abo ve , give s us a n accr etion r a d iu s of 1. 8 AU , w hich is s ign if i cantlysm al l er tha n in M 09 | a_value of_$\approx 44$. Self-gravity is_therefore not_expected_to lead_to_instabilities. As in_M09, we assume_that the initial mass_of the disc_is_1% of the mass of the star and we set the outer radius of_the_disc to_100_AU_and the inner radius to_10 AU. Realistically, the disc_should probably_extend inwards towards the radius of the star,_but_it is computationally_very expensive to simulate the disc on the orbital_time scales at these small radii._During the simulation,_particles_in_the disc will migrate_inwards due to viscous evolution of_the disc and are accreted onto_the star, resulting in a disc that_extends further inwards. This set-up allows_us to postpone the expensive_calculations towards_later times in our simulations,_thereby significantly decreasing_the duration_of our simulations.
We_position the disc perpendicular to the_flow (see Fig._\[fig:schematics\_eda1\]). This perfect alignment may not_occur_frequently in nature,_but_allows_to discern_the relevant physical_processes_more clearly._In_future work we will investigate the_influence_of giving the disc an inclination with_respect to the flow_direction.
###_Star {#sec:sink_eda1}
We assume that_the mass of the star_is concentrated in a single point,_which has_a mass_of $0.4\,{\mbox{$\mathrm{M}_{\odot}$}}$. This corresponds to the typical mass expected in the_early disc accretion scenario. The point_mass is added as_a collisionless_particle_to the SPH_code_and we_treat it as a sink particle, i.e._it can_accrete gas particles that fall within_a certain (fixed) radius._The_sink particle absorbs the mass and_momentum carried by the gas particles_it accretes.
We assume the radius_of_the_sink particle is 5% of_the Bondi-Hoyle accretion radius, $R_{\rm A}$,_defined as [see_also @Bondi52; @shima85]:
$$\label{eq:ra_eda1}
R_{\rm A} = \frac{2GM_{*}}{c_s^2 +v_{\rm_ISM}^2}\\$$
in_which $M_{*}$ is the mass of_the_star, $c_s$ is the sound speed_and_$v_{\rm_ISM}$ the velocity of the_ISM (with respect to the star)._Using the values mentioned above, gives us an accretion_radius of 1.8_AU, which is significantly smaller_than_in_M09 |
except that now the PDCZ convection retreats and the inner/upper borders are shifted outward/inward until $t=\tau_{\rm pdcz}$. The nuclear network is integrated over the pulse quenching phase and a final homogeneous chemical composition is obtained. This sets the chemical mixture of the material that may be brought up to the surface by the subsequent third dredge-up phase.
Despite its simplicity the PDCZ model yields results that nicely agree with those of full TP-AGB computations. A detailed discussion of the predictions and their main dependencies is given in Sect. \[ssec\_xpdcz\].
The synthetic module {#sec_synthmod}
====================
Most analytical ingredients of the `COLIBRI` code are formulas accurately fitting the results of full AGB models covering wide ranges of initial stellar mass and metallicity. The formulas are taken either from the extensive compilations by @Wagenhuber_96 [@WagenGroen_98; @Karakas_etal02; @Izzard_etal04; @Izzard_etal06], and other sources [@Straniero_etal03], or they are directly derived from AGB model data sets by using standard $\chi^2$-minimization techniques. New fits can be found in Appendix \[app\_fit\].
Importantly, all these analytic relations include a metallicity dependence, and take into account the peculiar behaviour of the first sub-luminous pulses while approaching the full-amplitude regime.
Among the most important prescriptions we mention the flash-driven luminosity variations as a function of the pulse-cycle phase [@WagenGroen_98], the core mass-interpulse period relation [@WagenGroen_98], the maximum mass of the PDCZ and its duration, the maximum temperature attained at the bottom of the PDCZ during a TP [@KarakasLattanzio_07][^7], the efficiency $\lambda$ of the third dredge-up [@Karakas_etal02].
Due to their particular relevance, below we will discuss in more detail a few analytic relations adopted in the present version of `COLIBRI`.
The third dredge-up: the need for a parametric description {#ssec_3dup}
----------------------------------------------------------
It is common practice describing the third dredge-up by means of two characteristic quantities, namely:
- $M_{\rm c}^{\rm min | except that now the PDCZ convection retreats and the inner / upper borders are transfer outward / in until $ t=\tau_{\rm pdcz}$. The nuclear network is integrated over the pulsation quenching phase and a concluding homogeneous chemical composition is obtained. This rig the chemical mixture of the material that may be fetch up to the surface by the subsequent third dredge - up phase.
Despite its simplicity the PDCZ mannequin yields results that nicely harmonize with those of wide TP - AGB computations. A detailed discussion of the predictions and their main dependencies is given in Sect. \[ssec\_xpdcz\ ].
The man-made module { # sec_synthmod }
= = = = = = = = = = = = = = = = = = = =
Most analytical ingredient of the 'COLIBRI 'code are formulas accurately fitting the results of entire AGB models covering wide ranges of initial stellar bulk and metallicity. The formulas are taken either from the extensive compilations by @Wagenhuber_96 [ @WagenGroen_98; @Karakas_etal02; @Izzard_etal04; @Izzard_etal06 ], and other sources [ @Straniero_etal03 ], or they are directly derived from AGB model data sets by using standard $ \chi^2$-minimization techniques. New fits can be found in Appendix \[app\_fit\ ].
Importantly, all these analytic relations admit a metallicity addiction, and take into account the peculiar behavior of the first sub - luminous pulses while border on the wide - amplitude regime.
Among the most authoritative prescriptions we mention the flash - driven luminosity variations as a function of the pulsation - cycle phase [ @WagenGroen_98 ], the core mass - interpulse period relation [ @WagenGroen_98 ], the maximum mass of the PDCZ and its duration, the maximum temperature achieve at the bottom of the PDCZ during a TP [ @KarakasLattanzio_07][^7 ], the efficiency $ \lambda$ of the third dredge - up [ @Karakas_etal02 ].
Due to their particular relevance, below we will discuss in more detail a few analytic relations espouse in the present version of 'COLIBRI `.
The third dredge - up: the need for a parametric description { # ssec_3dup }
----------------------------------------------------------
It is common exercise describing the third dredge - up by mean of two characteristic quantities, namely:
- $ M_{\rm c}^{\rm min | exfept that now the PDCZ cunvection retrecrs and the ihner/uppef borders are shifted outwarv/inwqrd ubtil $t=\tau_{\rm pdcz}$. The njclear nenwork is untejrated over the 'hlse qucuchinf phavx and a final hpmogeneous chemical compmsktnon is obtained. This sets the chemicwl mixtirf of the materyal uhae mag be brought up to the surface by fhe subvequent third dredge-up phase.
Despite its simolicity the PDCZ mldel yields restots that nicdly agree with those or full TP-AGB computations. A detxiled discussiob if hve predictiins agd their maik depengencies is given in Scct. \[ssxc\_xpecz\].
The synthetic modune {#sec_synthmod}
====================
Most analyticdl ingredients of tye `COLITRI` wode qre fodmnlaa accugatxly fitting the resultw of full AGB modelx sivering wide danges os initial stellar mass and metallicity. Uhe fkrmulas are taken eithee from the extensive fompilatijns by @Wagenhuber_96 [@WagenGroen_98; @Karakas_etal02; @Izzard_etdl04; @Izvafd_euao06], and itjer sources [@Straniero_etal03], or they are directlr dtrined from AGB modej data sets bj ifing standard $\chi^2$-miujmjzation techniques. New fijs can be found in Sppendix \[app\_fit\].
Importantly, aol these anajttic relations incpude a metaulicoty drpendence, and take into acckunt the pefuliar begxviour of the fifst sgb-luminous pulses while ap[roaching the full-amolitode regyme.
Among tje most important prescrlptious we mention tje flash-driven luminosity variavmons as a funvthon of the 'ulse-cncle phase [@WagegGroen_98], the corg mass-intzrpulsd period rvlation [@WegenGroen_98], thq maximum masv of the PDCZ and its durqtiob, the mxbimum temperatire attaiutd at the vottom of the PDCZ dutihg a TP [@KarakasOatuabzio_07][^7], the efficoenzy $\jalbva$ of dhe third dradge-jp [@Ysrakar_etal02].
Dut ti theor particular relevatce, gelow we will discisf in morg detail w few analytiv relations adoptef in vhe prxsent ferfion of `COLIBRI`.
The third dredgs-up: the nfed for a paramqtrig defcription {#xsec_3dup}
----------------------------------------------------------
It is common practice describing the third dredge-op by means of two chcrscteristic auantyties, namaly:
- $M_{\rm c}^{\rm min | except that now the PDCZ convection retreats inner/upper are shifted until $t=\tau_{\rm pdcz}$. over pulse quenching phase a final homogeneous composition is obtained. This sets the mixture of the material that may be brought up to the surface by subsequent third dredge-up phase. Despite its simplicity the PDCZ model yields results that agree those full computations. A detailed discussion of the predictions and their main dependencies is given in Sect. \[ssec\_xpdcz\]. synthetic module {#sec_synthmod} ==================== Most analytical ingredients of `COLIBRI` code are formulas fitting the results of full models wide ranges initial mass metallicity. The formulas taken either from the extensive compilations by @Wagenhuber_96 [@WagenGroen_98; @Karakas_etal02; @Izzard_etal04; @Izzard_etal06], and other sources [@Straniero_etal03], or are directly AGB model sets using $\chi^2$-minimization techniques. New be found in Appendix \[app\_fit\]. Importantly, relations include a metallicity dependence, and take into the peculiar of the first sub-luminous pulses while the full-amplitude regime. Among the most important prescriptions mention the flash-driven luminosity variations as a function of the pulse-cycle phase [@WagenGroen_98], the core relation [@WagenGroen_98], the maximum of the PDCZ its the temperature at the of the PDCZ during a TP [@KarakasLattanzio_07][^7], the efficiency $\lambda$ of third dredge-up [@Karakas_etal02]. Due to their particular relevance, below we in detail a few relations adopted in the version `COLIBRI`. The third dredge-up: for parametric It common describing the third dredge-up means of two characteristic quantities, - $M_{\rm c}^{\rm min | except that now the PDCZ conveCtion retreAts anD thE inNeR/uppEr boRders are shifteD OutwArd/inward until $t=\tau_{\rm pdCz}$. The NuCLear NEtWork iS integrATeD OVer ThE pUlsE qUEnChing PhaSe and a fInal homogeNeoUs Chemical compOSiTion is obtaIneD. This sets the CheMical mIxTurE Of the MatErial That maY Be brouGht up to thE sURface bY The subsEQUeNt thIrd dredge-up phase.
DESpITe its simplicitY the PDcZ MOdEL YieLds Results thaT nIcely AGree witH ThOSE Of fULl TP-AGB computAtions. A detaILed DiscusSiOn oF The preDictiOnS And Their main dePendEncies is gIven in sEct. \[ssec\_XPdcz\].
The SyntheTic ModUle {#sEC_sYnThmOd}
====================
mOst ANaLytICal IngredieNtS oF the `CoLIBri` CODe arE foRmulAs accUrately fittinG thE resULts Of fulL AGB mOdelS cOveriNg wide RangeS oF initial stellar Mass And metallIciTy. the FoRmulaS Are takEn eIthEr from tHe extenSIve CoMPILaTions by @Wagenhuber_96 [@WAgENgrOen_98; @KarakAs_etal02; @iZzArD_Etal04; @IzzaRd_EtaL06], and OTHer soUrceS [@stRaniero_eTal03], or tHEy ArE directLy DeriveD fRom aGB Model DAta sEts by uSing stanDard $\cHI^2$-minimization tEChniques. New fiTS cAN Be FOund In APpendix \[app\_fIt\].
ImPOrtaNtly, ALl TheSE analYtic rElATiONs include a metalliciTy DependEnce, aNd take into accOunt the pecULIAr behaviOur oF ThE First sub-luminoUs pulSes while apPRoaching The fuLl-amplitUde regime.
aMOng the moSt iMpoRtaNt pREScRiptions we menTIOn thE fLash-driVen LuminosIty VarIatIonS aS a functioN of the puLsE-cYcLe PhaSe [@WagENGroen_98], thE cOre MaSs-iNterpULse perIod reLatiOn [@waGEnGRoen_98], the MAxIMUm maSs Of The PdCZ AnD its dUratIOn, tHe maximUm temperaTurE AttaInEd At the boTtom of the PDCZ DuRing a TP [@KarAkAsLAttanzIO_07][^7], The efficIency $\lambda$ of the third drEDge-up [@KaRakAs_etaL02].
Due To their paRtiCular rEleVAnce, beLow we wIll diScUss IN More dETAiL a fEw Analytic reLATioNs adoPtEd in The presEnt version of `COLIBRi`.
the Third dredge-up: The Need FOR a ParAMeTRic DeSCriPTIon {#ssec_3dup}
----------------------------------------------------------
It is cOmmon practIcE DeScribing thE ThiRd Dredge-uP by meanS of twO CharactEristic quAntities, nAmEly:
- $M_{\RM C}^{\rm Min | except that now the PDCZconvection retr eat s a nd the inn er/upper borde r s ar e shifted outward/inwa rd un ti l $t= \ ta u_{\r m pdcz} $ .T h e n uc le arne t wo rk is in tegrate d over the pu ls e quenchingp ha se and a f ina l homogeneou s c hemica lcom p ositi onis ob tained . Thissets thech e micalm ixtureo f t he m aterial that mayb eb rought up to t he sur fa c eb y th e s ubsequentth ird d r edge-up ph a s e .
D espite its si mplicity th e PD CZ mod el yi e lds re sults t h atnicely agre e wi th thoseof ful l TP-AGB computa tions. Adet aile d d is cus si o n o f t hep red ictionsan dtheir mai n d e pend enc iesis gi ven in Sect.\[s sec\ _ xpd cz\].
The syn th eticmodule {#se c_ synthmod}
===== ==== ========= ==
M ost a nalyt i cal in gre die nts ofthe `CO L IBR I` c o de are formulas accu ra t e ly fitting the r e su lt s of full A GBmode l s cove ring wi de range s of i n it ia l stell ar massan d m eta llici t y. T he for mulas ar e tak e n either fromt he extensivec om p i la t ions by @Wagenhube r_96 [@Wa genG r oe n_9 8 ; @Ka rakas _e t al 0 2; @Izzard_etal04;@I zzard_ etal0 6], and other sources [ @ S t raniero_ etal 0 3] , or they are d irect ly derived from AGB mode l data s ets by us i n g standa rd$\c hi^ 2$- m i ni mization tech n i ques .New fit s c an be f oun d i n A ppe nd ix \[app\ _fit\].
I mp or ta ntl y, al l these a na lyt ic re latio n s incl ude a met al li c ity depend e nc e , and t ak e in toac count the pec uliar b ehaviouroft he f ir st sub-lu minous pulses w hile appro ac hin g thef u ll-ampli tude regime.
Among the most im por tantpres criptions we menti ont he fla sh-dri ven l um ino s i ty va r i at ion sas a funct i o n o f the p ulse -cyclephase [@WagenGroen _ 98] , the core ma ss- inte r p ul sep er i odre l ati o n [@WagenGroen_9 8], the ma xi m um mass of t h e P DC Z and i ts dura tion, the max imum temp erature a tt aine d atthe bottom of thePDCZ duri n g a T P [ @Kara kas Lattan zi o_0 7][^7 ], the eff icien cy $\l am bda$ o f the t hird dre dge-up [@Karakas_etal02 ].
Du e tothe ir partic ula r re levance,belo w we willdis cus s inmor e deta il a fe w a n alyti c re l ations ad o pt edi n t he presentv e r sio n of`CO L IBRI`.
Th e third dredge-up : the need fora pa r a met ric desc ri ption {#ssec_3 dup }- - -------- -- ----------- -------- -- - ----- ------ ------ ------- It is com monpra ctice des cri bi n g the t hi rd dredge -upby means of tw o cha r a cteristic quanti ties, n amely :
- $M _{ \rm c}^ { \rmmin | except_that now_the PDCZ convection retreats_and the_inner/upper_borders are_shifted_outward/inward until $t=\tau_{\rm_pdcz}$. The nuclear_network is integrated over_the pulse quenching_phase_and a final homogeneous chemical composition is obtained. This sets the chemical mixture of_the_material that_may_be_brought up to the surface_by the subsequent third dredge-up_phase.
Despite its_simplicity the PDCZ model yields results that nicely_agree_with those of_full TP-AGB computations. A detailed discussion of the predictions_and their main dependencies is given_in Sect. \[ssec\_xpdcz\].
The synthetic_module_{#sec_synthmod}
====================
Most_analytical ingredients of the_`COLIBRI` code are formulas accurately fitting_the results of full AGB models_covering wide ranges of initial stellar mass_and metallicity. The formulas are taken_either from the extensive compilations_by @Wagenhuber_96_[@WagenGroen_98; @Karakas_etal02; @Izzard_etal04; @Izzard_etal06], and_other sources [@Straniero_etal03],_or they_are directly derived_from AGB model data sets by_using standard $\chi^2$-minimization_techniques. New fits can be found_in_Appendix \[app\_fit\].
Importantly, all these_analytic_relations_include a_metallicity dependence, and_take_into account_the_peculiar behaviour of the first sub-luminous_pulses_while approaching the full-amplitude regime.
Among the most_important prescriptions we mention_the_flash-driven luminosity variations as_a function of the pulse-cycle_phase [@WagenGroen_98], the core mass-interpulse period_relation [@WagenGroen_98],_the maximum_mass of the PDCZ and its duration, the maximum temperature attained_at the bottom of the PDCZ_during a TP [@KarakasLattanzio_07][^7],_the efficiency_$\lambda$_of the third_dredge-up_[@Karakas_etal02].
Due to_their particular relevance, below we will discuss_in more_detail a few analytic relations adopted_in the present version_of_`COLIBRI`.
The third dredge-up: the need for_a parametric description {#ssec_3dup}
----------------------------------------------------------
It is common_practice describing the third dredge-up_by_means_of two characteristic quantities, namely:
-_ $M_{\rm c}^{\rm min |
widehat \gamma}$)
-----------------------------------------------------------------------------------------------------
Suppose first that $P_{\widehat \beta} = P_{\widehat \gamma}$. Since $w(\beta) = w(\gamma)$ is valid, assume $w(\beta) > w(\gamma)$. Prop. \[H3braidstdformprop\] shows that $p_{2,\,\widehat \beta} =0$ and $p_{0,\,\widehat \gamma} =0$, i.e. $\nabla_{\widehat \beta} = C_{w(\beta)-1}$ and $\nabla_{\widehat \gamma} = C_{w(\gamma)+1}$. Since $P_{\widehat \beta} = P_{\widehat \gamma}$, we have $C_{w(\beta)-1} = C_{w(\gamma)+1}$, i.e. $w(\beta) = w(\gamma)+2$ or $w(\beta) = -w(\gamma)$. These correspond to the second and third outcomes.
Conversely, when $\nabla_{\widehat \beta} = \nabla_{\widehat \gamma}$, the first condition implies $P_{\widehat \beta} = P_{\widehat \gamma}$ by Prop. \[Writhe3brPlusAny\]. Prop. \[H3braidstdformprop\] shows the remaining two cases also imply $P_{\widehat \beta} = P_{\widehat \gamma}$.
Proof of Proposition \[3brlandscape\] (3-Braid Degree Result)
-------------------------------------------------------------
It suffices to prove either $\deg p_0 \leq \deg p_1$, or $\deg p_2 \leq \deg p_1$, and to consider only the case in which $p_0$ and $p_2$ are both nonzero. By Prop. \[Homflycoeffprops\], the writhe may be assumed to be non-negative. Separate consideration of the cases $\nabla_{\beta} = 0$, $w=0$, $w=1$, $\deg \nabla_{\beta} > w$, $\deg \nabla_{\beta} = w$, $\deg \nabla_{\beta} = w-2$, and $\deg \nabla_{\beta} < w-2$ shows the result is true in all cases.
Proof of Lemma \[HJones3br\] ($V_{\widehat \beta} = V_{\widehat \gamma}$ properties for 3-braid links)
------------------------------------------------------------------------------------------------------
Observe that $V_{\widehat \beta} | widehat \gamma}$)
-----------------------------------------------------------------------------------------------------
Suppose first that $ P_{\widehat \beta } = P_{\widehat \gamma}$. Since $ w(\beta) = w(\gamma)$ is valid, assume $ w(\beta) > w(\gamma)$. Prop. \[H3braidstdformprop\ ] shows that $ p_{2,\,\widehat \beta } = 0 $ and $ p_{0,\,\widehat \gamma } = 0 $, i.e. $ \nabla_{\widehat \beta } = C_{w(\beta)-1}$ and $ \nabla_{\widehat \gamma } = C_{w(\gamma)+1}$. Since $ P_{\widehat \beta } = P_{\widehat \gamma}$, we get $ C_{w(\beta)-1 } = C_{w(\gamma)+1}$, i.e. $ w(\beta) = w(\gamma)+2 $ or $ w(\beta) = -w(\gamma)$. These represent to the second and third outcomes.
Conversely, when $ \nabla_{\widehat \beta } = \nabla_{\widehat \gamma}$, the first condition imply $ P_{\widehat \beta } = P_{\widehat \gamma}$ by Prop. \[Writhe3brPlusAny\ ]. Prop. \[H3braidstdformprop\ ] shows the remaining two case also imply $ P_{\widehat \beta } = P_{\widehat \gamma}$.
Proof of Proposition \[3brlandscape\ ] (3 - Braid Degree Result)
-------------------------------------------------------------
It suffices to prove either $ \deg p_0 \leq \deg p_1 $, or $ \deg p_2 \leq \deg p_1 $, and to consider entirely the case in which $ p_0 $ and $ p_2 $ are both nonzero. By Prop. \[Homflycoeffprops\ ], the writhe may be assumed to be non - minus. freestanding consideration of the cases $ \nabla_{\beta } = 0 $, $ w=0 $, $ w=1 $, $ \deg \nabla_{\beta } > w$, $ \deg \nabla_{\beta } = w$, $ \deg \nabla_{\beta } = w-2 $, and $ \deg \nabla_{\beta } < w-2 $ shows the result is true in all font.
Proof of Lemma \[HJones3br\ ] ($ V_{\widehat \beta } = V_{\widehat \gamma}$ properties for 3 - braid links)
------------------------------------------------------------------------------------------------------
Observe that $ V_{\widehat \beta } | widfhat \gamma}$)
-----------------------------------------------------------------------------------------------------
Suppose first uhat $P_{\widehat \betc} = P_{\widxhat \gajma}$. Sincd $w(\beta) = w(\gamma)$ is valid, assnme $q(\beta) > w(\gamma)$. Prop. \[H3braidstdwormprop\] dhows thqt $p_{2,\,\xidehat \beta} =0$ anv $p_{0,\,\widehat \gamjw} =0$, i.z. $\iabla_{\widehat \beja} = C_{w(\beta)-1}$ atd $\nabla_{\widehad \ealma} = C_{w(\gamma)+1}$. Since $P_{\widehat \beta} = P_{\ridehat \gwmma}$, we have $C_{r(\bets)-1} = C_{w(\famma)+1}$, i.e. $w(\beta) = w(\gamma)+2$ or $w(\beta) = -w(\famma)$. Tiese correspond to the second and third oktcoles.
Conversely, when $\nabla_{\widehqt \bqra} = \nabla_{\widdhat \gamma}$, the first ckndition implies $P_{\widehat \beta} = P_{\widzhat \gamma}$ vy Prlk. \[Writhe3brPluwAny\]. Irop. \[H3braidstdnprmpro[\] shows the remaining twm cqses also imply $P_{\wideiat \beta} = P_{\widehat \gwmma}$.
Proof oy Proposition \[3brlandsxape\] (3-Brdid Gegrdw Rdsumt)
-------------------------------------------------------------
Mt aufficfs vo prove eifher $\deg p_0 \oeq \deg p_1$, or $\deg p_2 \kez \deg p_1$, and to consiqew only the case in which $p_0$ and $p_2$ are bouh nohzero. By Prop. \[Homflycoefdprops\], the writhe may be assumqd to be non-negative. Separate consideration of tha casxs $\nayoa_{\betx} = 0$, $w=0$, $w=1$, $\deg \nabla_{\beta} > w$, $\deg \nabla_{\beta} = w$, $\deg \nafma_{\nena} = w-2$, and $\deg \nabja_{\beta} < w-2$ snoas jhe result is jrue in alm cases.
Proof of Lelma \[HJongs3br\] ($V_{\qidehat \btta} = F_{\widehat \gamma}$ properties fir 3-braid linhw)
------------------------------------------------------------------------------------------------------
Observe that $V_{\widzhat \beta} | widehat \gamma}$) ----------------------------------------------------------------------------------------------------- Suppose first that $P_{\widehat P_{\widehat Since $w(\beta) w(\gamma)$ is valid, \[H3braidstdformprop\] that $p_{2,\,\widehat \beta} and $p_{0,\,\widehat \gamma} i.e. $\nabla_{\widehat \beta} = C_{w(\beta)-1}$ and \gamma} = C_{w(\gamma)+1}$. Since $P_{\widehat \beta} = P_{\widehat \gamma}$, we have $C_{w(\beta)-1} = i.e. $w(\beta) = w(\gamma)+2$ or $w(\beta) = -w(\gamma)$. These correspond to the second third Conversely, $\nabla_{\widehat = \nabla_{\widehat \gamma}$, the first condition implies $P_{\widehat \beta} = P_{\widehat \gamma}$ by Prop. \[Writhe3brPlusAny\]. Prop. shows the remaining two cases also imply $P_{\widehat = P_{\widehat \gamma}$. Proof Proposition \[3brlandscape\] (3-Braid Degree Result) It to prove $\deg \leq p_1$, or $\deg \leq \deg p_1$, and to consider only the case in which $p_0$ and $p_2$ are both nonzero. Prop. \[Homflycoeffprops\], may be to non-negative. consideration of the = 0$, $w=0$, $w=1$, $\deg \nabla_{\beta} \nabla_{\beta} = w$, $\deg \nabla_{\beta} = w-2$, and \nabla_{\beta} < shows the result is true in cases. Proof of Lemma \[HJones3br\] ($V_{\widehat \beta} = \gamma}$ properties for 3-braid links) ------------------------------------------------------------------------------------------------------ Observe that $V_{\widehat \beta} | widehat \gamma}$)
-----------------------------------------------------------------------------------------------------
Suppose first tHat $P_{\widehaT \beta} = p_{\wiDehAt \GammA}$. SinCe $w(\beta) = w(\gamma)$ iS ValiD, assume $w(\beta) > w(\gamma)$. Prop. \[h3braiDsTDforMPrOp\] shoWs that $p_{2,\,\WIdEHAt \bEtA} =0$ aNd $p_{0,\,\WiDEhAt \gamMa} =0$, i.E. $\nabla_{\wIdehat \beta} = c_{w(\bEtA)-1}$ and $\nabla_{\widEHaT \gamma} = C_{w(\gaMma)+1}$. since $P_{\widehaT \beTa} = P_{\widEhAt \gAMma}$, we HavE $C_{w(\beTa)-1} = C_{w(\gaMMa)+1}$, i.e. $w(\bEta) = w(\gamma)+2$ Or $W(\Beta) = -w(\gAMma)$. ThesE COrRespOnd to the second and THiRD outcomes.
ConveRsely, wHeN $\NaBLA_{\wiDehAt \beta} = \nablA_{\wIdehaT \Gamma}$, thE FiRST ConDItion implies $P_{\Widehat \beta} = p_{\WidEhat \gaMmA}$ by pRop. \[WriThe3brplUSAnY\]. Prop. \[H3braidStdfOrmprop\] shOws the REmaininG Two caseS also iMplY $P_{\wIdehAT \bEtA} = P_{\wIdEHat \GAmMa}$.
PROof Of ProposItIoN \[3brlaNdscAPE\] (3-bRaid degRee REsult)
-------------------------------------------------------------
it suffices to pRovE eitHEr $\dEg p_0 \leQ \deg p_1$, Or $\deG p_2 \Leq \deG p_1$, and tO consIdEr only the case in WhicH $p_0$ and $p_2$ are BotH nOnzErO. By PrOP. \[HomflYcoEffProps\], thE writhe MAy bE aSSUMeD to be non-negative. SePaRATe ConsiderAtion oF ThE cASes $\nabla_{\BeTa} = 0$, $w=0$, $W=1$, $\deg \NABla_{\beTa} > w$, $\dEG \nAbla_{\beta} = W$, $\deg \naBLa_{\BeTa} = w-2$, and $\dEg \Nabla_{\bEtA} < w-2$ sHowS the rESult Is true In all casEs.
ProOF of Lemma \[HJones3BR\] ($V_{\widehat \beta} = v_{\WiDEHaT \GammA}$ prOperties for 3-BraiD LinkS)
------------------------------------------------------------------------------------------------------
ObsERvE thAT $V_{\widEhat \bEtA} | widehat \gamma}$)
-------- ---------- ----- --- --- -- ---- ---- -------------- - ---- ---------------------- ----- -- - ---- - -- ----- -
Supp o se f irs tth at$P _ {\ wideh at\beta}= P_{\wide hat \ gamma}$. Sin c e$w(\beta)= w (\gamma)$ is va lid, a ss ume $w(\b eta ) > w (\gamm a )$. Pr op. \[H3b ra i dstdfo r mprop\] s ho ws t hat $p_{2,\,\wide h at \beta} =0$ and $p_{0 ,\ , \w i d eha t \ gamma} =0$ ,i.e.$ \nabla_ { \w i d e hat \beta} = C_{w (\beta)-1}$ and $\nab la _{\ w idehat \gam ma } =C_{w(\gamma )+1} $. Since$P_{\w i dehat \ b eta} =P_{\wi deh at\gam m a} $, we h a ve$ C_ {w( \ bet a)-1} =C_ {w (\gam ma)+ 1 } $ , i.e . $ w(\b eta)= w(\gamma)+2 $ o r $w ( \be ta) = -w(\ gamm a) $. Th ese co rresp on d to the second and third ou tco me s.
C onver s ely, w hen $\ nabla_{ \wideha t \b et a } =\nabla_{\widehat \ ga m m a} $, the f irst c o nd it i on impli es $P _{\w i d ehat\bet a }= P_{\wi dehat\ ga mm a}$ byPr op. \[ Wr ith e3b rPlus A ny\] . Prop . \[H3br aidst d formprop\] sho w s the remaini n gt w oc ases al so imply $P _{\w i deha t \b e ta } = P_{\w ideha t\ ga m ma}$.
Proof of Pro po sition \[3b rlandscape\](3-Braid D e g r ee Resul t)
- - -- - -------------- ----- ---------- - -------- ----- -------- -----
It s ufficestopro veeit h e r$\deg p_0 \le q \deg p _1$, or $\ deg p_2 \l eq\de g p _1 $, and to conside ron ly t hecasei n which$p _0$ a nd$p_2$ are bo th no nzer o. B y Pr op. \[H o mf l y coef fp ro ps\] , t he writ he m a y b e assum ed to benon - nega ti ve . Separ ate considera ti on of theca ses $\nab l a _{\beta} = 0$, $w=0$, $w=1$, $\ d eg \nab la_ {\bet a} > w$, $\de g \ nabla_ {\b e ta} =w$, $\ deg \ na bla _ { \beta } =w-2 $, and $\deg \ nab la_{\ be ta}< w-2$shows the result i s tr ue in all cas es.
Pr o o fofL em m a \ [H J one s 3 br\] ($V_{\wide hat \beta} = V_ {\widehat\ gam ma }$ prop ertiesfor 3 - braid l inks)
--- --------- -- ---- - - --- ---------- -------- --------- - ----- - -- ----- --- ------ -- --- ----- ------ - --- ----- ----
Ob servethat$V _{\wideh at \beta} | widehat \gamma}$)
-----------------------------------------------------------------------------------------------------
Suppose_first that_$P_{\widehat \beta} = P_{\widehat_\gamma}$. Since_$w(\beta)_= w(\gamma)$_is_valid, assume $w(\beta)_> w(\gamma)$. Prop. \[H3braidstdformprop\]_shows that $p_{2,\,\widehat \beta}_=0$ and $p_{0,\,\widehat_\gamma}_=0$, i.e. $\nabla_{\widehat \beta} = C_{w(\beta)-1}$ and $\nabla_{\widehat \gamma} = C_{w(\gamma)+1}$. Since $P_{\widehat \beta}_=_P_{\widehat \gamma}$,_we_have_$C_{w(\beta)-1} = C_{w(\gamma)+1}$, i.e. $w(\beta)_= w(\gamma)+2$ or $w(\beta) =_-w(\gamma)$. These_correspond to the second and third outcomes.
Conversely, when_$\nabla_{\widehat_\beta} = \nabla_{\widehat_\gamma}$, the first condition implies $P_{\widehat \beta} = P_{\widehat_\gamma}$ by Prop. \[Writhe3brPlusAny\]. Prop. \[H3braidstdformprop\] shows the_remaining two cases_also_imply_$P_{\widehat \beta} = P_{\widehat_\gamma}$.
Proof of Proposition \[3brlandscape\] (3-Braid Degree_Result)
-------------------------------------------------------------
It suffices to prove either $\deg_p_0 \leq \deg p_1$, or $\deg p_2_\leq \deg p_1$, and to consider_only the case in which_$p_0$ and_$p_2$ are both nonzero. By_Prop. \[Homflycoeffprops\], the writhe_may be_assumed to be_non-negative. Separate consideration of the cases_$\nabla_{\beta} = 0$,_$w=0$, $w=1$, $\deg \nabla_{\beta} > w$,_$\deg_\nabla_{\beta} = w$,_$\deg_\nabla_{\beta}_= w-2$,_and $\deg \nabla_{\beta}_<_w-2$ shows_the_result is true in all cases.
Proof_of_Lemma \[HJones3br\] ($V_{\widehat \beta} = V_{\widehat \gamma}$ properties_for 3-braid links)
------------------------------------------------------------------------------------------------------
Observe that_$V_{\widehat_\beta} |
we varied the optical loss of the channel to Bob between $0\,\%$ and $16\,\%$ (see Figure \[fig:results\](b)), which is equivalent to a fiber length of up to $2.7$km when standard telecommunication fibers with an attenuation of $0.2$dB/km are used and a coupling efficiency of $95$$\%$ is taken into account. By measuring a total of $2\times 10^8$ samples we were still able to achieve a secret key rate of about $0.1$bit/sample at an equivalent fiber length of $2.7$km ($\approx 0.76$dB channel loss). This value, as well as the secret key sizes at the other attenuation values, were achieved by having a very high overall error reconciliation efficiency between $\beta = 94.3\,\%$ and $95.5\,\%$, again without a single frame error. The theoretical model shown in the figure reveals that even an optical transmission loss of almost $1.2$dB between Alice and Bob should be possible. This corresponds to an equivalent distance of about $4.8$km, which is already enough to implement CV QKD links with composable security against coherent attacks between parties in, for instance, a city’s central business district.
In conclusion, we have for the first time successfully implemented QKD with composable security against coherent attacks in a one-sided measurement-device independent manner. Along with the exploitation of strong Einstein-Podolsky-Rosen entanglement and a new highly efficient error reconciliation algorithm, the innovation of fast controlled random switching between the two measured quadrature angles made the implementation possible. Although attacks on the source of the quantum states are not covered by the security proof [@Furrer2012], these side-channel attacks can be avoided to arbitrarily low probabilities by adding optical isolators. While in our setup Alice and Bob were located on the same optical table, they could in principle be separated and connected by a standard telecommunication fiber (see Methods). Estimations show that our implementation is limited to about $4.8$km. Longer distances will be possible by using optical fibres with less loss, or by using reverse reconciliation [@Furrer2014].
#### **Acknowledgements**
This research was supported by the Deutsche Forschungsgemeinschaft (projects SCH 77/5-1 and WE 1240/20-1), the Centre for Quantum Engineering and Space-Time Research (QUEST | we varied the optical loss of the channel to Bob between $ 0\,\%$ and $ 16\,\%$ (visualize Figure \[fig: results\](b) ), which is equivalent to a roughage length of up to $ 2.7$km when standard telecommunication fibers with an attenuation of $ 0.2$dB / km are used and a yoke efficiency of $ 95$$\%$ is taken into account. By measuring a sum of $ 2\times 10 ^ 8 $ samples we were still able to achieve a clandestine key rate of approximately $ 0.1$bit / sample at an equivalent fiber length of $ 2.7$km ($ \approx 0.76$dB channel personnel casualty). This value, as well as the secret key sizes at the other attenuation value, were achieved by having a very high overall erroneousness reconciliation efficiency between $ \beta = 94.3\,\%$ and $ 95.5\,\%$, again without a single frame error. The theoretical model show in the figure reveals that even an optical transmission loss of about $ 1.2$dB between Alice and Bob should be possible. This corresponds to an equivalent distance of about $ 4.8$km, which is already enough to implement CV QKD liaison with composable security against coherent attacks between parties in, for instance, a city ’s central business district.
In conclusion, we have for the first time successfully implemented QKD with composable security system against coherent attack in a one - sided measurement - device independent manner. Along with the exploitation of impregnable Einstein - Podolsky - Rosen web and a new highly efficient error reconciliation algorithm, the innovation of fast controlled random switching between the two measured quadrature angles lay down the implementation possible. Although attacks on the source of the quantum states are not covered by the security proof [ @Furrer2012 ], these side - channel attacks can be debar to randomly abject probabilities by adding optical isolators. While in our setup Alice and Bob were located on the same optical mesa, they could in principle be separated and connected by a standard telecommunication fiber (experience Methods). estimate show that our implementation is restrict to about $ 4.8$km. long distances will be potential by using optical character with less loss, or by using inverse reconciliation [ @Furrer2014 ].
# # # # * * Acknowledgements * *
This inquiry was supported by the Deutsche Forschungsgemeinschaft (projects SCH 77/5 - 1 and WE 1240/20 - 1), the Centre for Quantum Engineering and Space - Time Research (QUEST | we varied the optical loss of the channel to Bob betwesn $0\,\%$ and $16\,\%$ (see Figure \[fig:results\](b)), which iw equuvalent to a fiber leneth of up to $2.7$km wyen wrandard telecommunleatioh fibzrw with an attekuation of $0.2$gB/km are used dna c coupling efficiency of $95$$\%$ is taken igto accpujt. By measurind a uotwl or $2\times 10^8$ samples we were still abls to aciieve a secret ley rate of about $0.1$bit/samplf at an equivalent fibfr length od $2.7$km ($\qpprox 0.76$dB chxnnel loss). This value, zs well as the secret key sizes at tke other atjzbuahhon values, xere abhieved by haypng a vary hign overall errov recmnculiation efficiency bxtween $\beta = 94.3\,\%$ and $95.5\,\%$, adain withmuc a single frame erroe. Rhe tveoraticxo mudem xhkwn in thx figure rebeals that wven an optical tramsipxsion loss or almoft $1.2$dB between Alice and Bob should be posvibme. This corresponds to qn equivalent distancg of about $4.8$km, which is already enough to implement CV QKD lhnks xigh eimposxvlf security against coherent attacks between pwdtoex in, for instakce, a city’s centrak huxyness districj.
In conemuaion, we have for tje firsj time successftlly implemented QKD with compowable securiny atainst coherent atcacks in a oue-sidec meaxurement-device independznt mahner. Along aith the sbploitation of sgrokg Ainstein-Kudolsky-Rosen entagglement end a new hiehly efficyent error recokwiliation algorithl, the itnovation lf fast controlled random switcimng between tne twm measurzd quacrature anglef made the impkementacion pussible. Alnhough atvacks on the source of tha quantum staves are njt civerwd by tfd security propf [@Furrer2012], these sidw-channel attacks cen bg zvoided to arbicxatuly low probaboligief hy aqging optical isouatuts. Whkle in our wdtup Alice and Bob were nocafed on the same opyigal table, they cotld in princille be separated ajd coinectev by a stwndard telecommunication fiber (see Methlds). Estimations shoq that our ikplementation is limited to about $4.8$km. Loiger distances will be possible by using oktigal fibres wmth lefs loss, os by using reverse rwconciliation [@Furrtr2014].
#### **Acknowledgements**
This reseasch wws supported by the Deutsche Forschungsgemeinschaft (projects SCH 77/5-1 and WE 1240/20-1), the Cxnere for Quahtum Engitezriug and Fpacx-Tmme Research (QUESN | we varied the optical loss of the Bob $0\,\%$ and (see Figure \[fig:results\](b)), fiber of up to when standard telecommunication with an attenuation of $0.2$dB/km are and a coupling efficiency of $95$$\%$ is taken into account. By measuring a of $2\times 10^8$ samples we were still able to achieve a secret key of $0.1$bit/sample an fiber length of $2.7$km ($\approx 0.76$dB channel loss). This value, as well as the secret key at the other attenuation values, were achieved by a very high overall reconciliation efficiency between $\beta = and again without single error. theoretical model shown the figure reveals that even an optical transmission loss of almost $1.2$dB between Alice and Bob should possible. This an equivalent of $4.8$km, is already enough CV QKD links with composable security between parties in, for instance, a city’s central district. In we have for the first time implemented QKD with composable security against coherent attacks a one-sided measurement-device independent manner. Along with the exploitation of strong Einstein-Podolsky-Rosen entanglement and a efficient error reconciliation algorithm, innovation of fast random between two quadrature angles the implementation possible. Although attacks on the source of the quantum are not covered by the security proof [@Furrer2012], these side-channel be to arbitrarily low by adding optical isolators. in setup Alice and Bob on same could principle separated and connected by standard telecommunication fiber (see Methods). show that our implementation Longer distances will be possible by using optical with less loss, or by using reverse [@Furrer2014]. #### **Acknowledgements** This research was supported by the Deutsche Forschungsgemeinschaft (projects 77/5-1 and the Centre for Quantum Engineering and Space-Time Research | we varied the optical loss of tHe channel tO Bob bEtwEen $0\,\%$ AnD $16\,\%$ (see figuRe \[fig:results\](b)), wHIch iS equivalent to a fiber lenGth of Up TO $2.7$km wHEn StandArd teleCOmMUNicAtIoN fiBeRS wIth an AttEnuatioN of $0.2$dB/km are UseD aNd a coupling eFFiCiency of $95$$\%$ is TakEn into accounT. By MeasurInG a tOTal of $2\TimEs 10^8$ samPles we WEre stiLl able to aChIEve a seCRet key rATE oF aboUt $0.1$bit/sample at an eqUIvALent fiber lengtH of $2.7$km ($\aPpROx 0.76$Db ChaNneL loss). This vAlUe, as wELl as the SEcRET Key SIzes at the otheR attenuatioN ValUes, werE aChiEVed by hAving A vERy hIgh overall eRror ReconciliAtion eFFiciencY Between $\Beta = 94.3\,\%$ anD $95.5\,\%$, agAin WithOUt A sIngLe FRamE ErRor. tHe tHeoreticAl MoDel shOwn iN THE FiguRe rEveaLs thaT even an opticaL trAnsmISsiOn losS of alMost $1.2$DB BetweEn AlicE and BOb Should be possiblE. ThiS correspoNds To An eQuIvaleNT distaNce Of aBout $4.8$km, wHich is aLReaDy ENOUgH to implement CV QKD lInKS WiTh composAble seCUrItY Against cOhEreNt atTACks beTweeN PaRties in, fOr instANcE, a City’s ceNtRal busInEss DisTrict.
iN conClusioN, we have fOr the FIrst time succesSFully implemenTEd qkd wITh coMpoSable securiTy agAInst CoheREnT atTAcks iN a one-SiDEd MEasurement-device indEpEndent ManneR. Along with the ExploitatiON OF strong EInstEIn-pOdolsky-Rosen enTanglEment and a nEW highly eFficiEnt error ReconciliATIon algorIthM, thE inNovATIoN of fast controLLEd raNdOm switcHinG betweeN thE twO meAsuReD quadratuRe angles MaDe ThE iMplEmentATion possIbLe. ALtHouGh attACks on tHe souRce oF tHe QUanTum statES aRE Not cOvErEd by The SeCuritY proOF [@FuRrer2012], theSe side-chaNneL AttaCkS cAn be avoIded to arbitraRiLy low probaBiLitIes by aDDIng opticAl isolators. While in our seTUp Alice And bob weRe loCated on thE saMe optiCal TAble, thEy coulD in prInCipLE Be sepARAtEd aNd Connected bY A StaNdard TeLecoMmunicaTion fiber (see MethodS). estImations show tHat Our iMPLeMenTAtIOn iS lIMitED To about $4.8$km. Longer Distances wIlL Be Possible by USinG oPtical fIbres wiTh lesS Loss, or bY using revErse reconCiLiatION [@FuRrer2014].
#### **AcknowLedgemenTs**
This resEArch wAS sUpporTed By the DEuTscHe ForSchungSGemEinscHaft (prOjEcts SCh 77/5-1 and We 1240/20-1), tHe Centre For Quantum Engineering anD Space-time REseArch (QUEST | we varied the optical los s of the c hanne l t o B ob bet ween $0\,\%$ and $ 1 6\,\ %$ (see Figure \[fig:r esult s\ ] (b)) , w hichis equi v al e n t t oafib er le ngthofup to $ 2.7$km whe n s ta ndard teleco m mu nication f ibe rs with an a tte nuatio nof$ 0.2$d B/k m are useda nd a c oupling e ff i ciency of $95$ $ \ %$ istaken into accoun t .B y measuring atotalof $2 \ t ime s 1 0^8$ sampl es we w e re stil l a b l e to achieve a sec ret key rat e of about $ 0.1 $ bit/sa mpleat anequivalentfibe r lengthof $2. 7 $km ($\ a pprox 0 .76$dB ch ann el l o ss ). Th is val u e, as wel l as the s ec ret k ey s i z e s atthe oth er at tenuation val ues , we r e a chiev ed by hav in g a v ery hi gh ov er all error recon cili ation eff ici en cybe tween $\beta =94. 3\,\%$and $95 . 5\, \% $ , ag ain without a sing le f ra me error . Thet he or e tical mo de l s hown i n the fig u re reveals thate ve nan opti ca l tran sm iss ion loss of a lmost$1.2$dBbetwe e n Alice and Bo b should be po s si b l e. This co rresponds t o an equi vale n tdis t anceof ab ou t $ 4 .8$km, which is alr ea dy eno ugh t o implement C V QKD link s w ith comp osab l es ecurity agains t coh erent atta c ks betwe en pa rties in , for ins t a nce, a c ity ’scen tra l bu siness distri c t .
I nconclus ion , we ha vefor th e f ir st time s uccessfu ll yim pl eme ntedQ KD withco mpo sa ble secu r ity ag ainst coh er en t at tacks i n a o ne-s id ed mea sur em ent-d evic e in depende nt manner . A l ongwi th the ex ploitation of s trong Eins te in- Podols k y -Rosen e ntanglement and a new h i ghly ef fic ienterro r reconci lia tion a lgo r ithm,the in novat io n o f fastc o nt rol le d random s w i tch ing b et ween the tw o measured quadrat u reangles made t heimpl e m en tat i on pos si b le. A lthough attacks on the so ur c eof the qua n tum s tates a re notcover e d by th e securit y proof [ @F urre r 2 012 ], these s ide-chan nel attac k s can be avoi ded to ar bi tra rilylow pr o bab iliti es byad ding o ptica lisolator s. While in our setup A lice a nd Bo b w ere locat edo n t he same o ptic al table,the y c ouldinp rinci pleb esep a rated and connected by as t an dard teleco m m u nic ation fi b er (se e Me thods). Estimatio n s show that ou r im p l eme nta t ionis limited to ab out $ 4 . 8$km. Lo ng er distance s will b ep ossib le byusingoptical f ib r es wit h le ssloss, orbyus i ng reve rs er econci liat io n [@Fu rrer20 1 4].# ### **Acknowledg ement s * *
Th i s r esear ch was su p port ed by theDeutsche Fo rschun gsge meins chaft ( pr ojects SC H77/5-1 and WE 1240/2 0-1), the Ce nt re f orQuantu m En g i neeri ng a nd Sp ace-TimeR e se a rc h( QUE ST | we_varied the_optical loss of the_channel to_Bob_between $0\,\%$_and_$16\,\%$ (see Figure \[fig:results\](b)),_which is equivalent_to a fiber length_of up to_$2.7$km_when standard telecommunication fibers with an attenuation of $0.2$dB/km are used and a coupling_efficiency_of $95$$\%$_is_taken_into account. By measuring a_total of $2\times 10^8$ samples_we were_still able to achieve a secret key rate_of_about $0.1$bit/sample at_an equivalent fiber length of $2.7$km ($\approx 0.76$dB channel_loss). This value, as well as_the secret key_sizes_at_the other attenuation values,_were achieved by having a very_high overall error reconciliation efficiency between_$\beta = 94.3\,\%$ and $95.5\,\%$, again without_a single frame error. The theoretical_model shown in the figure_reveals that_even an optical transmission loss_of almost $1.2$dB_between Alice_and Bob should_be possible. This corresponds to an_equivalent distance of_about $4.8$km, which is already enough_to_implement CV QKD_links_with_composable security_against coherent attacks_between_parties in,_for_instance, a city’s central business district.
In_conclusion,_we have for the first time successfully_implemented QKD with composable_security_against coherent attacks in_a one-sided measurement-device independent manner._Along with the exploitation of strong_Einstein-Podolsky-Rosen entanglement_and a_new highly efficient error reconciliation algorithm, the innovation of fast controlled_random switching between the two measured_quadrature angles made the_implementation possible._Although_attacks on the_source_of the_quantum states are not covered by the_security proof_[@Furrer2012], these side-channel attacks can be_avoided to arbitrarily low_probabilities_by adding optical isolators. While in_our setup Alice and Bob were_located on the same optical_table,_they_could in principle be separated_and connected by a standard telecommunication_fiber (see Methods)._Estimations show that our implementation is limited_to_about $4.8$km. Longer distances will be_possible_by using optical fibres with less_loss,_or_by using reverse reconciliation [@Furrer2014].
#### **Acknowledgements**
This_research was supported by the Deutsche_Forschungsgemeinschaft (projects SCH 77/5-1 and WE 1240/20-1), the Centre_for Quantum Engineering_and Space-Time Research (QUEST |
.]{} An example can be seen in the analysis made by @padwhite2008 in Fourier space and @xu2013 in configuration space. The latter recognize the need of introducing a two-step transformation, one isotropic dilation and one warping transformation, to transform from real fiducial to real space. However, the real fiducial space is actually redshift-space, and this identification is missing in these works.
Therefore, we argue (c.f., Section \[sec:sec2\]) that it is convenient to define the observable-redshift-space ${\bm \sigma}$ (${\sigma}$-space) given by the simple redshift differences and subtended angles that are truly observable, and that do not depend on any choice of cosmological parameters. Multiplying by a units function (scale factor) produces the physical redshift-space ($s-$space): ${\bm s} = K({\bm \Omega},z) ~ {\bm \sigma}$, that is isomorphic to the observable ${ \sigma}$-space, but has actual distance units that are dependent on a particular cosmological set of parameters [$\Omega$]{} and the redshift $z$. The $K({\bm \Omega},z)$ function is chosen so that the physical redshift-space is related to real space ${\bm r}$ by a unitary Jacobian independent of redshift. So that no additional scaling is needed, and the only remaining difference will be precisely in shape. That is why ${\bm \sigma}$ and ${\bm s}$ are more alike, and thus can both be named redshift-space; ${\bm \sigma}$ is the observable redshift-space while ${\bm s}$ is the physical redshift-space. Then, the transformation to real-space necessarily goes through redshift distortions.
Furthermore, when we introduce peculiar non-relativistic velocities in this scheme, we will show that it is possible to keep the same relation between observable and physical redshift-spaces, ${\bm s}$ and ${\bm \sigma}$, and that the @kaiser87 effect is recovered independently of redshift (see Section \[sec:sec3\]). That is, now redshift-space will also show an additional gravitational distortion with respect to real-space.
To solve for the relation between real-space and redshift-space, we will rely on projected correlations. Projections of the 2PCF in the plane of the sky have been widely used to avoid the complications of dealing with unknown components in redshift-space [e.g., @davis_peebles_83]. This has the advantage that | . ] { } An example can be seen in the analysis made by @padwhite2008 in Fourier quad and @xu2013 in shape space. The latter recognize the motivation of introduce a two - step transformation, one isotropic dilation and one warping transformation, to translate from real fiducial to real space. However, the actual fiducial space is actually red shift - space, and this identification is missing in these works.
consequently, we argue (c.f. , Section \[sec: sec2\ ]) that it is convenient to define the discernible - redshift - quad $ { \bm \sigma}$ ($ { \sigma}$-space) contribute by the simple redshift differences and subtend angles that are truly discernible, and that do not depend on any option of cosmological parameters. Multiplying by a units function (scale agent) produces the physical redshift - space ($ s-$space ): $ { \bm s } = K({\bm \Omega},z) ~ { \bm \sigma}$, that is isomorphic to the observable $ { \sigma}$-space, but has actual distance units that are dependent on a particular cosmological set of parameters [ $ \Omega$ ] { } and the redshift $ z$. The $ K({\bm \Omega},z)$ function is chosen so that the physical red shift - outer space is related to veridical space $ { \bm r}$ by a unitary Jacobian independent of redshift. therefore that no additional scaling is needed, and the only remaining remainder will be precisely in shape. That is why $ { \bm \sigma}$ and $ { \bm s}$ are more alike, and thus can both be named redshift - space; $ { \bm \sigma}$ is the observable red shift - space while $ { \bm s}$ is the physical redshift - space. Then, the transformation to real - space necessarily run low through redshift distortions.
Furthermore, when we introduce peculiar non - relativistic velocities in this schema, we will show that it is possible to keep the like relation between discernible and physical redshift - spaces, $ { \bm s}$ and $ { \bm \sigma}$, and that the @kaiser87 effect is recovered independently of red shift (see Section \[sec: sec3\ ]). That is, now redshift - space will also prove an additional gravitational distorted shape with respect to veridical - space.
To solve for the relation back between actual - space and red shift - space, we will rely on projected correlations. projection of the 2PCF in the airplane of the sky have been widely used to debar the complications of dealing with unknown components in redshift - space [ e.g., @davis_peebles_83 ]. This have the advantage that | .]{} An example can be seen in uhe analysis made by @padxhite2008 ih Fourief space and @xu2013 in configuratmon wpace. The latter recognize ghe need lf introeuciig a two-step traiaformatljn, ohc isocripic dilation snd one wasping transforkagiln, to transform from real fiducial eo real soace. However, tre rtal fidhbicl space is actually redshift-spzce, and this identifocation is missing in thesf wogks.
Therefore, we arhue (c.f., Sectuon \[sqx:sec2\]) that it is convenpznt to defihe the observable-redshift-space ${\cm \siyma}$ (${\sigma}$-spqcw) gleen by the wimplv redshift dingerencas and xubtended anglcs thet aee truly observable, aid that do not depenq on any whkice of cosmologixao pardmetars. Oyltkplgiig gy a ujita function (scale factir) produces the phyxisql redshift-spzce ($s-$s[ase): ${\bm s} = K({\bm \Omega},z) ~ {\bm \sigma}$, that is ivomkrphic to the observablw ${ \sigma}$-space, but has actual dystance units that are dependent on a particular wosmomugieql seg ov parameters [$\Omega$]{} and the redshift $z$. The $K({\bm \Kmtga},d)$ function is chofen so that tje khysical redshkft-spaes js related to real space ${\fm r}$ vy a unitwry Kacobian independent of redwhift. So than no additional scaliny is needed, cnd thg only remaining difference wnll be precisely ln shape. Ffat is why ${\bm \siema}$ atd ${\bm s}$ are more alike, and thus can botk be namdd rgdshift-fpace; ${\bm \slgma}$ lv the observable rfdshiyt-spawe while ${\bl s}$ is the physical redshift-sparx. Then, the trsnvfogmation tj reak-space necesswrily goes thrpugh rebshift distortiohs.
Furthxrmore, when re introduce [gculiar non-renativistyc vwlocuties iv this scheme, wr will show that it is possible to kecp thg aame relation bzuwetn observable anc pfysycwl rqgshift-spaces, ${\bm r}$ avc ${\bm \rigma}$, and tmat the @kaiser87 effect is rewoveded independently pf redshifj (see Seceion \[sec:sec3\]). Thst is, now redshift-dpace wiul alxo fhow an additional gravitationzl distorhiok with respece to real-space.
To solve for the relation between real-spare and redshift-space, wg will rely on projecjed correlatiois. Protections mf the 2PCF in the plqne of the sky haye been widely used to avoid the fomplications of dealing with unknown components in redshift-space [e.g., @daviw_peeblxs_83]. This has tge acvantdgz tkat | .]{} An example can be seen in made @padwhite2008 in space and @xu2013 recognize need of introducing two-step transformation, one dilation and one warping transformation, to from real fiducial to real space. However, the real fiducial space is actually and this identification is missing in these works. Therefore, we argue (c.f., Section that is to the observable-redshift-space ${\bm \sigma}$ (${\sigma}$-space) given by the simple redshift differences and subtended angles that are observable, and that do not depend on any of cosmological parameters. Multiplying a units function (scale factor) the redshift-space ($s-$space): s} K({\bm ~ {\bm \sigma}$, is isomorphic to the observable ${ \sigma}$-space, but has actual distance units that are dependent on a cosmological set [$\Omega$]{} and redshift The \Omega},z)$ function is that the physical redshift-space is related ${\bm r}$ by a unitary Jacobian independent of So that additional scaling is needed, and the remaining difference will be precisely in shape. That why ${\bm \sigma}$ and ${\bm s}$ are more alike, and thus can both be named \sigma}$ is the observable while ${\bm s}$ the redshift-space. the to real-space goes through redshift distortions. Furthermore, when we introduce peculiar non-relativistic velocities this scheme, we will show that it is possible to same between observable and redshift-spaces, ${\bm s}$ and \sigma}$, that the @kaiser87 effect independently redshift That now will also show an gravitational distortion with respect to To solve for the we will rely on projected correlations. Projections of 2PCF in the plane of the sky been widely used to avoid the complications of dealing with unknown components redshift-space [e.g., has the advantage that | .]{} An example can be seen in the anAlysis made By @padWhiTe2008 iN FOuriEr spAce and @xu2013 in confIGuraTion space. The latter recoGnize ThE Need OF iNtrodUcing a tWO-sTEP trAnSfOrmAtIOn, One isOtrOpic dilAtion and onE waRpIng transformATiOn, to transfOrm From real fiduCiaL to reaL sPacE. howevEr, tHe reaL fiducIAl spacE is actualLy REdshifT-Space, anD THiS ideNtification is missINg IN these works.
TheRefore, We ARgUE (C.f., SEctIon \[sec:sec2\]) tHaT it is COnvenieNT tO DEFinE The observable-Redshift-spaCE ${\bm \Sigma}$ (${\sIgMa}$-sPAce) givEn by tHe SImpLe redshift dIffeRences and SubtenDEd angleS That are Truly oBseRvaBle, aND tHaT do NoT DepENd On aNY chOice of coSmOlOgicaL parAMETErs. MUltIplyIng by A units functioN (scAle fACtoR) prodUces tHe phYsIcal rEdshifT-spacE ($s-$Space): ${\bm s} = K({\bm \OmegA},z) ~ {\bm \Sigma}$, that Is iSoMorPhIc to tHE obserVabLe ${ \sIgma}$-spaCe, but haS ActUaL DIStAnce units that are dePeNDEnT on a partIcular COsMoLOgical seT oF paRameTERs [$\OmeGa$]{} anD ThE redshifT $z$. The $K({\BM \OMeGa},z)$ funcTiOn is chOsEn sO thAt the PHysiCal redShift-spaCe is rELated to real spaCE ${\bm r}$ by a unitarY jaCOBiAN indEpeNdent of redsHift. sO thaT no aDDiTioNAl scaLing iS nEEdED, and the only remaininG dIffereNce wiLl be precisely In shape. ThaT IS Why ${\bm \sigMa}$ anD ${\Bm S}$ Are more alike, anD thus Can both be nAMed redshIft-spAce; ${\bm \sigMa}$ is the obSERvable reDshIft-SpaCe wHILe ${\Bm s}$ is the physiCAL redShIft-spacE. ThEn, the trAnsForMatIon To Real-space NecessarIlY gOeS tHroUgh reDShift disToRtiOnS.
FuRtherMOre, wheN we inTrodUcE pECulIar non-rELaTIVistIc VeLociTieS iN this ScheME, we Will shoW that it is PosSIble To KeEp the saMe relation betWeEn observabLe And PhysicAL Redshift-Spaces, ${\bm s}$ and ${\bm \sigma}$, and tHAt the @kaIseR87 effeCt is Recovered IndEpendeNtlY Of redsHift (seE SectIoN \[seC:SEc3\]). ThaT IS, nOw rEdShift-space WILl aLso shOw An adDitionaL gravitational distORtiOn with respect To rEal-sPACe.
to sOLvE For ThE RelATIon between real-sPace and redShIFt-Space, we wilL RelY oN projecTed corrElatiONs. ProjeCtions of tHe 2PCF in thE pLane OF The Sky have beeN widely uSed to avoiD The coMPlIcatiOns Of dealInG wiTh unkNown coMPonEnts iN redshIfT-space [E.g., @davIs_Peebles_83]. THis has the advantage that | .]{} An example can be see n in the a nalys ismad eby @ padw hite2008 in Fo u rier space and @xu2013 inconfi gu r atio n s pace. The la t te r rec og ni zeth e n eed o f i ntroduc ing a two- ste ptransformati o n, one isotr opi c dilation a ndone wa rp ing trans for matio n, tot ransfo rm from r ea l fiduc i al to r e a lspac e. However, the r e al fiducial space is ac tu a ll y red shi ft-space,an d thi s identi f ic a t i oni s missing inthese works .
T herefo re , w e argue (c.f ., Sec tion \[sec: sec2 \]) thatit isc onvenie n t to de fine t heobs erva b le -r eds hi f t-s p ac e $ { \bm \sigma} $($ {\sig ma}$ - s p a ce)giv en b y the simple redsh ift dif f ere ncesand s ubte nd ed an gles t hat a re truly observab le,and thatdono t d ep end o n any c hoi ceof cosm ologica l pa ra m e t er s. Multiplying byau n it s functi on (sc a le f a ctor) pr od uce s th e physi calr ed shift-sp ace ($ s -$ sp ace): $ {\ bm s}=K({ \bm \Ome g a},z ) ~ {\ bm \sigm a}$,t hat is isomorp h ic to the obs e rv a b le ${ \ sig ma}$-space, but hasactu a ldis t anceunits t h at are dependent on apa rticul ar co smological se t of param e t e rs [$\Om ega$ ] {} and the redshi ft $z $. The $K( { \bm \Ome ga},z )$ funct ion is ch o s en so th atthe ph ysi c a lredshift-spac e is r el ated to re al spac e $ {\b m r }$by a unitar y Jacobi an i nd ep end ent o f redshif t. So t hat no a d dition al sc alin gis nee ded, an d t h e onl yre main ing d iffer ence wil l be pr ecisely i n s h ape. T ha t is wh y ${\bm \sigm a} $ and ${\b ms}$ are m o r e alike, and thus can both be n a med red shi ft-sp ace; ${\bm \s igm a}$ is th e obser vableredsh if t-s p a ce wh i l e${\ bm s}$ is th e phy sical r edsh ift-spa ce. Then, the tran s for mation to rea l-s pace n ec ess a ri l y g oe s th r o ugh redshift di stortions.
F ur thermore,w hen w e intro duce pe culia r non-re lativisti c velocit ie s in t his scheme, w e will s how thati t isp os sible to keepth e s ame r elatio n be tween obser va ble an d phy si cal reds hift-spaces, ${\bm s}$and ${ \bm \ sig ma}$, and th a t t he @kaise r87effect isrec ove red i nde p enden tlyo fred s hift(see Section \ [ se c:s e c 3\ ]). That is , n owredsh ift - spacewill also show an add i tional gravita tion a l di sto r tion w ith respect to re al - s pace.
T osolve for t he relat io n betw een re al-spa ce andr e ds h ift-sp ace, we will rel y o np rojecte dco r relati ons. P roject ions o f the 2 PCF in the plane of t h e skyh ave been w idely u s ed t o avoid th e complicat ions o f de aling with u nk nown c omp on ents in re d shift-spa ce [e .g., @d av is_p eeb les_83 ]. T h i s has the a dva ntage tha t | .]{} An_example can_be seen in the_analysis made_by_@padwhite2008 in_Fourier_space and @xu2013_in configuration space._The latter recognize the_need of introducing_a_two-step transformation, one isotropic dilation and one warping transformation, to transform from real fiducial_to_real space._However,_the_real fiducial space is actually_redshift-space, and this identification is_missing in_these works.
Therefore, we argue (c.f., Section \[sec:sec2\]) that it_is_convenient to define_the observable-redshift-space ${\bm \sigma}$ (${\sigma}$-space) given by the simple_redshift differences and subtended angles that_are truly observable,_and_that_do not depend on_any choice of cosmological parameters. Multiplying_by a units function (scale factor)_produces the physical redshift-space ($s-$space): ${\bm s}_= K({\bm \Omega},z) ~ {\bm \sigma}$,_that is isomorphic to the_observable ${_\sigma}$-space, but has actual distance_units that are_dependent on_a particular cosmological_set of parameters [$\Omega$]{} and the_redshift $z$. The_$K({\bm \Omega},z)$ function is chosen so_that_the physical redshift-space_is_related_to real_space ${\bm r}$_by_a unitary_Jacobian_independent of redshift. So that no_additional_scaling is needed, and the only remaining_difference will be precisely_in_shape. That is why_${\bm \sigma}$ and ${\bm s}$_are more alike, and thus can_both be_named redshift-space;_${\bm \sigma}$ is the observable redshift-space while ${\bm s}$ is the_physical redshift-space. Then, the transformation to_real-space necessarily goes through_redshift distortions.
Furthermore,_when_we introduce peculiar_non-relativistic_velocities in_this scheme, we will show that it_is possible_to keep the same relation between_observable and physical redshift-spaces,_${\bm_s}$ and ${\bm \sigma}$, and that_the @kaiser87 effect is recovered independently_of redshift (see Section \[sec:sec3\]). That_is,_now_redshift-space will also show an_additional gravitational distortion with respect to_real-space.
To solve for_the relation between real-space and redshift-space, we_will_rely on projected correlations. Projections of_the_2PCF in the plane of the_sky_have_been widely used to avoid_the complications of dealing with unknown_components in redshift-space [e.g., @davis_peebles_83]. This has the advantage_that |
ua and BPS spectrum, which implies that one can study the geometry of the vacuum bundle either in one or two dimensions [@rif10]. The superpotential in SQM is dimensionless and arises from the compactification of the $2d$ theory. The matching scale is given by the radious $R$ of the $tt^{*}$ circle and we have $\mathcal{W}_{\mathrm{SQM}}=-\mathcal{W}_{2d}/\epsilon_{2}$. According to the AGT correspondence, the amplitudes of the Penner matrix models compute the conformal blocks of the Liouville CFT and hence the FQHE wave functions in the Vafa’s model. To match the normalization of the Liouville correlators given in [@rif1] one has to identify $-m_{a}/\epsilon_{2}=1$ and $1/\nu=\epsilon_{1}/\epsilon_{2}$. The AGT correspondence has been generalized in [@rif29] also to class $\mathcal{S}$ theories with irregular punctures in the Gaiotto surface. It turns out that the Nekrasov partition functions of these models reproduce irregular conformal blocks of the Liouville CFT. The presence of higher degree poles in the quadratic differential corresponds to insertions in the Liouville correlators of the so called irregular vertex operators [@rif30; @rif31]. As for the mass parameters, the relevant coupling $\mu_{2d}$ of the $2d$ theory is related to its dimensionless counterpart $\mu_{1d}$ in SQM by $\mu_{1d}=-\mu_{2d}/\epsilon_{2}$.
$tt^{*}$ Geometry of the One Electron Model with Two Quasi-Holes
================================================================
We consider a class of $2d$ $\mathcal{N}=(2,2)$ Landau-Ginzburg theory described by the superpotential
$$W(z)= \mu z + \log z + \log(z-\rho).$$
The $4$-SQM version of this theory corresponds to the Vafa model for a single electron on the plane with two quasi-holes at distinct positions $0,\rho \in \mathbb{C}$ and a background electrostatic field $\mu \in \mathbb{C}^{\times}$. The above function is multi-valued and cannot be strictly considered as the superpotential of the model. We want to define this theory on the Abelian universal cover of the target space and study the correspoding $tt^{*}$ geometry. | ua and BPS spectrum, which implies that one can study the geometry of the vacuum bundle either in one or two dimension [ @rif10 ]. The superpotential in SQM is dimensionless and arise from the compactification of the $ 2d$ theory. The matching scale is given by the radious $ R$ of the $ tt^{*}$ lap and we have $ \mathcal{W}_{\mathrm{SQM}}=-\mathcal{W}_{2d}/\epsilon_{2}$. According to the AGT parallelism, the amplitude of the Penner matrix models compute the conformal blocks of the Liouville CFT and hence the FQHE wave function in the Vafa ’s model. To match the standardization of the Liouville correlators given in [ @rif1 ] one has to identify $ -m_{a}/\epsilon_{2}=1 $ and $ 1/\nu=\epsilon_{1}/\epsilon_{2}$. The AGT correspondence has been generalized in [ @rif29 ] also to class $ \mathcal{S}$ theory with irregular punctures in the Gaiotto surface. It turn out that the Nekrasov partition functions of these models reproduce atypical conformal blocks of the Liouville CFT. The bearing of higher degree poles in the quadratic differential corresponds to insertions in the Liouville correlators of the then called irregular vertex operators [ @rif30; @rif31 ]. As for the mass parameters, the relevant pair $ \mu_{2d}$ of the $ 2d$ theory is related to its dimensionless counterpart $ \mu_{1d}$ in SQM by $ \mu_{1d}=-\mu_{2d}/\epsilon_{2}$.
$ tt^{*}$ Geometry of the One Electron Model with Two Quasi - Holes
= = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = =
We consider a class of $ 2d$ $ \mathcal{N}=(2,2)$ Landau - Ginzburg theory described by the superpotential
$ $ W(z)= \mu z + \log z + \log(z-\rho).$$
The $ 4$-SQM version of this theory corresponds to the Vafa model for a individual electron on the plane with two quasi - holes at clear-cut position $ 0,\rho \in \mathbb{C}$ and a background electrostatic field $ \mu \in \mathbb{C}^{\times}$. The above function is multi - prize and cannot be rigorously considered as the superpotential of the model. We want to specify this theory on the Abelian universal cover of the target space and study the correspoding $ tt^{*}$ geometry. | ua wnd BPS spectrum, which ioplies that one can stndy the geometrh of the vacuum bundle eithec in one ir two dimensions [@rif10]. Ghe superiotential in WWM is dimeiaionless and zvises hrom the compacjification ox the $2d$ theory. Tfe matching scale is given by the radyous $R$ pf the $tt^{*}$ circle and re hzne $\mathcal{W}_{\mathrm{SQM}}=-\mathcal{W}_{2d}/\epailon_{2}$. Abcording to the ABT correspondence, the amplltudfs of the Penner mwtrix modelw coipute the conwormal blobns of the Miouville CFT and hence the FQHD wavz functions ib tjg Vafa’s modeo. To iatch the novkalizadion of the Liouville cocrelqtors given in [@rif1] onx has to identify $-m_{a}/\gpsilon_{2}=1$ ang $1/\uu=\epsilon_{1}/\epsilon_{2}$. The QGR cortespotdenzw hxs gexn fenerapizxd in [@rif29] amso to clasw $\mathcal{S}$ theories wynn irregular lunctuwef in the Gaiotto surface. It turns out tvat the Nekrasov partition functions of these mldels repwoduce irregular conformal blocks of the Liouvilla CFT. Ghe pvesevxe of higher degree poles in the quadratic diffqdemtpal corresponds tj insertionx ln jhe Liouville zorrelcfods of the so callef irregolar vwrtex opewatots [@rif30; @rif31]. As for the mass parameters, tke eelevant coupling $\lu_{2d}$ of the $2b$ theoty is telated to its dimensiouless dounterpart $\mu_{1d}$ in SSO by $\mu_{1d}=-\mu_{2d}/\epsilun_{2}$.
$tn^{*}$ Gemmetry of the One Electron Model wivh Twp Quasi-Folex
================================================================
We cogsider a cpass of $2d$ $\mathcal{N}=(2,2)$ Landak-Ginzyurg dheory desfribed by the superpotential
$$W(z)= \mu z + \log z + \lpg(s-\rhm).$$
The $4$-SQM verslon of this thejry correspondx to thz Vafa model for a singne electron on the plane aith two quavi-holes wt dustibct poskgions $0,\rho \in \msthbb{C}$ anb a bqckground electrosbatic rield $\mu \in \matkyb{X}^{\times}$. The abovr fjnceiln if multi-valued and cavmot bd stricujy zonsodered as the superpmtenfial of the model. Ee want to define ehis theory om the Abelian univtrsal rover uf thr twrget space and study the corrsspoding $ht^{*}$ neometry. | ua and BPS spectrum, which implies that study geometry of vacuum bundle either [@rif10]. superpotential in SQM dimensionless and arises the compactification of the $2d$ theory. matching scale is given by the radious $R$ of the $tt^{*}$ circle and have $\mathcal{W}_{\mathrm{SQM}}=-\mathcal{W}_{2d}/\epsilon_{2}$. According to the AGT correspondence, the amplitudes of the Penner matrix compute conformal of Liouville CFT and hence the FQHE wave functions in the Vafa’s model. To match the normalization the Liouville correlators given in [@rif1] one has identify $-m_{a}/\epsilon_{2}=1$ and $1/\nu=\epsilon_{1}/\epsilon_{2}$. AGT correspondence has been generalized [@rif29] to class theories irregular in the Gaiotto It turns out that the Nekrasov partition functions of these models reproduce irregular conformal blocks of the CFT. The higher degree in quadratic corresponds to insertions Liouville correlators of the so called [@rif30; @rif31]. As for the mass parameters, the coupling $\mu_{2d}$ the $2d$ theory is related to dimensionless counterpart $\mu_{1d}$ in SQM by $\mu_{1d}=-\mu_{2d}/\epsilon_{2}$. $tt^{*}$ of the One Electron Model with Two Quasi-Holes ================================================================ We consider a class of $2d$ theory described by the $$W(z)= \mu z \log + The version of theory corresponds to the Vafa model for a single electron on plane with two quasi-holes at distinct positions $0,\rho \in \mathbb{C}$ background field $\mu \in The above function is and be strictly considered as of model. define theory the Abelian universal cover the target space and study correspoding $tt^{*}$ geometry. | ua and BPS spectrum, which implIes that one Can stUdy The GeOmetRy of The vacuum bundlE EithEr in one or two dimensions [@Rif10]. ThE sUPerpOTeNtial In SQM is DImENSioNlEsS anD aRIsEs froM thE compacTification Of tHe $2D$ theory. The maTChIng scale is GivEn by the radioUs $R$ Of the $tT^{*}$ cIrcLE and wE haVe $\matHcal{W}_{\mAThrm{SQm}}=-\mathcal{W}_{2D}/\ePSilon_{2}$. ACCording TO ThE AGT Correspondence, the AMpLItudes of the PenNer matRiX MoDELs cOmpUte the confOrMal blOCks of thE liOUVIllE cFT and hence thE FQHE wave fuNCtiOns in tHe vafA’S model. to matCh THe nOrmalizatioN of tHe LiouvilLe corrELators gIVen in [@riF1] one haS to IdeNtifY $-M_{a}/\EpSilOn_{2}=1$ ANd $1/\nU=\EpSilON_{1}/\epSilon_{2}$. The aGt cOrresPondENCE Has bEen GeneRalizEd in [@rif29] also to ClaSs $\maTHcaL{S}$ theOries With IrRegulAr puncTures In The Gaiotto surfaCe. It Turns out tHat ThE NeKrAsov pARtitioN fuNctIons of tHese modELs rEpRODUcE irregular conformaL bLOCkS of the LiOuvillE cFt. THE presencE oF hiGher DEGree pOles IN tHe quadraTic difFErEnTial corReSponds To InsErtIons iN The LIouvilLe correlAtors OF the so called irREgular vertex oPErATOrS [@Rif30; @rIf31]. AS for the mass ParaMEterS, the RElEvaNT coupLing $\mU_{2d}$ OF tHE $2d$ theory is related to ItS dimenSionlEss counterparT $\mu_{1d}$ in SQM bY $\MU_{1D}=-\mu_{2d}/\epsiLon_{2}$.
$tT^{*}$ geOMetry of the One ELectrOn Model witH two Quasi-holes
================================================================
we considEr a class oF $2D$ $\Mathcal{N}=(2,2)$ lanDau-ginZbuRG ThEory described BY The sUpErpotenTiaL
$$W(z)= \mu z + \lOg z + \Log(Z-\rhO).$$
ThE $4$-SqM version Of this thEoRy CoRrEspOnds tO The Vafa mOdEl fOr A siNgle eLEctron On the PlanE wItH Two Quasi-hoLEs AT DistInCt PosiTioNs $0,\Rho \in \MathBB{C}$ aNd a backGround eleCtrOStatIc FiEld $\mu \in \Mathbb{C}^{\times}$. THe Above functIoN is Multi-vALUed and caNnot be strictly considereD As the suPerPotenTial Of the modeL. We Want to DefINe this Theory On the abEliAN UniveRSAl CovEr Of the targeT SPacE and sTuDy thE corresPoding $tt^{*}$ geometry. | ua and BPS spectrum, which implies t hat o necan s tudy the geometry of t h e va cuum bundle either inone o rt wo d i me nsion s [@rif 1 0] . The s up erp ot e nt ial i n S QM is d imensionle ssan d arises fro m t he compact ifi cation of th e $ 2d$ th eo ry. The m atc hingscalei s give n by thera d ious $ R $ of th e $t t^{* }$ circle and weh av e $\mathcal{W}_ {\math rm { SQ M } }=- \ma thcal{W}_{ 2d }/\ep s ilon_{2 } $. A c cor d ing to the AG T correspon d enc e, the a mpl i tudesof th eP enn er matrix m odel s compute the c o nformal blocksof the Li ouv ille CF Tand h e nce th e F Q HEwave fun ct io ns in the V a f a’smod el.To ma tch the norma liz atio n of theLiouv ille c orrel atorsgiven i n [@rif1] one h as t o identif y $ -m _{a }/ \epsi l on_{2} =1$ an d $1/\n u=\epsi l on_ {1 } / \ ep silon_{2}$. The AG Tc o rr esponden ce has be en generali ze d i n [@ r i f29]also to class $ \mathc a l{ S} $ theor ie s with i rre gul ar pu n ctur es inthe Gaio tto s u rface. It turn s out that the Ne k r as o v pa rti tion functi onso f th esem od els repro duceir r eg u lar conformal block sof the Liou ville CFT. Th e presence o f higherdegr e ep oles in the qu adrat ic differe n tial cor respo nds to i nsertions i n the Li ouv ill e c orr e l at ors of the so c alle dirregul arvertexope rat ors [@ ri f30; @rif 31]. Asfo rth emas s par a meters,th e r el eva nt co u pling$\mu_ {2d} $of the $2d$ t h eo r y isre la tedtoit s dim ensi o nle ss coun terpart $ \mu _ {1d} $in SQM by $\mu_{1d}=-\ mu _{2d}/\eps il on_ {2}$.$ tt^{*}$Geometry of the One Ele c tron Mo del with Two Quasi-Ho les
===== === = ====== ====== ===== == === = = ===== = = == === == ========== = = === ==
W econs ider aclass of $2d$ $\ma t hca l{N}=(2,2)$ L and au-G i n zb urg th e ory d e scr i b ed by the super potential
$ $ W( z)= \mu z+ \l og z + \l og(z-\r ho).$ $
The $ 4$-SQM ve rsion ofth is t h e ory correspon ds to th e Vafa mo d el fo r a sing leelectr on on theplanew ith twoquasi- ho les at dist in ct posit ions $0,\rho \in \mathb b{C}$and a ba ckgroundele c tro static fi eld$\mu \in \ mat hbb {C}^{ \ti m es}$. The ab ove funct ioni s multi-v a lu eda n dcannot be s t r i ctl y con sid e red as the superpotential o f the model. We wan t todef i ne t hi s theory on th e A be l i an unive rs al cover of the tar ge t spac e andstudythe cor r e sp o ding $ tt^{ *}$ geometry . | ua and_BPS spectrum,_which implies that one_can study_the_geometry of_the_vacuum bundle either_in one or_two dimensions [@rif10]. The_superpotential in SQM_is_dimensionless and arises from the compactification of the $2d$ theory. The matching scale is_given_by the_radious_$R$_of the $tt^{*}$ circle and_we have $\mathcal{W}_{\mathrm{SQM}}=-\mathcal{W}_{2d}/\epsilon_{2}$. According to_the AGT_correspondence, the amplitudes of the Penner matrix models_compute_the conformal blocks_of the Liouville CFT and hence the FQHE wave_functions in the Vafa’s model. To_match the normalization_of_the_Liouville correlators given in_[@rif1] one has to identify $-m_{a}/\epsilon_{2}=1$_and $1/\nu=\epsilon_{1}/\epsilon_{2}$. The AGT correspondence has_been generalized in [@rif29] also to class_$\mathcal{S}$ theories with irregular punctures in_the Gaiotto surface. It turns_out that_the Nekrasov partition functions of_these models reproduce_irregular conformal_blocks of the_Liouville CFT. The presence of higher_degree poles in_the quadratic differential corresponds to insertions_in_the Liouville correlators_of_the_so called_irregular vertex operators_[@rif30;_@rif31]. As_for_the mass parameters, the relevant coupling_$\mu_{2d}$_of the $2d$ theory is related to_its dimensionless counterpart $\mu_{1d}$_in_SQM by $\mu_{1d}=-\mu_{2d}/\epsilon_{2}$.
$tt^{*}$ Geometry_of the One Electron Model_with Two Quasi-Holes
================================================================
We consider a class_of $2d$_$\mathcal{N}=(2,2)$ Landau-Ginzburg_theory described by the superpotential
$$W(z)= \mu z + \log z +_\log(z-\rho).$$
The $4$-SQM version of this theory_corresponds to the Vafa_model for_a_single electron on_the_plane with_two quasi-holes at distinct positions $0,\rho \in_\mathbb{C}$ and_a background electrostatic field $\mu \in_\mathbb{C}^{\times}$. The above function_is_multi-valued and cannot be strictly considered_as the superpotential of the model._We want to define this_theory_on_the Abelian universal cover of_the target space and study the_correspoding $tt^{*}$ geometry. |
deriving an asymptotic expansion for the energy functional associated with the equation, which we apply to a suitable family of test functions depending on a real parameter (see for the energy functional and for the family of test functions). To simplify the calculations of curvature terms, we use the conformal normal coordinates introduced by Lee and Parker [@LeePar] and later improved by Cao [@Cao] and Günther [@Gun]. Our proof also crucially relies on the derivation of an expression for the highest-order terms of the GJMS operators (see ), which we obtain by using Juhl’s formulae [@Juhl]. Similarly as in the lower-order cases, we find that the proof essentially reduces to determine the sign of a constant $C{\left(}n,k{\right)}$ depending only on $n$ and $k$, which appears in the energy expansion (see ). In particular, we recover the values found in [@ChenHou; @EspRob] for $C{\left(}n,k{\right)}$ with $k\in\left\{2,3\right\}$. We then conclude the proof by using a general minimization result in the spirit of Aubin [@Aub] (see Mazumdar [@Maz]\*[Theorem 3]{}). We point out that at one place in the proof, namely in the very last computation to determine the sign of $C{\left(}n,k{\right)}$ (see ), we have used the computation software [*Maple*]{} to expand a complicated polynomial with integer coefficients.
Energy expansion and proof of Theorem \[Th\] {#Sec2}
============================================
This section is devoted to the proof of Theorem \[Th\]. Given an integer $k\ge1$ and a smooth positive function $f$ in $M$, we let $I_{k,f}$ be the energy functional defined as $$\label{Eq2}
I_{k,f,g}{\left(}u{\right)}:=\frac{\displaystyle\int_MuP_{2k}u\,dv_g}{\displaystyle{\left(}\int_Mf\left|u\right|^{2^*_k}dv_g{\right)}^{\frac{n-2k}{n}}}$$ for all functions $u\in C^{2k}{\left(}M{\right)}$ such that $u\not\equiv0$. We fix a point $\xi\in M$. | deriving an asymptotic expansion for the energy functional associated with the equation, which we lend oneself to a desirable family of test function count on a real parameter (experience for the department of energy functional and for the family of test functions). To simplify the calculations of curvature terms, we practice the conformal normal coordinates introduced by Lee and Parker [ @LeePar ] and later improved by Cao [ @Cao ] and Günther [ @Gun ]. Our validation also crucially relies on the deriving of an expression for the highest - ordering terms of the GJMS operators (see), which we prevail by using Juhl ’s formulae [ @Juhl ]. Similarly as in the lower - order casing, we find that the proof essentially reduces to determine the signboard of a constant $ C{\left(}n, k{\right)}$ depending only on $ n$ and $ k$, which appears in the energy expansion (determine). In particular, we recover the values found in [ @ChenHou; @EspRob ] for $ C{\left(}n, k{\right)}$ with $ k\in\left\{2,3\right\}$. We then conclude the proof by using a general minimization result in the spirit of Aubin [ @Aub ] (attend Mazumdar [ @Maz]\*[Theorem 3 ] { }). We point out that at one place in the validation, namely in the very final calculation to determine the sign of $ C{\left(}n, k{\right)}$ (determine), we have used the computation software [ * Maple * ] { } to expand a complicated polynomial with integer coefficients.
department of energy expansion and proof of Theorem \[Th\ ] { # Sec2 }
= = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = =
This section is devoted to the proof of Theorem \[Th\ ]. Given an integer $ k\ge1 $ and a smooth positive function $ f$ in $ M$, we let $ I_{k, f}$ be the energy functional defined as $ $ \label{Eq2 }
I_{k, f, g}{\left(}u{\right)}:=\frac{\displaystyle\int_MuP_{2k}u\,dv_g}{\displaystyle{\left(}\int_Mf\left|u\right|^{2^*_k}dv_g{\right)}^{\frac{n-2k}{n}}}$$ for all functions $ u\in C^{2k}{\left(}M{\right)}$ such that $ u\not\equiv0$. We fix a distributor point $ \xi\in M$. | degiving an asymptotic expxnsion for the gnwrgy fnnctionzl assockated with the equation, whici we applt to a suitable family of test vunctionw dekending on a real parametcx (see nor tke energy functipnal and fmr the family mf tzst functions). To simplify the calculwtions pf curvature teris, wt ufe tgv gonformal normal coordinates intdoduced by Lee and Psrker [@LeePar] and later imprlved by Cao [@Cao] and Güntjer [@Gun]. Our kdoos also cruciauly relies on the deribation of an expression for the highzst-order tetma lx the GJMS iperanors (see ), whign we ottain bu using Juhl’s normunae [@Huhl]. Similarly as in vhe lower-order cases, we find dhct the proof essentiaolt redoces do ddreroint tie aign ov a constant $D{\left(}n,k{\righr)}$ depending only on $n$ qnd $k$, which alpears ig the energy expansion (see ). In particuldr, se recover the values fiund in [@ChenHou; @EspRob] for $C{\lefe(}n,k{\right)}$ with $k\in\left\{2,3\right\}$. We then conclude the psoof uy usnkn a ewnfral minimization result in the spirit of Aubyh [@Aib] (see Mazumdar [@Mcz]\*[Theorem 3]{}). We poimt oij that at one klace iu tge proof, namely in the vety lasr computauion yo determine the sign of $C{\lwft(}n,k{\right)}$ (sve ), qe have used the clmputation rofteare [*Kaple*]{} to expand a complncated polynomial with intseer coefficients.
Dneggy axpansion and proof of Thejrem \[Th\] {#Ser2}
============================================
This sectiov is devotqd to the oroof of Theorem \[Th\]. Given an iuteges $k\ge1$ and w smooth positive function $f$ in $M$, we let $I_{k,f}$ ne thv energy yunctipnal defined ws $$\label{Eq2}
I_{k,f,g}{\keft(}u{\riyht)}:=\fraz{\displaystjle\int_MuP_{2n}u\,dv_g}{\displarstyle{\left(}\int_Kv\left|u\right|^{2^*_k}vv_g{\right)}^{\fwac{n-2j}{n}}}$$ fir all wjnctions $u\in C^{2l}{\left(}M{\rigku)}$ such thar $u\not\equiv0$. We fix a kojnt $\xi\in M$. | deriving an asymptotic expansion for the energy with equation, which apply to a depending a real parameter for the energy and for the family of test To simplify the calculations of curvature terms, we use the conformal normal coordinates by Lee and Parker [@LeePar] and later improved by Cao [@Cao] and Günther Our also relies the derivation of an expression for the highest-order terms of the GJMS operators (see ), which obtain by using Juhl’s formulae [@Juhl]. Similarly as the lower-order cases, we that the proof essentially reduces determine sign of constant depending on $n$ and which appears in the energy expansion (see ). In particular, we recover the values found in [@ChenHou; for $C{\left(}n,k{\right)}$ We then the by a general minimization the spirit of Aubin [@Aub] (see We point out that at one place in proof, namely the very last computation to determine sign of $C{\left(}n,k{\right)}$ (see ), we have used computation software [*Maple*]{} to expand a complicated polynomial with integer coefficients. Energy expansion and proof \[Th\] {#Sec2} ============================================ This is devoted to proof Theorem Given integer $k\ge1$ a smooth positive function $f$ in $M$, we let $I_{k,f}$ be energy functional defined as $$\label{Eq2} I_{k,f,g}{\left(}u{\right)}:=\frac{\displaystyle\int_MuP_{2k}u\,dv_g}{\displaystyle{\left(}\int_Mf\left|u\right|^{2^*_k}dv_g{\right)}^{\frac{n-2k}{n}}}$$ for all functions $u\in that We fix a $\xi\in M$. | deriving an asymptotic expanSion for the EnergY fuNctIoNal aSsocIated with the eqUAtioN, which we apply to a suitabLe famIlY Of teST fUnctiOns depeNDiNG On a ReAl ParAmETeR (see fOr tHe energY functionaL anD fOr the family oF TeSt functionS). To Simplify the cAlcUlatioNs Of cURvatuRe tErms, wE use thE ConforMal normal CoORdinatES introdUCEd By LeE and Parker [@LeePar] aND lATer improved by CAo [@Cao] aNd gÜnTHEr [@GUn]. OUr proof alsO cRuciaLLy relieS On THE DerIVation of an expRession for tHE hiGhest-oRdEr tERms of tHe GJMs oPEraTors (see ), whicH we oBtain by usIng JuhL’S formulAE [@Juhl]. SiMilarlY as In tHe loWEr-OrDer CaSEs, wE FiNd tHAt tHe proof eSsEnTiallY redUCES To deTerMine The siGn of a constant $c{\leFt(}n,k{\RIghT)}$ depeNding Only On $N$ and $k$, Which aPpearS iN the energy expanSion (See ). In partIcuLaR, we ReCover THe valuEs fOunD in [@Chenhou; @EspROB] foR $C{\LEFT(}n,K{\right)}$ with $k\in\left\{2,3\rIgHT\}$. we Then concLude thE PrOoF By using a GeNerAl miNIMizatIon rESuLt in the sPirit oF auBiN [@Aub] (see maZumdar [@maZ]\*[ThEorEm 3]{}). We pOInt oUt that At one plaCe in tHE proof, namely in THe very last comPUtATIoN To deTerMine the sign Of $C{\lEFt(}n,k{\RighT)}$ (SeE ), we HAve usEd the CoMPuTAtion software [*Maple*]{} tO eXpand a ComplIcated polynomIal with intEGER coefficIentS.
enERgy expansion anD prooF of Theorem \[tH\] {#Sec2}
============================================
This SectiOn is devoTed to the pROOf of TheoRem \[th\]. GIveN an INTeGer $k\ge1$ and a smoOTH posItIve funcTioN $f$ in $M$, we Let $i_{k,f}$ Be tHe eNeRgy functiOnal defiNeD aS $$\lAbEl{EQ2}
I_{k,f,g}{\LEft(}u{\righT)}:=\fRac{\DiSplAystyLE\int_Mup_{2k}u\,dv_G}{\disPlAySTylE{\left(}\inT_mf\LEFt|u\rIgHt|^{2^*_K}dv_g{\RigHt)}^{\Frac{n-2K}{n}}}$$ foR All FunctioNs $u\in C^{2k}{\leFt(}M{\RIght)}$ SuCh That $u\noT\equiv0$. We fix a pOiNt $\xi\in M$. | deriving an asymptotic ex pansion fo r the en erg yfunc tion al associatedw iththe equation, which we appl yt o as ui table family of t est f un cti on s d epend ing on a r eal parame ter ( see for thee ne rgy functi ona l and for th e f amilyof te s t fun cti ons). To si m plifythe calcu la t ions o f curvat u r eterm s, we use the con f or m al normal coor dinate si nt r o duc edby Lee and P arker [@LeePa r ]a n d la t er improved b y Cao [@Cao ] an d Günt he r [ @ Gun].Our p ro o f a lso crucial ly r elies onthe de r ivation of an e xpress ion fo r th e h ig hes t- o rde r t erm s of the GJM Sop erato rs ( s e e ), w hic h we obta in by using J uhl ’s f o rmu lae [ @Juhl ]. S im ilarl y as i n the l ower-order case s, w e find th atth e p ro of es s ential lyred uces to determ i neth e s ig n of a constant $C {\ l e ft (}n,k{\r ight)} $ d ep e nding on ly on $n$ a nd $k $, w h ic h appear s in t h een ergy ex pa nsion(s ee).In pa r ticu lar, w e recove r the values found i n [@ChenHou; @ E sp R o b] for$C{ \left(}n,k{ \rig h t)}$ wit h $ k\i n \left \{2,3 \r i gh t \}$. We then conclu de the p roofby using a ge neral mini m i z ation re sult in the spirit ofAubin [@Aub] (s e e Mazumd ar [@ Maz]\*[T heorem 3] { } ). We po int ou t t hat a tone place int h e pr oo f, name lyin thever y l ast co mp utation t o determ in eth esig n of$ C{\left( }n ,k{ \r igh t)}$( see ), we h aveus ed the comput a ti o n sof tw ar e [* Map le *]{}to e x pan d a com plicatedpol y nomi al w ith int eger coeffici en ts.
Energ yexp ansion a nd proof of Theorem \[Th\] {#Se c 2}
==== === ===== ==== ========= === ====== === = ======
This sect io n i s devot e d t o t he proof ofT h eor em \[ Th \].Given a n integer $k\ge1$a nda smooth posi tiv e fu n c ti on$ f$ in$M $ , w e let $I_{k,f}$ b e the ener gy fu nctional d e fin ed as $$\ label{E q2}
I _ {k,f,g} {\left(}u {\right)} := \fra c { \di splaystyle \int_MuP _{2k}u\,d v _g}{\ d is plays tyl e{\lef t( }\i nt_Mf \left| u \ri ght|^ {2^*_k }d v_g{\r ight) }^ {\frac{n -2k}{n}}}$$ for all fun ctions $u\i n C ^{2k}{\le ft( } M{\ right)}$such that $u\n ot\ equ iv0$. We fix a poi n t$\x i \in M $. | deriving_an asymptotic_expansion for the energy_functional associated_with_the equation,_which_we apply to_a suitable family_of test functions depending_on a real_parameter_(see for the energy functional and for the family of test functions). To simplify_the_calculations of_curvature_terms,_we use the conformal normal_coordinates introduced by Lee and_Parker [@LeePar] and_later improved by Cao [@Cao] and Günther [@Gun]. Our proof_also_crucially relies on_the derivation of an expression for the highest-order terms_of the GJMS operators (see ),_which we obtain_by_using_Juhl’s formulae [@Juhl]. Similarly as_in the lower-order cases, we find_that the proof essentially reduces to_determine the sign of a constant $C{\left(}n,k{\right)}$_depending only on $n$ and $k$,_which appears in the energy_expansion (see_). In particular, we recover_the values found_in [@ChenHou; @EspRob]_for $C{\left(}n,k{\right)}$ with_$k\in\left\{2,3\right\}$. We then conclude the proof_by using a_general minimization result in the spirit_of_Aubin [@Aub] (see Mazumdar [@Maz]\*[Theorem 3]{})._We_point_out that_at one place_in_the proof,_namely_in the very last computation to_determine_the sign of $C{\left(}n,k{\right)}$ (see ), we_have used the computation_software_[*Maple*]{} to expand a_complicated polynomial with integer coefficients.
Energy_expansion and proof of Theorem \[Th\] {#Sec2}
============================================
This_section is_devoted to_the proof of Theorem \[Th\]. Given an integer $k\ge1$ and a smooth_positive function $f$ in $M$, we_let $I_{k,f}$ be the_energy functional_defined_as $$\label{Eq2}
I_{k,f,g}{\left(}u{\right)}:=\frac{\displaystyle\int_MuP_{2k}u\,dv_g}{\displaystyle{\left(}\int_Mf\left|u\right|^{2^*_k}dv_g{\right)}^{\frac{n-2k}{n}}}$$ for_all_functions $u\in_C^{2k}{\left(}M{\right)}$ such that $u\not\equiv0$. We fix a_point $\xi\in_M$. |
- 'A. Kappes'
- 'T. Karg'
- 'A. Karle'
- 'M. Kauer'
- 'J. L. Kelley'
- 'A. Kheirandish'
- 'J. Kiryluk'
- 'J. Kläs'
- 'S. R. Klein'
- 'J.-H. Köhne'
- 'G. Kohnen'
- 'H. Kolanoski'
- 'A. Koob'
- 'L. Köpke'
- 'C. Kopper'
- 'S. Kopper'
- 'D. J. Koskinen'
- 'M. Kowalski'
- 'A. Kriesten'
- 'K. Krings'
- 'G. Kroll'
- 'M. Kroll'
- 'J. Kunnen'
- 'N. Kurahashi'
- 'T. Kuwabara'
- 'M. Labare'
- 'D. T. Larsen'
- 'M. J. Larson'
- 'M. Lesiak-Bzdak'
- 'M. Leuermann'
- 'J. Leute'
- 'J. Lünemann'
- 'O. Mac[í]{}as'
- 'J. Madsen'
- 'G. Maggi'
- 'R. Maruyama'
- 'K. Mase'
- 'H. S. Matis'
- 'F. McNally'
- 'K. Meagher'
- 'M. Medici'
- 'A. Meli'
- 'T. Meures'
- 'S. Miarecki'
- 'E. Middell'
- 'E. Middlemas'
- 'N. Milke'
- 'J. Miller'
- 'L. Mohrmann'
- 'T. Montaruli' | -' A. Kappes'
-' T. Karg'
-' A. Karle'
-' M. Kauer'
-' J. L. Kelley'
-' A. Kheirandish'
-' J. Kiryluk'
-' J. Kläs'
-' S. R. Klein'
-' J.-H. Köhne'
-' G. Kohnen'
-' H. Kolanoski'
-' A. Koob'
-' L. Köpke'
-' C. Kopper'
-' S. Kopper'
-' D. J. Koskinen'
-' M. Kowalski'
-' A. Kriesten'
-' K. Krings'
-' G. Kroll'
-' M. Kroll'
-' J. Kunnen'
-' N. Kurahashi'
-' T. Kuwabara'
-' M. Labare'
-' D. T. Larsen'
-' M. J. Larson'
-' M. Lesiak - Bzdak'
-' M. Leuermann'
-' J. Leute'
-' J. Lünemann'
-' O. Mac[í]{}as'
-' J. Madsen'
-' G. Maggi'
-' R. Maruyama'
-' K. Mase'
-' H. S. Matis'
-' F. McNally'
-' K. Meagher'
-' M. Medici'
-' A. Meli'
-' T. Meures'
-' S. Miarecki'
-' E. Middell'
-' E. Middlemas'
-' N. Milke'
-' J. Miller'
-' L. Mohrmann'
-' T. Montaruli' |
- 'A. Kwppes'
- 'T. Karg'
- 'A. Karle'
- 'M. Kauev'
- 'J. L. Kelley'
- 'A. Kheitabdish'
- 'O. Kirylui'
- 'J. Kläs'
- 'S. F. Klein'
- 'J.-H. Köhne'
- 'G. Kohnen'
- 'H. Kolanlsji'
- 'A. Kiob'
- 'L. Köpke'
- 'C. Kopper'
- 'S. Kopoer'
- 'D. J. Kosninen'
- 'M. Kiwalwji'
- 'A. Kriestxh'
- 'K. Krinnf'
- 'G. Kdlll'
- 'K. Jroll'
- 'J. Kunnen'
- 'K. Kurahashi'
- 'D. Kuwabara'
- 'M. Labdrd'
- 'B. T. Larsen'
- 'M. J. Larson'
- 'M. Lesiak-Bzdak'
- 'M. Leuewmann'
- 'J. Kekte'
- 'J. Lünemann'
- 'O. Iac[í]{}ax'
- 'J. Massen'
- 'G. Maggi'
- 'R. Maruyama'
- 'K. Mase'
- 'H. S. Matia'
- 'F. McNanly'
- 'K. Meagher'
- 'K. Medici'
- 'A. Meli'
- 'T. Meures'
- 'S. Miageckl'
- 'E. Middell'
- 'E. Middlelas'
- 'N. Milke'
- 'H. Miljwr'
- 'L. Mohrmann'
- 'T. Montarulp' | - 'A. Kappes' - 'T. Karg' - - Kauer' - L. Kelley' - - Kläs' - 'S. Klein' - 'J.-H. - 'G. Kohnen' - 'H. Kolanoski' 'A. Koob' - 'L. Köpke' - 'C. Kopper' - 'S. Kopper' - 'D. Koskinen' - 'M. Kowalski' - 'A. Kriesten' - 'K. Krings' - 'G. Kroll' 'M. - Kunnen' 'N. Kurahashi' - 'T. Kuwabara' - 'M. Labare' - 'D. T. Larsen' - 'M. J. Larson' 'M. Lesiak-Bzdak' - 'M. Leuermann' - 'J. Leute' 'J. Lünemann' - 'O. - 'J. Madsen' - 'G. - Maruyama' - Mase' 'H. Matis' - 'F. - 'K. Meagher' - 'M. Medici' - 'A. Meli' - 'T. Meures' - 'S. Miarecki' - 'E. - 'E. 'N. Milke' 'J. - Mohrmann' - 'T. |
- 'A. Kappes'
- 'T. Karg'
- 'A. Karle'
- 'M. Kauer'
- 'J. l. Kelley'
- 'A. KhEiranDisH'
- 'J. KIrYluk'
- 'j. KläS'
- 'S. R. Klein'
- 'J.-H. KöhnE'
- 'g. KohNen'
- 'H. Kolanoski'
- 'A. Koob'
- 'L. KöpKe'
- 'C. KoPpER'
- 'S. KoPPeR'
- 'D. J. KoSkinen'
- 'M. kOwALSki'
- 'a. KRiEstEn'
- 'k. krIngs'
- 'G. kroLl'
- 'M. KrolL'
- 'J. Kunnen'
- 'N. KUraHaShi'
- 'T. Kuwabara'
- 'm. laBare'
- 'D. T. LarsEn'
- 'M. j. Larson'
- 'M. LesiAk-BZdak'
- 'M. LEuErmANn'
- 'J. LeUte'
- 'j. LüneMann'
- 'O. MAC[í]{}as'
- 'J. MAdsen'
- 'G. MagGi'
- 'r. maruyaMA'
- 'K. Mase'
- 'H. s. mAtIs'
- 'F. MCNally'
- 'K. Meagher'
- 'M. MeDIcI'
- 'a. Meli'
- 'T. Meures'
- 'S. MIareckI'
- 'E. mIdDELl'
- 'E. midDlemas'
- 'N. MilKe'
- 'j. MillER'
- 'L. MohrmANn'
- 't. mONtaRUli' |
- 'A. Kappes'
- 'T. Karg'
- 'A. Kar le'
- 'M . K au er'- 'J . L. Kelley'
- 'A.Kheirandish'
- 'J. Kir yluk'
- 'J.K lä s'
-'S. R.K le i n '
- ' J. -H. K ö hn e'
-'G. Kohnen '
- 'H. Ko lan os ki'
- 'A. Ko o b'
- 'L. Köp ke'
- 'C. Koppe r'- 'S.Ko ppe r '
- ' D.J. Ko skinen '
- 'M. Kowalski '- 'A. K r iesten' - ' K. K rings'
- 'G. Krol l '- 'M. Kroll'
-'J. Ku nn e n' - 'N . K urahashi'-'T. K u wabara' -' M . La b are'
- 'D. T. Larsen'
-' M.J. Lar so n'- 'M. L esiak -B z dak '
- 'M. Leu erma nn'
- 'J. Leute '
- 'J.L ünemann '
- 'O . M ac[ í]{} a s'
- 'J .M ads e n'
-' G.Maggi'
- ' R. Maru yama ' - 'K.Mas e'
- 'H.S. Matis'
- ' F.McNa l ly'
- 'K . Mea gher '- 'M. Medic i'
-'A . Meli'
- 'T. M eure s'
- 'S.Mia re cki '- 'E. Middel l'- ' E. Midd lemas'- 'N .M i l ke '
- 'J. Miller'
-'L . Mo hrmann'- 'T.M on ta r uli' |
- 'A. Kappes'
-_'T. Karg'
- 'A. Karle'
-_'M. Kauer'
- 'J. L. Kelley'
- 'A. Kheirandish'
- 'J. Kiryluk'
-_'J. Kläs'
- 'S. R. Klein'
-_'J.-H. Köhne'
-_'G. Kohnen'
- 'H. Kolanoski'
-_'A. Koob'
-_'L. Köpke'
- 'C. Kopper'
- 'S. Kopper'
-_'D. J. Koskinen'
- 'M. Kowalski'
- 'A. Kriesten'
-_'K. Krings'
- 'G. Kroll'
- 'M. Kroll'
- 'J. Kunnen'
-_'N. Kurahashi'
- 'T. Kuwabara'
- 'M. Labare'
-_'D. T. Larsen'
-_'M. J. Larson'
- 'M. Lesiak-Bzdak'
- 'M. Leuermann'
- 'J. Leute'
- 'J. Lünemann'
- 'O. Mac[í]{}as'
- 'J. Madsen'
- 'G. Maggi'
- 'R. Maruyama'
- 'K. Mase'
- 'H. S. Matis'
- 'F. McNally'
- 'K. Meagher'
- 'M. Medici'
-_'A. Meli'
-_'T. Meures'
- 'S. Miarecki'
-_'E. Middell'
-_'E. Middlemas'
-_'N. Milke'
- 'J. Miller'
- 'L. Mohrmann'
- 'T. Montaruli' |
strength of the interaction.]{}
Given that a squirmer travels its own size in a time $\tau = \frac{\sigma}{v_s}$, we perform simulations from $1450$ up to $3000$ $\tau$. Once the system reaches steady state, in a time window between $1000$ and $2000$ $\tau$, we carry out a systematic analysis of the dynamics of the squirmer suspension, considering $\xi$ from 0.1 to 10 and $\beta$ from -3 up to 3.
\[sec:Tools\]Analysis Tools
===========================
In order to be able to establish the effect played by the anisotropy of the interaction between swimmers, we follow the procedure in Ref. [@SoftMatter17] and study the degree of aggregation as a function of $\beta$ and $\xi$. Whenever the system forms a steady-state cluster distribution, we explore the clusters morphology by computing their mean size, the cluster-size distribution, their radius of gyration, and both the polar and nematic order parameter. When clusters are present, we identify them following a distance criterion: at each time step, two particles belong to the same cluster of size $s$ whenever their distance is smaller than $r_{cl} = 1.75 \sigma$ (first minimum of the $g(r)$, see appendix \[Appendix:Clusters\]).
To calculate the cluster-size distribution $f(s)$ we apply a criterion based on [@chantalbins2012]: (i) We arbitrarily subdivide the range of $s$-values into intervals $\Delta s_i = (s_{i,max} - s_{i,min})$, where $n_i^t$ is the total number of clusters within each interval $\Delta s$; (ii) we assign the value $n_i=n^t_i/\Delta s_i$ to every $s$ within $\Delta s_i$, and compute the fraction of clusters of size $s$ as $f(s)=n_i/N_c$, where $N_c=\sum_i n_i \Delta s_i$ is the total number of clusters.
To characterise the clusters’ morphology, we compute the radius of gyration $$\label{eq:Rg}
R_g(s) = \sqrt{\sum_{i,j=1}^s \frac{\left(\vec{r}_i-\vec{r}_j\right)^2 | strength of the interaction. ] { }
Given that a squirmer travels its own size in a meter $ \tau = \frac{\sigma}{v_s}$, we do simulations from $ 1450 $ up to $ 3000 $ $ \tau$. Once the system reach firm state, in a time windowpane between $ 1000 $ and $ 2000 $ $ \tau$, we carry out a systematic analysis of the moral force of the wiggler suspension, considering $ \xi$ from 0.1 to 10 and $ \beta$ from -3 up to 3.
\[sec: Tools\]Analysis cock
= = = = = = = = = = = = = = = = = = = = = = = = = = =
In order to be able to establish the effect play by the anisotropy of the interaction between swimmers, we follow the procedure in Ref. [ @SoftMatter17 ] and analyze the degree of aggregation as a function of $ \beta$ and $ \xi$. Whenever the organization forms a steady - state cluster distribution, we research the clusters morphology by calculate their mean size, the cluster - size distribution, their radius of gyration, and both the arctic and nematic order parameter. When clusters are present, we identify them following a distance standard: at each clock time step, two particles belong to the same cluster of size $ s$ whenever their distance is smaller than $ r_{cl } = 1.75 \sigma$ (first minimum of the $ g(r)$, see appendix \[Appendix: Clusters\ ]).
To calculate the bunch - size distribution $ f(s)$ we use a criterion free-base on [ @chantalbins2012 ]: (i) We arbitrarily subdivide the range of $ s$-values into intervals $ \Delta s_i = (s_{i, max } - s_{i, min})$, where $ n_i^t$ is the entire number of clusters within each interval $ \Delta s$; (ii) we assign the value $ n_i = n^t_i/\Delta s_i$ to every $ s$ within $ \Delta s_i$, and calculate the fraction of clusters of size $ s$ as $ f(s)=n_i / N_c$, where $ N_c=\sum_i n_i \Delta s_i$ is the total number of clusters.
To characterise the clusters ’ morphology, we compute the radius of gyration $ $ \label{eq: Rg }
R_g(s) = \sqrt{\sum_{i, j=1}^s \frac{\left(\vec{r}_i-\vec{r}_j\right)^2 | stgength of the interactiok.]{}
Given that a sqoiemer tcavels jts own rize in a time $\tau = \frac{\sigme}{v_s}$, qe peeform simulations from $1450$ up to $3000$ $\nau$. Once rhe wtstem reaciss steady stafc, in c vime window beteeen $1000$ and $2000$ $\tau$, we carry mug c systematic analysis of the dynamicf of thr dquirmer suspegsiom, conapdtring $\xi$ from 0.1 to 10 and $\beta$ from -3 up to 3.
\[vec:Tools\]Analyxis Tools
===========================
In order to be abpe tl establish the efvect played by eye anisotroph of the interaction bgtween swimmers, we follow the pruceduxe in Ref. [@SodtNathgr17] and study the qegree of agngegatiot as a gunction of $\beba$ anv $\xi$. Whenever the system horms a steady-state sluster dhscribution, we explore rhw cluvterv mofphouogg uy domputlng their meah size, the xluster-size distribityin, their radihs of dywation, and both the polar and nematic ogder parameter. When clusterw are present, we idenjify them sollowing a distance criterion: at each time step, dwo pergicoef vepong to the same cluster of size $s$ whenever trsit cistance is smcller than $r_{cl} = 1.75 \slgkw$ (first minimom of tks $f(r)$, see appendix \[Apoendix:Cjustees\]).
To calctlatr the cluster-size distributuon $f(s)$ we apily q criterion based ln [@chantalbnns2012]: (i) Ee arnitrarily subdivide the ranfe of $s$-valufs into ihgervals $\Delta s_i = (s_{p,max} - s_{i,min})$, where $n_i^t$ is the tjtal numbxr of clustefs wothin qach intergal $\Dcnta s$; (ii) we assign the ralue $n_i=n^t_i/\Deltw s_i$ to every $s$ within $\Delta s_i$, and compute tne frdction oy clusbers of size $s$ ws $f(s)=n_i/N_c$, wherg $N_c=\sum_i u_i \Delga s_i$ is tge totan number of clusters.
To cvwracterise tie clustews’ mirphilogy, wd compute the rsdius of yvration $$\lqbel{eq:Rg}
R_g(s) = \sqrt{\smm_{i,j=1}^s \rrac{\left(\vec{r}_i-\vee{x}_j\tught)^2 | strength of the interaction.]{} Given that a its size in time $\tau = $1450$ to $3000$ $\tau$. the system reaches state, in a time window between and $2000$ $\tau$, we carry out a systematic analysis of the dynamics of squirmer suspension, considering $\xi$ from 0.1 to 10 and $\beta$ from -3 up 3. Tools In to be able to establish the effect played by the anisotropy of the interaction between swimmers, follow the procedure in Ref. [@SoftMatter17] and study degree of aggregation as function of $\beta$ and $\xi$. the forms a cluster we the clusters morphology computing their mean size, the cluster-size distribution, their radius of gyration, and both the polar and nematic parameter. When present, we them a criterion: at each two particles belong to the same $s$ whenever their distance is smaller than $r_{cl} 1.75 \sigma$ minimum of the $g(r)$, see appendix To calculate the cluster-size distribution $f(s)$ we apply criterion based on [@chantalbins2012]: (i) We arbitrarily subdivide the range of $s$-values into intervals $\Delta (s_{i,max} - s_{i,min})$, where is the total of within interval s$; (ii) assign the value $n_i=n^t_i/\Delta s_i$ to every $s$ within $\Delta s_i$, compute the fraction of clusters of size $s$ as $f(s)=n_i/N_c$, n_i s_i$ is the number of clusters. To the morphology, we compute the gyration R_g(s) | strength of the interaction.]{}
GIven that a sQuirmEr tRavElS its Own sIze in a time $\tau = \fRAc{\siGma}{v_s}$, we perform simulatiOns frOm $1450$ UP to $3000$ $\tAU$. ONce thE system REaCHEs sTeAdY stAtE, In A time WinDow betwEen $1000$ and $2000$ $\tau$, wE caRrY out a systemaTIc Analysis of The Dynamics of thE sqUirmer SuSpeNSion, cOnsIderiNg $\xi$ frOM 0.1 to 10 and $\Beta$ from -3 uP tO 3.
\[Sec:TooLS\]AnalysIS toOls
===========================
IN order to be able to eSTaBLish the effect pLayed bY tHE aNISotRopY of the inteRaCtion BEtween sWImMERS, we FOllow the proceDure in Ref. [@SoFTMaTter17] anD sTudY The degRee of AgGRegAtion as a funCtioN of $\beta$ anD $\xi$. WheNEver the SYstem foRms a stEadY-stAte cLUsTeR diStRIbuTIoN, we EXplOre the clUsTeRs morPholOGY BY comPutIng tHeir mEan size, the cluSteR-sizE DisTribuTion, tHeir RaDius oF gyratIon, anD bOth the polar and nEmatIc order paRamEtEr. WHeN clusTErs are PreSenT, we idenTify theM FolLoWING a Distance criterion: aT eACH tIme step, tWo partIClEs BElong to tHe SamE cluSTEr of sIze $s$ WHeNever theIr distANcE iS smalleR tHan $r_{cl} = 1.75 \SiGma$ (FirSt minIMum oF the $g(r)$, See appenDix \[ApPEndix:Clusters\]).
TO Calculate the cLUsTER-sIZe diStrIbution $f(s)$ we ApplY A criTeriON bAseD On [@chaNtalbInS2012]: (I) WE Arbitrarily subdividE tHe rangE of $s$-vAlues into inteRvals $\Delta S_I = (S_{I,max} - s_{i,miN})$, wheRE $n_I^T$ is the total numBer of Clusters wiTHin each iNtervAl $\Delta s$; (Ii) we assigN THe value $n_I=n^t_I/\DeLta S_i$ tO EVeRy $s$ within $\DeltA S_I$, and CoMpute thE frAction oF clUstErs Of sIzE $s$ as $f(s)=n_i/N_C$, where $N_c=\SuM_i N_i \deLta S_i$ is tHE total nuMbEr oF cLusTers.
TO CharacTerisE the ClUsTErs’ MorpholOGy, WE CompUtE tHe raDiuS oF gyraTion $$\LAbeL{eq:Rg}
R_g(S) = \sqrt{\sum_{i,J=1}^s \fRAc{\leFt(\VeC{r}_i-\vec{r}_J\right)^2 | strength of the interacti on.]{}
Gi ven t hat asq uirm er t ravels its own size in a time $\tau = \fr ac{\s ig m a}{v _ s} $, we perfor m s i m ula ti on s f ro m $ 1450$ up to $30 00$ $\tau$ . O nc e the system re aches stea dystate, in atim e wind ow be t ween$10 00$ a nd $20 0 0$ $\t au$, we c ar r y outa system a t ic ana lysis of the dyna m ic s of the squirm er sus pe n si o n , c ons idering $\ xi $ fro m 0.1 to 10 a n d $ \ beta$ from -3 up to 3.
\ [se c:Tool s\ ]An a lysisTools
= = === =========== ==== =======
In ord e r to be able to estab lis h t he e f fe ct pl ay e d b y t hea nis otropy o fth e int erac t i o n bet wee n sw immer s, we followthe pro c edu re in Ref. [@S of tMatt er17]and s tu dy the degree o f ag gregation as a fu nc tiono f $\be ta$ an d $\xi$ . Whene v erth e s ys tem forms a steady -s t a te cluster distr i bu ti o n, we ex pl ore the c luste rs m o rp hology b y comp u ti ng theirme an siz e, th e c luste r -siz e dist ribution , the i r radius of gy r ation, and bo t ht h ep olar an d nematic o rder para mete r .Whe n clus tersar e p r esent, we identifyth em fol lowin g a distancecriterion: a t each ti me s t ep , two particles belo ng to thes ame clus ter o f size $ s$ whenev e r their d ist anc e i s s m a ll er than $r_{c l } = 1 .7 5 \sigm a$(firstmin imu m o f t he $g(r)$,see appe nd ix \ [A ppe ndix: C lusters\ ]) .
To ca lcula t e theclust er-s iz ed ist ributio n $ f ( s)$we a pply acr iteri on b a sed on [@c hantalbin s20 1 2]:(i )We arbi trarily subdi vi de the ran ge of $s$-v a l ues into intervals $\Delta s_i= (s_{i, max } - s _{i, min})$, w her e $n_i ^t$ is the total numb er of c luste r s w ith in each inte r v al$\Del ta s$; (ii) w e assign the value $n_ i=n^t_i/\Delt a s _i$t o e ver y $ s $ w it h in$ \ Delta s_i$, and compute t he fr action ofc lus te rs of s ize $s$ as $ f (s)=n_i /N_c$, wh ere $N_c= \s um_i n _i\Delta s_i $ is the total nu m ber o f c luste rs.
To c ha rac teris e thec lus ters’ morph ol ogy, w e com pu te the r adius of gyration $$\la bel{eq :Rg}R_g (s) = \sq rt{ \ sum _{i,j=1}^ s \f rac{\left( \ve c{r }_i-\ vec { r}_j\ righ t )^ 2 | strength_of the_interaction.]{}
Given that a squirmer_travels its_own_size in_a_time $\tau =_\frac{\sigma}{v_s}$, we perform_simulations from $1450$ up_to $3000$ $\tau$._Once_the system reaches steady state, in a time window between $1000$ and $2000$ $\tau$,_we_carry out_a_systematic_analysis of the dynamics of_the squirmer suspension, considering $\xi$_from 0.1_to 10 and $\beta$ from -3 up to_3.
\[sec:Tools\]Analysis_Tools
===========================
In order to_be able to establish the effect played by the_anisotropy of the interaction between swimmers,_we follow the_procedure_in_Ref. [@SoftMatter17] and study the_degree of aggregation as a function_of $\beta$ and $\xi$. Whenever the_system forms a steady-state cluster distribution, we_explore the clusters morphology by computing_their mean size, the cluster-size_distribution, their_radius of gyration, and both_the polar and_nematic order_parameter. When clusters_are present, we identify them following_a distance criterion:_at each time step, two particles_belong_to the same_cluster_of_size $s$_whenever their distance_is_smaller than_$r_{cl}_= 1.75 \sigma$ (first minimum of_the_$g(r)$, see appendix \[Appendix:Clusters\]).
To calculate the cluster-size_distribution $f(s)$ we apply_a_criterion based on [@chantalbins2012]:_(i) We arbitrarily subdivide the_range of $s$-values into intervals $\Delta_s_i =_(s_{i,max} -_s_{i,min})$, where $n_i^t$ is the total number of clusters within each_interval $\Delta s$; (ii) we assign_the value $n_i=n^t_i/\Delta s_i$_to every_$s$_within $\Delta s_i$,_and_compute the_fraction of clusters of size $s$ as_$f(s)=n_i/N_c$, where_$N_c=\sum_i n_i \Delta s_i$ is the_total number of clusters.
To_characterise_the clusters’ morphology, we compute the_radius of gyration $$\label{eq:Rg}
R_g(s) = \sqrt{\sum_{i,j=1}^s_\frac{\left(\vec{r}_i-\vec{r}_j\right)^2 |