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
|
|---|---|---|---|---|---|---|
a})
[u_i(\vec{a})]-
\delta\sum_{t=0}^{\lceil n/\delta\rceil} (1-\delta)^t
\sum_{h \in H^{t}, \vec{a} \in A} \rho_{\vec{M}}^{t+1}(h \cdot \vec{a})
[u_i(\vec{a})]\\
= &\delta\sum_{t=\lceil n/\delta\rceil+1}^{\infty} (1-\delta)^t
\sum_{h \in H^{t}, \vec{a} \in A} \rho_{\vec{M}}^{t+1}(h \cdot \vec{a})
[u_i(\vec{a})]\\
\leq& \delta\sum_{t=\lceil n/\delta\rceil}^{\infty} (1-\delta)^t a\\
= &\delta(1-\delta)^{\lceil n/\delta\rceil}\sum_{t=0}^{\infty}(1-\delta)^ta
=\delta(1-\delta)^{\lceil n/\delta\rceil}\frac{a}{\delta}\leq \frac{a}{e^n}.
\end{array}$$
### The $\epsilon$-NE strategy and the algorithm {#sec:computingNE}
Let be a non-empty set and let.[^6] A player can broadcast an $m$-bit string by using his actions for $m$ rounds, by treating actions from $A_i^0$ as 0 and actions from $A_i^1$ as 1. Given a polynomial $\phi$ (with natural coefficients), let $({{\text{Gen}}},{{\text{Enc}}},{{\text{Dec}}})$ be a multi-message multi-key secure $\phi$-bit, if the security parameter is $k$, the length of an encrypted message is $z(k)$ for some polynomial $z$. Let ${\mathit{sq}}=(s_1,s_2\ldots,s_m)$ be a fixed sequence of action profiles. Fix a polynomial-time pseudorandom function ensemble $\{{\mathit{PS}}_s: s \in
\{0,1\}^*\}$. For a game $G$ such that $|G|=n$, consider the strategy $\sigma^{{\mathit{NE}}}$ for player $i$ in $G^{\infty}(\delta)$ that has the following three phases. Phase 1 explains what to do if | a })
[ u_i(\vec{a})]-
\delta\sum_{t=0}^{\lceil n/\delta\rceil } (1-\delta)^t
\sum_{h \in H^{t }, \vec{a } \in A } \rho_{\vec{M}}^{t+1}(h \cdot \vec{a })
[ u_i(\vec{a})]\\
= & \delta\sum_{t=\lceil n/\delta\rceil+1}^{\infty } (1-\delta)^t
\sum_{h \in H^{t }, \vec{a } \in A } \rho_{\vec{M}}^{t+1}(h \cdot \vec{a })
[ u_i(\vec{a})]\\
\leq & \delta\sum_{t=\lceil n/\delta\rceil}^{\infty } (1-\delta)^t a\\
= & \delta(1-\delta)^{\lceil n/\delta\rceil}\sum_{t=0}^{\infty}(1-\delta)^ta
= \delta(1-\delta)^{\lceil n/\delta\rceil}\frac{a}{\delta}\leq \frac{a}{e^n }.
\end{array}$$
# # # The $ \epsilon$-NE strategy and the algorithm { # sec: computingNE }
Let be a non - empty set and let.[^6 ] A player can air an $ m$-bit chain by using his actions for $ m$ rounds, by process actions from $ A_i^0 $ as 0 and actions from $ A_i^1 $ as 1. give a polynomial $ \phi$ (with natural coefficients), permit $ ({ { \text{Gen}}},{{\text{Enc}}},{{\text{Dec}}})$ be a multi - message multi - key secure $ \phi$-bit, if the security system parameter is $ k$, the length of an encrypted message is $ z(k)$ for some polynomial $ z$. Let $ { \mathit{sq}}=(s_1,s_2\ldots, s_m)$ embody a fixed succession of action profile. Fix a polynomial - time pseudorandom function ensemble $ \{{\mathit{PS}}_s: s \in
\{0,1\}^*\}$. For a game $ G$ such that $ |G|=n$, regard the strategy $ \sigma^{{\mathit{NE}}}$ for player $ i$ in $ G^{\infty}(\delta)$ that has the following three phase. Phase 1 explains what to do if | a})
[u_i(\gec{a})]-
\delta\sum_{t=0}^{\lceil n/\deltx\rceil} (1-\delta)^t
\sum_{h \in H^{v}, \vec{a} \jn A} \rho_{\xec{M}}^{t+1}(h \cdot \vec{a})
[u_i(\vec{a})]\\
= &\delta\dun_{t=\lceul n/\delta\rceil+1}^{\infty} (1-\deuta)^t
\sum_{h \pn H^{t}, \vec{q} \in Q} \rho_{\vec{M}}^{t+1}(i \cdot \vce{a})
[u_i(\vsg{a})]\\
\leq& \velta\sum_{t=\lceil k/\delta\rceil}^{\hnfty} (1-\delta)^t a\\
= &\ddlca(1-\delta)^{\lceil n/\delta\rceil}\sum_{t=0}^{\infty}(1-\delea)^ta
=\delya(1-\felta)^{\lceil n/\dejta\rbeyl}\frzb{a}{\belta}\leq \frac{a}{e^n}.
\end{array}$$
### The $\epailon$-NE strategy and the algorithm {#sec:computinhNE}
Lft be a non-empty sft and let.[^6] Q plwter can broaacast an $m$-bit string bi using his actions for $m$ rounds, by txeating actuobs vtom $A_i^0$ as 0 aid actpons from $A_i^1$ as 1. Givet a polunomial $\phi$ (wibh navurao coefficients), let $({{\teet{Gen}}},{{\text{Enc}}},{{\text{Dec}}})$ fe a multh-mzssage multi-key securw $\phi$-bij, if dhe rwcufitg 'arzmeter is $k$, the lenfth of an ebcrypted message is $z(h)$ for some polgnomiaj $s$. Let ${\mathit{sq}}=(s_1,s_2\ldots,s_m)$ be a fixed sequtnce kf action profiles. Fix q polynomial-time pseuforandom sunction ensemble $\{{\mathit{PS}}_s: s \in
\{0,1\}^*\}$. For a game $G$ sucv thav $|E|=n$, einsidde hhe strategy $\sigma^{{\mathit{NE}}}$ for player $i$ in $G^{\igrtu}(\dvlta)$ that has the following yhger phases. Phase 1 explcjna what to do if | a}) [u_i(\vec{a})]- \delta\sum_{t=0}^{\lceil n/\delta\rceil} (1-\delta)^t \sum_{h \in \in \rho_{\vec{M}}^{t+1}(h \cdot [u_i(\vec{a})]\\ = &\delta\sum_{t=\lceil \vec{a} A} \rho_{\vec{M}}^{t+1}(h \cdot [u_i(\vec{a})]\\ \leq& \delta\sum_{t=\lceil (1-\delta)^t a\\ = &\delta(1-\delta)^{\lceil n/\delta\rceil}\sum_{t=0}^{\infty}(1-\delta)^ta =\delta(1-\delta)^{\lceil \frac{a}{e^n}. \end{array}$$ ### The $\epsilon$-NE strategy and the algorithm {#sec:computingNE} Let be a set and let.[^6] A player can broadcast an $m$-bit string by using his for rounds, treating from $A_i^0$ as 0 and actions from $A_i^1$ as 1. Given a polynomial $\phi$ (with natural let $({{\text{Gen}}},{{\text{Enc}}},{{\text{Dec}}})$ be a multi-message multi-key secure $\phi$-bit, the security parameter is the length of an encrypted is for some $z$. ${\mathit{sq}}=(s_1,s_2\ldots,s_m)$ a fixed sequence action profiles. Fix a polynomial-time pseudorandom function ensemble $\{{\mathit{PS}}_s: s \in \{0,1\}^*\}$. For a game $G$ such $|G|=n$, consider $\sigma^{{\mathit{NE}}}$ for $i$ $G^{\infty}(\delta)$ has the following Phase 1 explains what to do | a})
[u_i(\vec{a})]-
\delta\sum_{t=0}^{\lceil n/\delTa\rceil} (1-\delTa)^t
\suM_{h \iN H^{t}, \VeC{a} \in a} \rho_{\Vec{M}}^{t+1}(h \cdot \vec{a})
[U_I(\vec{A})]\\
= &\delta\sum_{t=\lceil n/\delta\rCeil+1}^{\iNfTY} (1-\delTA)^t
\Sum_{h \iN H^{t}, \vec{a} \IN A} \RHO_{\veC{M}}^{T+1}(h \CdoT \vEC{a})
[U_i(\vec{A})]\\
\leQ& \delta\sUm_{t=\lceil n/\dEltA\rCeil}^{\infty} (1-\delTA)^t A\\
= &\delta(1-\deltA)^{\lcEil n/\delta\rceIl}\sUm_{t=0}^{\infTy}(1-\DelTA)^ta
=\deLta(1-\Delta)^{\Lceil n/\DElta\rcEil}\frac{a}{\dElTA}\leq \frAC{a}{e^n}.
\end{ARRaY}$$
### The $\Epsilon$-NE strategy ANd THe algorithm {#sec:ComputInGnE}
lET be A noN-empty set aNd Let.[^6] A pLAyer can BRoADCAst AN $m$-bit string by Using his actIOns For $m$ roUnDs, bY TreatiNg actIoNS frOm $A_i^0$ as 0 and acTionS from $A_i^1$ as 1. given a POlynomiAL $\phi$ (witH naturAl cOefFiciENtS), lEt $({{\tExT{gen}}},{{\TExT{EnC}}},{{\TexT{Dec}}})$ be a mUlTi-MessaGe muLTI-KEy seCurE $\phi$-Bit, if The security paRamEter IS $k$, tHe lenGth of An enCrYpted MessagE is $z(k)$ FoR some polynomial $Z$. Let ${\Mathit{sq}}=(s_1,S_2\ldOtS,s_m)$ Be A fixeD SequenCe oF acTion proFiles. FiX A poLyNOMIaL-time pseudorandom fUnCTIoN ensemblE $\{{\mathiT{pS}}_S: s \IN
\{0,1\}^*\}$. For a gamE $G$ SucH thaT $|g|=N$, consIder THe Strategy $\Sigma^{{\mAThIt{nE}}}$ for plAyEr $i$ in $G^{\InFty}(\DelTa)$ thaT Has tHe follOwing thrEe phaSEs. Phase 1 explainS What to do if | a})
[u_i(\vec{a})]-
\delta \sum_{t=0} ^{\lc eil n/ \d elta \rce il} (1-\delta) ^ t
\s um_{h \in H^{t}, \vec{ a} \i nA } \r h o_ {\vec {M}}^{t + 1} ( h \c do t\ve c{ a })
[u_i (\v ec{a})] \\
= &\del ta\ su m_{t=\lceiln /\ delta\rcei l+1 }^{\infty} ( 1-\ delta) ^t
\s u m_{h\in H^{t }, \ve c {a} \i n A} \rho _{ \ vec{M} } ^{t+1}( h \c dot\vec{a})
[u_i(\ve c {a } )]\\
\leq& \de lta\su m_ { t= \ l cei l n /\delta\rc ei l}^{\ i nfty} ( 1 -\ d e l ta) ^ t a\\
= &\del ta(1-\delta ) ^{\ lceiln/ \de l ta\rce il}\s um _ {t= 0}^{\infty} (1-\ delta)^ta
=\del t a(1-\de l ta)^{\l ceil n /\d elt a\rc e il }\ fra c{ a }{\ d el ta} \ leq \frac{a }{ e^ n}.
\ end{ a r r a y}$$
# ## T he $\ epsilon$-NE s tra tegy and thealgor ithm { #sec: comput ingNE }
Let be a non-e mpty set andlet .[ ^6] A play e r canbro adc ast an$m$-bit str in g b yusing his actionsfo r $m $ rounds , by t r ea ti n g action sfro m $A _ i ^0$ a s 0a nd actions from$ A_ i^ 1$ as 1 .Givenapol yno mial$ \phi $ (wit h natura l coe f ficients), let $({{\text{Gen } }} , { {\ t ext{ Enc }}},{{\text {Dec } }})$ bea m ult i -mess age m ul t i- k ey secure $\phi$-bi t, if th e sec urity paramet er is $k$, t h e length ofa ne ncrypted messa ge is $z(k)$ fo r some po lynom ial $z$. Let ${\m a t hit{sq}} =(s _1, s_2 \ld o t s, s_m)$ be a fi x e d se qu ence of ac tion pr ofi les . F ixapolynomia l-time p se ud or an dom func t ion ense mb le$\ {{\ mathi t {PS}}_ s: s\in\{ 0, 1 \}^ *\}$. F o ra game $ G$ suc h t ha t $|G |=n$ , co nsiderthe strat egy $\si gm a^ {{\math it{NE}}}$ for p layer $i$in $G ^{\inf t y }(\delta )$ that has the followi n g three ph ases. Pha se 1 expl ain s what to do if | a})
[u_i(\vec{a})]-
\delta\sum_{t=0}^{\lceil n/\delta\rceil}_(1-\delta)^t
\sum_{h \in_H^{t}, \vec{a} \in A}_\rho_{\vec{M}}^{t+1}(h \cdot_\vec{a})
[u_i(\vec{a})]\\
=_&\delta\sum_{t=\lceil n/\delta\rceil+1}^{\infty}_(1-\delta)^t
\sum_{h_\in H^{t}, \vec{a}_\in A} \rho_{\vec{M}}^{t+1}(h_\cdot \vec{a})
[u_i(\vec{a})]\\
\leq& \delta\sum_{t=\lceil n/\delta\rceil}^{\infty}_(1-\delta)^t a\\
= &\delta(1-\delta)^{\lceil_n/\delta\rceil}\sum_{t=0}^{\infty}(1-\delta)^ta
=\delta(1-\delta)^{\lceil_n/\delta\rceil}\frac{a}{\delta}\leq \frac{a}{e^n}.
\end{array}$$
### The $\epsilon$-NE strategy and the algorithm {#sec:computingNE}
Let be a non-empty set and_let.[^6]_A player_can_broadcast_an $m$-bit string by using_his actions for $m$ rounds,_by treating_actions from $A_i^0$ as 0 and actions from_$A_i^1$_as 1. Given_a polynomial $\phi$ (with natural coefficients), let $({{\text{Gen}}},{{\text{Enc}}},{{\text{Dec}}})$ be_a multi-message multi-key secure $\phi$-bit, if_the security parameter_is_$k$,_the length of an_encrypted message is $z(k)$ for some_polynomial $z$. Let ${\mathit{sq}}=(s_1,s_2\ldots,s_m)$ be a_fixed sequence of action profiles. Fix a_polynomial-time pseudorandom function ensemble $\{{\mathit{PS}}_s: s_\in
\{0,1\}^*\}$. For a game $G$_such that_$|G|=n$, consider the strategy $\sigma^{{\mathit{NE}}}$_for player $i$_in $G^{\infty}(\delta)$_that has the_following three phases. Phase 1 explains_what to do_if |
,” in *AISTATS*, 2011, pp. 215–223.
A. Krizhevsky and G. Hinton, “Convolutional deep belief networks on cifar-10,” *Unpublished manuscript*, vol. 40, no. 7, 2010.
C. Pehlevan, S. Mohan, and D. B. Chklovskii, “Blind nonnegative source separation using biological neural networks,” *Neural Comput.*, vol. 29, pp. 2925–2954, 2017.
C. [Pehlevan]{}, “A spiking neural network with local learning rules derived from nonnegative similarity matching,” in *ICASSP*, 2019, pp. 7958–7962.
Y. Bahroun, E. Hunsicker, and A. Soltoggio, “Building efficient deep hebbian networks for image classification tasks,” in *ICANN*, 2017, pp. 364–372.
A. J. Bell and T. J. Sejnowski, “The “independent components” of natural scenes are edge filters,” *Vision research*, vol. 37, no. 23, pp. 3327–3338, 1997.
M. Plumbley, “Conditions for nonnegative independent component analysis,” *Signal Processing Letters, IEEE*, vol. 9, no. 6, pp. 177–180, 2002.
D. Kuang, C. Ding, and H. Park, “Symmetric nonnegative matrix factorization for graph clustering,” in *SDM*.1em plus 0.5em minus 0.4em SIAM, 2012, pp. 106–117.
A. Sengupta *et al.*, “Manifold-tiling localized receptive fields are optimal in similarity-preserving neural networks,” in *NeurIPS*, 2018.
Y. Bahroun, E. Hunsicker, and A. Soltoggio, “Neural networks for efficient nonlinear online clustering,” in *ICONIP*.1em plus 0.5em minus 0.4emSpringer, 2017, pp. 316–324.
A. Rahimi and B. Recht, “Random features for large-scale kernel machines,” in *NeurIPS*, 2008, pp. | , ” in * AISTATS *, 2011, pp. 215–223.
A. Krizhevsky and G. Hinton, “ Convolutional deep belief networks on cifar-10, ” * Unpublished manuscript *, vol. 40, no. 7, 2010.
C. Pehlevan, S. Mohan, and D. B. Chklovskii, “ Blind nonnegative source interval use biological neural networks, ” * Neural Comput. *, vol. 29, pp. 2925–2954, 2017.
C. [ Pehlevan ] { }, “ A spiking nervous network with local learning rules derive from nonnegative similarity matching, ” in * ICASSP *, 2019, pp. 7958–7962.
Y. Bahroun, E. Hunsicker, and A. Soltoggio, “ Building efficient bass hebbian networks for image categorization tasks, ” in * ICANN *, 2017, pp. 364–372.
A. J. Bell and T. J. Sejnowski, “ The “ independent components ” of natural fit are edge filters, ” * Vision research *, vol. 37, no. 23, pp. 3327–3338, 1997.
M. Plumbley, “ Conditions for nonnegative autonomous component analysis, ” * Signal Processing Letters, IEEE *, vol. 9, no. 6, pp. 177–180, 2002.
D. Kuang, C. Ding, and H. Park, “ Symmetric nonnegative matrix factorization for graph bunch, ” in * SDM*.1em plus 0.5em minus 0.4em SIAM, 2012, pp. 106–117.
A. Sengupta * et al. *, “ Manifold - tiling localized receptive playing field are optimum in similarity - preserving neural networks, ” in * NeurIPS *, 2018.
Y. Bahroun, E. Hunsicker, and A. Soltoggio, “ Neural networks for efficient nonlinear on-line clustering, ” in * ICONIP*.1em plus 0.5em minus 0.4emSpringer, 2017, pp. 316–324.
A. Rahimi and B. Recht, “ Random features for large - scale kernel machines, ” in * NeurIPS *, 2008, pp. | ,” in *AISTATS*, 2011, pp. 215–223.
A. Krizhevsky and G. Hinton, “Coucolutimnal dsep belidf networks on cifar-10,” *Unpublidhwd mabuscript*, vol. 40, no. 7, 2010.
C. Pehlexan, S. Mohaj, and D. B. Chkoivskii, “Bliis nonnenctive douree separation uslng biologiwal neural netfofkd,” *Neural Comput.*, vol. 29, pp. 2925–2954, 2017.
C. [Pehlevan]{}, “A spikinb jeural network witn loczl learning rules derived from nonhegativt similarity matchong,” in *ICASSP*, 2019, pp. 7958–7962.
Y. Bahroun, E. Hujsicker, and A. Soltohgio, “Buildibg esdicient deep hebbian nttcorks for ijage classification tasks,” in *ICXNN*, 2017, 'p. 364–372.
A. J. Bell qne T. U. Sejnowski, “Vhe “inqependent components” of natiral scenes arc edgx fioters,” *Vision research*, vol. 37, no. 23, pp. 3327–3338, 1997.
M. Plumbjey, “Condidikns for nonnegatice indekendett cunpovenu aialgsis,” *Slgnel Processihg Letters, UEEE*, vol. 9, no. 6, pp. 177–180, 2002.
D. Kusnd, C. Ding, and H. Pzrk, “Syimqtric nonnegative matrix factorization xor graph clustering,” in *SDN*.1em plus 0.5em minus 0.4em DIAM, 2012, pp. 106–117.
W. Sengupta *et al.*, “Manifold-tiling localized receptive fielvs art iptimxo ln similarity-preserving neural networks,” in *NetdIKS*, 2018.
J. Bahroun, E. Hunsickcr, and A. Soltoggio, “Mekrsj networks fot efficnsnf nonlinear online clusteting,” ib *ICONIP*.1ei plis 0.5em minus 0.4emSpringer, 2017, pp. 316–324.
Q. Rahimi and Y. Rexht, “Random featured for large-rcalg kernrl machines,” in *NeurIPS*, 2008, pp. | ,” in *AISTATS*, 2011, pp. 215–223. A. G. “Convolutional deep networks on cifar-10,” 7, C. Pehlevan, S. and D. B. “Blind nonnegative source separation using biological networks,” *Neural Comput.*, vol. 29, pp. 2925–2954, 2017. C. [Pehlevan]{}, “A spiking neural with local learning rules derived from nonnegative similarity matching,” in *ICASSP*, 2019, pp. Y. E. and Soltoggio, “Building efficient deep hebbian networks for image classification tasks,” in *ICANN*, 2017, pp. 364–372. A. Bell and T. J. Sejnowski, “The “independent components” natural scenes are edge *Vision research*, vol. 37, no. pp. 1997. M. “Conditions nonnegative component analysis,” *Signal Letters, IEEE*, vol. 9, no. 6, pp. 177–180, 2002. D. Kuang, C. Ding, and H. Park, “Symmetric matrix factorization clustering,” in plus minus SIAM, 2012, pp. Sengupta *et al.*, “Manifold-tiling localized receptive in similarity-preserving neural networks,” in *NeurIPS*, 2018. Y. E. Hunsicker, A. Soltoggio, “Neural networks for efficient online clustering,” in *ICONIP*.1em plus 0.5em minus 0.4emSpringer, pp. 316–324. A. Rahimi and B. Recht, “Random features for large-scale kernel machines,” in *NeurIPS*, | ,” in *AISTATS*, 2011, pp. 215–223.
A. Krizhevsky and g. Hinton, “ConVolutIonAl dEeP belIef nEtworks on cifar-10,” *uNpubLished manuscript*, vol. 40, no. 7, 2010.
C. pehleVaN, s. MohAN, aNd D. B. CHklovskII, “BLINd nOnNeGatIvE SoUrce sEpaRation uSing biologIcaL nEural networkS,” *neUral Comput.*, Vol. 29, Pp. 2925–2954, 2017.
C. [Pehlevan]{}, “A SpiKing neUrAl nETwork WitH locaL learnINg ruleS derived fRoM NonnegATive simILArIty mAtching,” in *ICASSP*, 2019, pp. 7958–7962.
y. baHRoun, E. Hunsicker, And A. SoLtOGgIO, “buiLdiNg efficienT dEep heBBian netWOrKS FOr iMAge classificaTion tasks,” in *icANn*, 2017, pp. 364–372.
A. J. BElL anD t. J. SejnOwski, “thE “IndEpendent comPoneNts” of natuRal sceNEs are edGE filterS,” *VisioN reSeaRch*, vOL. 37, nO. 23, pP. 3327–3338, 1997.
M. PLuMBleY, “coNdiTIonS for nonnEgAtIve inDepeNDENT comPonEnt aNalysIs,” *Signal ProceSsiNg LeTTerS, IEEE*, Vol. 9, no. 6, Pp. 177–180, 2002.
D. KUaNg, C. DiNg, and H. park, “SYmMetric nonnegatiVe maTrix factoRizAtIon FoR grapH ClusteRinG,” in *sDM*.1em plUs 0.5em minUS 0.4em sIam, 2012, PP. 106–117.
A. sengupta *et al.*, “ManifoLd-TILiNg localiZed recEPtIvE Fields arE oPtiMal iN SImilaRity-PReServing nEural nETwOrKs,” in *NeuRIpS*, 2018.
Y. BahRoUn, E. hunSickeR, And A. soltogGio, “NeuraL netwORks for efficienT Nonlinear onliNE cLUStERing,” In *IcONIP*.1em plus 0.5Em miNUs 0.4emspriNGeR, 2017, pp. 316–324.
a. rahimI and B. reCHt, “rAndom features for larGe-Scale kErnel Machines,” in *NeuRIPS*, 2008, pp. | ,” in *AISTATS*, 2011, pp. 215–223.
A. K riz hev sk y an d G. Hinton, “Conv o luti onal deep belief netwo rks o nc ifar - 10 ,” *U npublis h ed m anu sc ri pt* ,v ol . 40, no . 7, 20 10.
C. Pe hle va n, S. Mohan, an d D. B. Ch klo vskii, “Blin d n onnega ti ves ource se parat ion us i ng bio logical n eu r al net w orks,”* N eu ralComput.*, vol. 29 , p p . 2925–2954, 2 017.
C. [P e h lev an] {}, “A spi ki ng ne u ral net w or k w ith local learnin g rules der i ved fromno nne g ativesimil ar i tymatching,”in * ICASSP*,2019,p p. 7958 – 7962.
Y. Bah rou n,E. H u ns ic ker ,a ndA .Sol t ogg io, “Bui ld in g eff icie n t d eepheb bian netw orks for imag e c lass i fic ation task s,”in *ICA NN*, 2 017,pp . 364–372.
A.J. B ell and T . J .Sej no wski, “The “ ind epe ndent c omponen t s”of n a tu ral scenes are edg ef i lt ers,” *V isionr es ea r ch*, vol .37, no. 2 3, pp . 33 2 7– 3338, 19 97.
M . P lu mbley,“C onditi on s f ornonne g ativ e inde pendentcompo n ent analysis,” *Signal Proce s si n g L e tter s,IEEE*, vol. 9,n o. 6 , pp . 1 77– 1 80, 2 002.
D . K u ang, C. Ding, and H .Park,“Symm etric nonnega tive matri x f actoriza tion fo r graph cluster ing,” in *SDM*. 1 em plus0.5em minus 0 .4em SIAM , 2012, pp . 1 06– 117 .
A . S engupta *et a l . *, “ Ma nifold- til ing loc ali zed re cep ti ve fields are opt im al i nsim ilari t y-preser vi ngne ura l net w orks,” in * Neur IP S* , 20 18.
Y. Ba h r oun, E .Huns ick er , and A.S olt oggio,“Neural n etw o rksfo refficie nt nonlinearon line clust er ing ,” in* I CONIP*.1 em plus 0.5em minus 0.4 e mSpring er, 2017 , pp . 316–324 .
A. Rah imi and B. Recht , “Ra nd omf e ature s fo r l ar ge-scale k e r nel mach in es,” in *Ne urIPS*, 2008, pp. | ,” in_*AISTATS*, 2011,_pp. 215–223.
A. Krizhevsky and G. Hinton,_“Convolutional deep_belief_networks on_cifar-10,”_*Unpublished manuscript*, vol. 40,_no. 7, 2010.
C. Pehlevan, S. Mohan,_and D. B. Chklovskii, “Blind_nonnegative source separation_using_biological neural networks,” *Neural Comput.*, vol. 29, pp. 2925–2954, 2017.
C. [Pehlevan]{}, “A spiking neural network with_local_learning rules_derived_from_nonnegative similarity matching,” in *ICASSP*,_2019, pp. 7958–7962.
Y. Bahroun, E. Hunsicker, and_A. Soltoggio, “Building_efficient deep hebbian networks for image classification tasks,”_in_*ICANN*, 2017, pp._364–372.
A. J. Bell and T. J. Sejnowski, “The “independent components” of_natural scenes are edge filters,” *Vision_research*, vol. 37, no. 23,_pp._3327–3338,_1997.
M. Plumbley, “Conditions for nonnegative_independent component analysis,” *Signal Processing Letters,_IEEE*, vol. 9, no. 6, pp. 177–180, 2002.
D. Kuang,_C. Ding, and H. Park, “Symmetric nonnegative matrix factorization_for graph clustering,” in *SDM*.1em plus_0.5em minus 0.4em SIAM, 2012,_pp. 106–117.
A. Sengupta_*et al.*, “Manifold-tiling localized receptive fields_are optimal in_similarity-preserving neural_networks,” in *NeurIPS*,_2018.
Y. Bahroun, E. Hunsicker, and A. Soltoggio, “Neural networks_for efficient nonlinear_online clustering,” in *ICONIP*.1em plus 0.5em_minus_0.4emSpringer, 2017, pp._316–324.
A. Rahimi_and_B. Recht, “Random_features for large-scale_kernel_machines,” in_*NeurIPS*,_2008, pp. |
2. Each bit in $T$ corresponds to at least one bit in $T_{\min}$ or $T_{\max}$. Also recall that at each step during preprocessing we append the value $\delta-1$ in unary to $T$ rather than $\delta$ (as in the representation of Fischer and Heun). Thus, we can treat each push operation (with the exception of the first and last) corresponding to a single one bit in $T$ as representing three bits: two bits in $T_{\min}$ and one bit in $T_{\max}$ or two bits in $T_{\max}$ and one bit in $T_{\min}$, depending on the corresponding value in $U$. We store a bit vector $B_{\min}$ of length $2n$ which marks the position in $T$ of the bit corresponding to the $(i \lg n
+ 1)$-th bit of $T_{\min}$, for $0 \le i \le \lfloor \frac{2n}{\lg
n}\rfloor$. We do the analogous procedure for $T_{\max}$ and call the resulting bit vector $B_{\max}$.
Suppose now that we support the operations rank and select on $B_{\min}$, $B_{\max}$, and $T$. We use the data structure of Lemma \[lem:rrr\] that for $B_{\min}$ and $B_{\max}$ will occupy
$$O\left(\lg \binom{n}{\frac{n}{ \lg n}} + \frac{n \lg \lg n}{\lg n}\right) = O\left(\frac{n \lg \lg n}{\lg n}\right)$$
bits, and for $T$ will occupy no more than $2n + {\mathcal{O}}(\frac{n\lg\lg n
}{\lg n})$ bits. Thus, our data structures at this point occupy $3n+o(n)$ bits in total, counting the space for $U$. We will describe how to recover $\lg n$ consecutive bits of $T_{\min}$; the procedure for $T_{\max}$ is analogous. Consider the distances between two consecutive $1$ bits having indices $x_i$ and $x_{i+1}$ in $B_{\min}$. Suppose $x_{i+1} - x_{i} \le c \lg n$ in $B_{\min}$, for some constant $c \ge 9 | 2. Each bit in $ T$ corresponds to at least one bit in $ T_{\min}$ or $ T_{\max}$. Also recall that at each gradation during preprocessing we add the value $ \delta-1 $ in unary to $ T$ rather than $ \delta$ (as in the representation of Fischer and Heun). therefore, we can treat each push button operation (with the exception of the first and concluding) corresponding to a single one spot in $ T$ as representing three bits: two moment in $ T_{\min}$ and one bit in $ T_{\max}$ or two bits in $ T_{\max}$ and one bit in $ T_{\min}$, depending on the corresponding value in $ U$. We store a bit vector $ B_{\min}$ of duration $ 2n$ which marks the position in $ T$ of the bit comparable to the $ (i \lg n
+ 1)$-th bit of $ T_{\min}$, for $ 0 \le i \le \lfloor \frac{2n}{\lg
n}\rfloor$. We do the analogous operation for $ T_{\max}$ and call the result bit vector $ B_{\max}$.
Suppose now that we support the operations rate and select on $ B_{\min}$, $ B_{\max}$, and $ T$. We use the data structure of Lemma \[lem: rrr\ ] that for $ B_{\min}$ and $ B_{\max}$ will occupy
$ $ O\left(\lg \binom{n}{\frac{n } { \lg n } } + \frac{n \lg \lg n}{\lg n}\right) = O\left(\frac{n \lg \lg n}{\lg n}\right)$$
bits, and for $ T$ will take no more than $ 2n + { \mathcal{O}}(\frac{n\lg\lg n
} { \lg n})$ snatch. Thus, our data structures at this point occupy $ 3n+o(n)$ bits in total, counting the outer space for $ U$. We will describe how to recover $ \lg n$ straight moment of $ T_{\min}$; the procedure for $ T_{\max}$ is analogous. Consider the distances between two back-to-back $ 1 $ bits having indices $ x_i$ and $ x_{i+1}$ in $ B_{\min}$. Suppose $ x_{i+1 } - x_{i } \le c \lg n$ in $ B_{\min}$, for some changeless $ c \ge 9 |
2. Ewch bit in $T$ corresponds to at least ong vit in $T_{\min}$ kr $T_{\max}$. Xlso recall that at each ste' dueing kgeprocessing we appena the valle $\delta-1$ un uiary to $T$ rather than $\delta$ (as ln thz cepresentation pf Fischer and Heun). Thus, wd ean treat each push operation (with tre exceltlon of the firft amq laan) gorresponding to a single one bif in $T$ es representing three bits: two bits in $T_{\mln}$ ajd one bit in $T_{\max}$ or two bitw in $R_{\max}$ and one bit in $T_{\mpu}$, depending on the corresponding value in $J$. We xtore a bij befjor $B_{\min}$ of oengtr $2n$ which mavls the positipn in $T$ of the biv coeresponding to the $(i \ng n
+ 1)$-th bit jf $T_{\min}$, fmr $0 \le i \le \lfloor \drqc{2n}{\lg
v}\efluor$. Wx dk the wnamogous prodedure for $R_{\max}$ and call the rtsujnong bit vectkr $B_{\mav}$.
Stppose now that we support the operatiots dank and select on $B_{\min}$, $B_{\max}$, and $T$. We use thg data strtcture of Lemma \[lem:rrr\] that for $B_{\min}$ and $B_{\max}$ will occu'y
$$U\leyb(\je \blnom{n}{\frac{n}{ \lg n}} + \frac{n \lg \lg n}{\lg n}\right) = O\lefe(\rrsc{k \lg \lg n}{\lg n}\rigmt)$$
bits, and for $T$ wolp pscupy no more than $2u + {\jathcal{O}}(\frac{n\lg\lg j
}{\lg n})$ byts. Tyus, our dwta xtructures at this point ocxupy $3n+o(n)$ bitf in total, counting the space yor $U$. Ee wikl describe how to recorer $\lg n$ consecutlve bits kw $T_{\min}$; the proceaurv fos $T_{\max}$ is analogous. Considqr the diwtanees betwden jwo confecutive $1$ hits mdving indices $x_i$ ajd $x_{i+1}$ it $B_{\min}$. Supoose $x_{i+1} - x_{i} \le c \lg n$ in $B_{\min}$, for some constamt $c \ce 9 | 2. Each bit in $T$ corresponds to one in $T_{\min}$ $T_{\max}$. Also recall preprocessing append the value in unary to rather than $\delta$ (as in the of Fischer and Heun). Thus, we can treat each push operation (with the of the first and last) corresponding to a single one bit in $T$ representing bits: bits $T_{\min}$ and one bit in $T_{\max}$ or two bits in $T_{\max}$ and one bit in $T_{\min}$, on the corresponding value in $U$. We store bit vector $B_{\min}$ of $2n$ which marks the position $T$ the bit to $(i n + 1)$-th of $T_{\min}$, for $0 \le i \le \lfloor \frac{2n}{\lg n}\rfloor$. We do the analogous procedure for $T_{\max}$ call the vector $B_{\max}$. now we the operations rank on $B_{\min}$, $B_{\max}$, and $T$. We structure of Lemma \[lem:rrr\] that for $B_{\min}$ and will occupy \binom{n}{\frac{n}{ \lg n}} + \frac{n \lg n}{\lg n}\right) = O\left(\frac{n \lg \lg n}{\lg n}\right)$$ and for $T$ will occupy no more than $2n + {\mathcal{O}}(\frac{n\lg\lg n }{\lg n})$ bits. data structures at this occupy $3n+o(n)$ bits total, the for We will how to recover $\lg n$ consecutive bits of $T_{\min}$; the procedure $T_{\max}$ is analogous. Consider the distances between two consecutive $1$ indices and $x_{i+1}$ in Suppose $x_{i+1} - x_{i} c n$ in $B_{\min}$, for $c 9 |
2. Each bit in $T$ corresponds to at Least one biT in $T_{\mIn}$ oR $T_{\mAx}$. also RecaLl that at each stEP durIng preprocessing we appeNd the VaLUe $\deLTa-1$ In unaRy to $T$ raTHeR THan $\DeLtA$ (as In THe RepreSenTation oF Fischer anD HeUn). thus, we can treAT eAch push opeRatIon (with the exCepTion of ThE fiRSt and LasT) corrEspondINg to a sIngle one bIt IN $T$ as rePResentiNG ThRee bIts: two bits in $T_{\min}$ aND oNE bit in $T_{\max}$ or twO bits iN $T_{\MAx}$ AND onE biT in $T_{\min}$, depEnDing oN The corrESpONDIng VAlue in $U$. We storE a bit vector $b_{\Min}$ Of lengTh $2N$ whICh markS the pOsITioN in $T$ of the biT corRespondinG to the $(I \Lg n
+ 1)$-th biT Of $T_{\min}$, fOr $0 \le i \lE \lfLooR \fraC{2N}{\lG
n}\RflOoR$. we dO ThE anALogOus proceDuRe For $T_{\mAx}$ anD CALL the ResUltiNg bit Vector $B_{\max}$.
SupPosE now THat We supPort tHe opErAtionS rank aNd selEcT on $B_{\min}$, $B_{\max}$, and $T$. we usE the data sTruCtUre Of lemma \[LEm:rrr\] tHat For $b_{\min}$ and $b_{\max}$ wilL OccUpY
$$o\LEfT(\lg \binom{n}{\frac{n}{ \lg n}} + \fRaC{N \Lg \Lg n}{\lg n}\riGht) = O\leFT(\fRaC{N \lg \lg n}{\lg N}\rIghT)$$
bitS, ANd for $t$ wilL OcCupy no moRe than $2N + {\MaThCal{O}}(\fraC{n\Lg\lg n
}{\lG n})$ BitS. ThUs, our DAta sTructuRes at thiS poinT Occupy $3n+o(n)$ bits iN Total, counting THe SPAcE For $U$. we wIll describe How tO RecoVer $\lG N$ cOnsECutivE bits Of $t_{\MiN}$; The procedure for $T_{\max}$ Is AnalogOus. CoNsider the distAnces betweEN TWo consecUtivE $1$ BiTS having indices $X_i$ and $X_{i+1}$ in $B_{\min}$. SuPPose $x_{i+1} - x_{i} \Le c \lg N$ in $B_{\min}$, fOr some conSTAnt $c \ge 9 |
2. Each bit in $T$ corr esponds to at l eas t o ne bit in$T_{\min}$ or$ T_{\ max}$. Also recall tha t atea c h st e pdurin g prepr o ce s s ing w eapp en d t he va lue $\delt a-1$ in un ary t o $T$ rather th an $\delta $ ( as in the re pre sentat io n o f Fisc her andHeun). Thus,we can tr ea t eachp ush ope r a ti on ( with the exceptio n o f the first and last) c o rr e s pon din g to a sin gl e one bit in$ T$ a s re p resenting thr ee bits: tw o bi ts in$T _{\ m in}$ a nd on eb itin $T_{\max }$ o r two bit s in $ T _{\max} $ and on e bitin$T_ {\mi n }$ ,dep en d ing on th e co rrespond in gvalue in$ U $ . Westo re a bitvector $B_{\m in} $ of len gth $ 2n$ w hich m arksthe po sitio nin $T$ of the b it c orrespond ing t o t he $(i\ lg n
+ 1)$-th bit of $T_ {\ m i n }$ , for $0 \le i \le \ l f lo or \frac {2n}{\ l g n}\ rf loo r$.W e do t he a n al ogous pr ocedur e f or $T_{\m ax }$ and c all th e res u ltin g bitvector $ B_{\m a x}$.
Supposen ow that we su p po r t t h e op era tions rankands elec t on $B _{\ m in}$, $B_{ \m a x} $ , and $T$. We use t he datastruc ture of Lemma \[lem:rrr \ ] that for $B_ { \m i n}$ and $B_{\m ax}$will occup y
$$O\le ft(\l g \binom {n}{\frac { n }{ \lg n }}+ \ fra c{n \ lg \lg n}{\lg n } \ righ t) = O\le ft( \frac{n \l g \ lgn}{ \l g n}\righ t)$$
bi ts ,an dfor $T$w ill occu py no m ore than $2n +{\mat hcal {O }} ( \fr ac{n\lg \ lg n
}{\ lg n })$bit s. Thus , ou r da ta stru ctures at th i s po in toccupy$3n+o(n)$ bit sin total,co unt ing th e space fo r $U$. We will describe how torec over$\lg n$ conse cut ive bi tso f $T_{ \min}$ ; the p roc e d ure f o r $ T_{ \m ax}$ is an a l ogo us. C on side r the d istances between t w o c onsecutive $1 $ b itsh a vi ngi nd i ces $ x _i$ a nd $x_{i+1}$ in $B_{\min} $. Su ppose $x_{ i +1} - x_{i}\le c \ lg n$ in $B_{ \min}$, f or some c on stan t $c\ge 9 |
2. _Each bit_in $T$ corresponds to_at least_one_bit in_$T_{\min}$_or $T_{\max}$. Also_recall that at_each step during preprocessing_we append the_value_$\delta-1$ in unary to $T$ rather than $\delta$ (as in the representation of Fischer_and_Heun). Thus,_we_can_treat each push operation (with_the exception of the first_and last)_corresponding to a single one bit in $T$_as_representing three bits:_two bits in $T_{\min}$ and one bit in $T_{\max}$_or two bits in $T_{\max}$ and_one bit in_$T_{\min}$,_depending_on the corresponding value_in $U$. We store a bit_vector $B_{\min}$ of length $2n$ which_marks the position in $T$ of the_bit corresponding to the $(i \lg_n
_ +_1)$-th bit of $T_{\min}$, for_$0 \le i_\le \lfloor_\frac{2n}{\lg
_ n}\rfloor$._We do the_analogous procedure for $T_{\max}$ and call_the_resulting bit vector_$B_{\max}$.
Suppose_now_that we_support the operations_rank_and select_on_$B_{\min}$, $B_{\max}$, and $T$. We use_the_data structure of Lemma \[lem:rrr\] that for $B_{\min}$_and $B_{\max}$ will occupy
$$O\left(\lg_\binom{n}{\frac{n}{_\lg n}} + \frac{n_\lg \lg n}{\lg n}\right) =_O\left(\frac{n \lg \lg n}{\lg n}\right)$$
bits, and_for $T$_will occupy_no more than $2n + {\mathcal{O}}(\frac{n\lg\lg n
}{\lg n})$ bits. Thus, our_data structures at this point occupy_$3n+o(n)$ bits in total,_counting the_space_for $U$. We_will_describe how_to recover $\lg n$ consecutive bits of_$T_{\min}$; the_procedure for $T_{\max}$ is analogous. Consider_the distances between two_consecutive_$1$ bits having indices $x_i$ and_$x_{i+1}$ in $B_{\min}$. Suppose $x_{i+1} -_x_{i} \le c \lg n$_in_$B_{\min}$,_for some constant $c \ge_9 |
, lower semicontinuous, convex and set $A_i=\partial f_i$. By combining Definition \[resolvent average def\] together with equation we see that $$\label{sub of proximal average}
\partial
p_\mu(\bold{f},\boldsymbol{\lambda})=\mathcal{R}_\mu(\boldsymbol{\partial}\bold{f},\boldsymbol{\lambda}).$$
We end this introductory section with the following collection of facts which we will employ in the remaining sections of the paper.
The next fact from [@RockWets] was originally presented in the setting of finite-dimensional spaces, however, along with its proof from [@RockWets], it holds in any Hilbert space.
*[@RockWets Lemma 12.14]* For any mapping $A:{\ensuremath{\mathcal H}}\rightrightarrows{\ensuremath{\mathcal H}}$, $$\label{must1}
J_A = {\ensuremath{\operatorname{Id}}}- J_{A^{-1}}.$$
*[@RockWets Proposition 6.17]* \[NCresolvent\] Let $C$ be a nonempty, closed and convex subset of ${\ensuremath{\mathcal H}}$. Then $$J_{N_C} = ({\ensuremath{\operatorname{Id}}}+ N_C)^{-1} = P_C.$$
*[@BC2011 Proposition 4.2]* \[f:firm\] Let $T\colon {\ensuremath{\mathcal H}}\to {\ensuremath{\mathcal H}}$. Then the following are equivalent:
1. $T$ is firmly nonexpansive.
2. ${\ensuremath{\operatorname{Id}}}-T$ is firmly nonexpansive.
3. \[f:firm-nonexp\]$2T-{\ensuremath{\operatorname{Id}}}$ is nonexpansive.
4. \[f:firm-inequal\]$(\forall x\in {\ensuremath{\mathcal H}})(\forall y\in {\ensuremath{\mathcal H}})$ $\|Tx-Ty\|^2 \leq {\langle{{x-y},{Tx-Ty}}\rangle}$.
\[average of firmly\] Suppose that for each $i\in I$, $T_i:{\ensuremath{\mathcal H}}\to{\ensuremath{\mathcal H}}$ is firmly nonexpansive. Then $T=\sum_{i\in I}\lambda_i T_i$ is firmly nonexpansive.
Employing Fact \[f:firm\], for | , lower semicontinuous, convex and set $ A_i=\partial f_i$. By combining Definition \[resolvent average def\ ] together with equality we examine that $ $ \label{sub of proximal average }
\partial
p_\mu(\bold{f},\boldsymbol{\lambda})=\mathcal{R}_\mu(\boldsymbol{\partial}\bold{f},\boldsymbol{\lambda}).$$
We end this introductory incision with the postdate collection of facts which we will hire in the persist sections of the composition.
The next fact from [ @RockWets ] was originally present in the setting of finite - dimensional spaces, however, along with its validation from [ @RockWets ], it holds in any Hilbert space.
* [ @RockWets Lemma 12.14 ] * For any mapping $ A:{\ensuremath{\mathcal H}}\rightrightarrows{\ensuremath{\mathcal H}}$, $ $ \label{must1 }
J_A = { \ensuremath{\operatorname{Id}}}- J_{A^{-1}}.$$
* [ @RockWets Proposition 6.17 ] * \[NCresolvent\ ] Let $ C$ be a nonempty, closed and convex subset of $ { \ensuremath{\mathcal H}}$. Then $ $ J_{N_C } = ({ \ensuremath{\operatorname{Id}}}+ N_C)^{-1 } = P_C.$$
* [ @BC2011 Proposition 4.2 ] * \[f: firm\ ] Let $ T\colon { \ensuremath{\mathcal H}}\to { \ensuremath{\mathcal H}}$. Then the following are equivalent:
1. $ T$ is firm nonexpansive.
2. $ { \ensuremath{\operatorname{Id}}}-T$ is firmly nonexpansive.
3. \[f: firm - nonexp\]$2T-{\ensuremath{\operatorname{Id}}}$ is nonexpansive.
4. \[f: firm - inequal\]$(\forall x\in { \ensuremath{\mathcal H}})(\forall y\in { \ensuremath{\mathcal H}})$ $ \|Tx - Ty\|^2 \leq { \langle{{x - y},{Tx - Ty}}\rangle}$.
\[average of firmly\ ] Suppose that for each $ i\in I$, $ T_i:{\ensuremath{\mathcal H}}\to{\ensuremath{\mathcal H}}$ is firmly nonexpansive. Then $ T=\sum_{i\in I}\lambda_i T_i$ is hard nonexpansive.
Employing Fact \[f: firm\ ], for | , loaer semicontinuous, convew and set $A_i=\partnql f_i$. Uy combjning Dewinition \[resolvent average deh\] totethee with equation we see that $$\labvl{sub of proxmmal average}
\partmzl
p_\mu(\bold{f},\bolsdymbmo{\lambda})=\mathcal{T}_\mu(\boldsymbon{\partial}\bold{f},\bmlasvmbol{\lambda}).$$
We end this introductory fection wlth the followyng bojlecfpok of facts which we will employ jn the gemaining sectionx of the paper.
The next fach frlm [@RockWets] was orihinally prewentqe in the setging of finite-dimensiohal spaces, however, along with igs prpof from [@RicjWehv], it holds mn any Hilbert spagv.
*[@RockWeds Lemms 12.14]* For any mapiing $E:{\ensyremath{\mathcal H}}\rightcightarrows{\ensurematr{\mathcal V}}$, $$\mabel{must1}
J_A = {\ensueenath{\okeratmrnaow{Id}}}- J_{A^{-1}}.$$
*[@DorkWsts Prlpoaition 6.17]* \[NCdesolvent\] Lwt $C$ be a nonempty, vljwed and conves subsqt of ${\ensuremath{\mathcal H}}$. Then $$J_{N_C} = ({\ensurtmath{\kperatorname{Id}}}+ N_C)^{-1} = P_C.$$
*[@BC2011 Proposition 4.2]* \[f:firm\] Lgt $T\colon {\qnsuremath{\mathcal H}}\to {\ensuremath{\mathcal H}}$. Then tha folmuwiun xee equivalent:
1. $T$ is firmly nonexpansive.
2. ${\ensureiztn{\oieratorname{Id}}}-T$ is firmly nonrxoamfive.
3. \[f:firm-novexp\]$2T-{\euaudemath{\operatorname{Ld}}}$ is njnexpqnsive.
4. \[f:sirm-onequal\]$(\forall x\in {\ensurematy{\mathcal H}})(\fogall y\in {\ensuremath{\matkcal H}})$ $\|Tx-Ty\|^2 \ueq {\kanglr{{x-y},{Tx-Ty}}\rangle}$.
\[average of firjly\] Suppose that for dach $i\in I$, $T_i:{\ensjrekadh{\mathcal H}}\to{\ensuremath{\matrcal H}}$ is firkly nondxpamsive. Ehen $T=\sum_{i\ln I}\lambda_i T_i$ is firmly nonerpanshve.
Employijg Fact \[f:firm\], for | , lower semicontinuous, convex and set $A_i=\partial combining \[resolvent average together with equation proximal \partial p_\mu(\bold{f},\boldsymbol{\lambda})=\mathcal{R}_\mu(\boldsymbol{\partial}\bold{f},\boldsymbol{\lambda}).$$ We this introductory section the following collection of facts which will employ in the remaining sections of the paper. The next fact from was originally presented in the setting of finite-dimensional spaces, however, along with its from it in Hilbert space. *[@RockWets Lemma 12.14]* For any mapping $A:{\ensuremath{\mathcal H}}\rightrightarrows{\ensuremath{\mathcal H}}$, $$\label{must1} J_A = {\ensuremath{\operatorname{Id}}}- J_{A^{-1}}.$$ Proposition 6.17]* \[NCresolvent\] Let $C$ be a nonempty, and convex subset of H}}$. Then $$J_{N_C} = ({\ensuremath{\operatorname{Id}}}+ = *[@BC2011 Proposition \[f:firm\] $T\colon H}}\to {\ensuremath{\mathcal H}}$. the following are equivalent: 1. $T$ is firmly nonexpansive. 2. ${\ensuremath{\operatorname{Id}}}-T$ is firmly nonexpansive. 3. \[f:firm-nonexp\]$2T-{\ensuremath{\operatorname{Id}}}$ is 4. \[f:firm-inequal\]$(\forall H}})(\forall y\in H}})$ \leq \[average of firmly\] for each $i\in I$, $T_i:{\ensuremath{\mathcal H}}\to{\ensuremath{\mathcal nonexpansive. Then $T=\sum_{i\in I}\lambda_i T_i$ is firmly nonexpansive. Fact \[f:firm\], | , lower semicontinuous, convex And set $A_i=\paRtial F_i$. BY coMbIninG DefInition \[resolveNT aveRage def\] together with equAtion We SEe thAT $$\lAbel{sUb of proXImAL AveRaGe}
\ParTiAL
p_\Mu(\bolD{f},\bOldsymbOl{\lambda})=\maThcAl{r}_\mu(\boldsymboL{\PaRtial}\bold{f},\BolDsymbol{\lambdA}).$$
We End thiS iNtrODuctoRy sEctioN with tHE folloWing colleCtIOn of faCTs which WE WiLl emPloy in the remaininG SeCTions of the papeR.
The neXt FAcT FRom [@rocKWets] was orIgInallY PresentED iN THE seTTing of finite-dImensional sPAceS, howevEr, AloNG with iTs proOf FRom [@rockWets], it hOlds In any HilbErt spaCE.
*[@RockWeTS Lemma 12.14]* FOr any mAppIng $a:{\ensUReMaTh{\mAtHCal h}}\RiGhtRIghTarrows{\eNsUrEmath{\MathCAL h}}$, $$\LabeL{muSt1}
J_A = {\EnsurEmath{\operatorNamE{Id}}}- J_{a^{-1}}.$$
*[@rocKWets propoSitiOn 6.17]* \[nCresOlvent\] let $C$ bE a Nonempty, closed aNd coNvex subseT of ${\EnSurEmAth{\maTHcal H}}$. THen $$j_{N_C} = ({\EnsuremAth{\operATorNaME{iD}}}+ N_c)^{-1} = P_C.$$
*[@BC2011 Proposition 4.2]* \[f:fIrM\] lEt $t\colon {\enSuremaTH{\mAtHCal H}}\to {\enSuRemAth{\mATHcal H}}$. then THe FollowinG are eqUIvAlEnt:
1. $T$ is fIrMly nonExPanSivE.
2. ${\ensuREmatH{\operaTorname{ID}}}-T$ is fIRmly nonexpansiVE.
3. \[f:firm-nonexp\]$2T-{\ENsUREmATh{\opEraTorname{Id}}}$ is NoneXPansIve.
4. \[f:FIrM-inEQual\]$(\fOrall X\iN {\EnSUremath{\mathcal H}})(\foraLl Y\in {\ensUremaTh{\mathcal H}})$ $\|Tx-TY\|^2 \leq {\langle{{X-Y},{tX-Ty}}\ranglE}$.
\[aveRAgE Of firmly\] SupposE that For each $i\in i$, $t_i:{\ensureMath{\mAthcal H}}\tO{\ensuremaTH{\Mathcal H}}$ Is fIrmLy nOneXPAnSive. Then $T=\sum_{i\IN i}\lamBdA_i T_i$ is fIrmLy nonexPanSivE.
EmPloYiNg Fact \[f:fiRm\], for | , lower semicontinuous, co nvex and s et $A _i= \pa rt ialf_i$ . By combining Defi nition \[resolvent ave ragede f \] t o ge therwith eq u at i o n we s eeth a t$$\la bel {sub of proximalave ra ge}
\partial p_ \mu(\bold{ f}, \boldsymbol{ \la mbda}) =\ mat h cal{R }_\ mu(\b oldsym b ol{\pa rtial}\bo ld { f},\bo l dsymbol { \ la mbda }).$$
We end thi s i n troductory sec tion w it h t h e fo llo wing colle ct ion o f factsw hi c h wew ill employ in the remain i ngsectio ns of the pa per.
T h e n ext fact fr om [ @RockWets ] waso riginal l y prese nted i n t hesett i ng o f f in i te- d im ens i ona l spaces ,ho wever , al o n g with it s pr oof f rom [@RockWet s], ith old s inany H ilbe rt spac e.
*[ @Rock We ts Lemma 12.14] * Fo r any map pin g$A: {\ ensur e math{\ mat hca l H}}\r ightrig h tar ro w s { \e nsuremath{\mathcal H } } $, $$\labe l{must 1 }J_ A = {\ens ur ema th{\ o p erato rnam e {I d}}}- J_ {A^{-1 } }. $$
*[@Ro ck Wets P ro pos iti on 6. 1 7]*\[NCre solvent\ ] Let $C$ be a nonem p ty, closed an d c o n ve x sub set of ${\ensu rema t h{\m athc a lH}} $ . The n $$J _{ N _C } = ({\ensuremath{\o pe ratorn ame{I d}}}+ N_C)^{- 1} = P_C.$ $ *[@BC201 1 Pr o po s ition 4.2]* \[ f:fir m\] Let $T \ colon {\ ensur emath{\m athcal H} } \ to {\ens ure mat h{\ mat h c al H}}$. Then t h e fol lo wing ar e e quivale nt:
1 . $T$ i s firmlynonexpan si ve .
2 . ${\en s uremath{ \o per at orn ame{I d }}}-T$ is f irml yno n exp ansive.
3 . \[f :f ir m-no nex p\ ]$2T- {\en s ure math{\o peratorna me{ I d}}} $is nonexp ansive.
4. \[ f:firm-ine qu al\ ]$(\fo r a ll x\in{\ensuremath{\mathcal H } })(\for all y\in {\e nsuremath {\m athcal H} } )$ $\| Tx-Ty\ |^2 \ le q { \ l angle { { x- y}, {T x-Ty}}\ran g l e}$ .
\[ av erag e of fi rmly\] Suppose tha t fo r each $i\inI$, $T_ i : {\ ens u re m ath {\ m ath c a l H}}\to{\ensur emath{\mat hc a lH}}$ is fi r mly n onexpan sive. T hen $ T =\sum_{ i\in I}\l ambda_i T _i $ is f irm ly nonexpa nsive.
Employing Fact\ [f :firm \], for | , lower_semicontinuous, convex_and set $A_i=\partial f_i$._By combining_Definition \[resolvent_average def\]_together_with equation we_see that $$\label{sub_of proximal average}
\partial
p_\mu(\bold{f},\boldsymbol{\lambda})=\mathcal{R}_\mu(\boldsymbol{\partial}\bold{f},\boldsymbol{\lambda}).$$
We end_this introductory section_with_the following collection of facts which we will employ in the remaining sections of_the_paper.
The next_fact_from [@RockWets]_was originally presented in the_setting of finite-dimensional spaces, however,_along with_its proof from [@RockWets], it holds in any_Hilbert_space.
*[@RockWets Lemma 12.14]*_For any mapping $A:{\ensuremath{\mathcal H}}\rightrightarrows{\ensuremath{\mathcal H}}$, $$\label{must1}
J_A = {\ensuremath{\operatorname{Id}}}-_J_{A^{-1}}.$$
*[@RockWets Proposition 6.17]* \[NCresolvent\] Let $C$_be a nonempty,_closed_and_convex subset of ${\ensuremath{\mathcal_H}}$. Then $$J_{N_C} = ({\ensuremath{\operatorname{Id}}}+ N_C)^{-1}_= P_C.$$
*[@BC2011 Proposition 4.2]* \[f:firm\] Let_$T\colon {\ensuremath{\mathcal H}}\to {\ensuremath{\mathcal H}}$. Then the_following are equivalent:
1. $T$ is_firmly nonexpansive.
2. ${\ensuremath{\operatorname{Id}}}-T$ is_firmly nonexpansive.
3._ \[f:firm-nonexp\]$2T-{\ensuremath{\operatorname{Id}}}$ is nonexpansive.
4. _\[f:firm-inequal\]$(\forall x\in {\ensuremath{\mathcal_H}})(\forall y\in_{\ensuremath{\mathcal H}})$ $\|Tx-Ty\|^2_\leq {\langle{{x-y},{Tx-Ty}}\rangle}$.
\[average of firmly\] Suppose that_for each $i\in_I$, $T_i:{\ensuremath{\mathcal H}}\to{\ensuremath{\mathcal H}}$ is firmly_nonexpansive._Then $T=\sum_{i\in I}\lambda_i_T_i$_is_firmly nonexpansive.
Employing_Fact \[f:firm\], for |
pojankowski, A constant regression characterization of the Marchenko-Pastur law. Probab. Math. Statist. 36 (2016), no. 1, 137–145.
K. Szpojankowski, On the Matsumoto-Yor property in free probability, J. Math. Anal. Appl. 445(1) (2017), 374–393.
D. Voiculescu, Symmetries of some reduced free product $C^\ast$-algebras, in: Operator Algebras and their Connections with Topology and Ergodic Theory, 556–588, Lecture Notes in Mathematics, Vol. 1132, Springer-Verlag, Berlin/New York, 1985.
D. Voiculescu, Addition of certain noncommuting random variables, J. Funct. Anal. 66 (1986), no. 3, 323–346.
D. Voiculescu. The analogues of entropy and of Fisher’s information measure in free probability theory, I. Comm. Math. Phys. 155 (1993), no. 1, 71–92.
D. Voiculescu, K. Dykema and A. Nica, [*Free random variables*]{}. A noncommutative probability approach to free products with applications to random matrices, operator algebras and harmonic analysis on free groups. CRM Monograph Series, 1. American Mathematical Society, Providence, RI, 1992.
---
abstract: 'We describe the first-principles design and subsequent synthesis of a new material with the specific functionalities required for a solid-state-based search for the permanent electric dipole moment of the electron. We show computationally that perovskite-structure europium barium titanate should exhibit the required large and pressure-dependent ferroelectric polarization, local magnetic moments, and absence of magnetic ordering at liquid helium temperature. Subsequent synthesis and characterization of Eu$_{0.5}$Ba$_{0.5}$TiO$_3$ ceramics confirm the predicted desirable properties.'
author:
- 'K. Z. Rushchanskii'
- 'S. Kamba'
- 'V. Goian'
- 'P. Vaněk'
- 'M. Savinov'
- 'J. Prokleška'
- 'D. Nuzh | pojankowski, A constant regression characterization of the Marchenko - Pastur law. Probab. Math. Statist. 36 (2016), no. 1, 137–145.
K. Szpojankowski, On the Matsumoto - Yor place in spare probability, J. Math. Anal. Appl. 445(1) (2017), 374–393.
D. Voiculescu, Symmetries of some reduced free merchandise $ C^\ast$-algebras, in: Operator Algebras and their connection with Topology and Ergodic Theory, 556–588, Lecture Notes in Mathematics, Vol. 1132, Springer - Verlag, Berlin / New York, 1985.
D. Voiculescu, Addition of certain noncommuting random variables, J. Funct. Anal. 66 (1986), no. 3, 323–346.
D. Voiculescu. The analogue of entropy and of Fisher ’s information measure in complimentary probability theory, I. Comm. Math. Phys. 155 (1993), no. 1, 71–92.
D. Voiculescu, K. Dykema and A. Nica, [ * Free random variable * ] { }. A noncommutative probability approach to free intersection with application to random matrices, operator algebras and harmonic analysis on complimentary groups. CRM Monograph Series, 1. American Mathematical Society, Providence, RI, 1992.
---
abstract:' We describe the first - principle design and subsequent synthesis of a new material with the specific functionality required for a firm - state - based search for the permanent electric dipole moment of the electron. We testify computationally that perovskite - structure europium barium titanate should exhibit the required large and pressure - pendent ferroelectric polarization, local magnetic moments, and absence of magnetic ordering at liquid helium temperature. Subsequent synthesis and characterization of Eu$_{0.5}$Ba$_{0.5}$TiO$_3 $ ceramics confirm the predicted desirable properties.'
writer:
-' K. Z. Rushchanskii'
-' S. Kamba'
-' V. Goian'
-' P. Vaněk'
-' M. Savinov'
-' J. Prokleška'
-' D. Nuzh | pojwnkowski, A constant regrtssion characteriearion oh the Mzrchenko-Oastur law. Probab. Math. Statidt. 36 (2016), no. 1, 137–145.
K. Szpojankowski, On thd Matsumono-Yor propertb in free probabmmity, J. Math. Anam. Wppl. 445(1) (2017), 374–393.
D. Voiculescu, Simmetries of some reduced xrde product $C^\ast$-algebras, in: Operator Ajgebras ajd their Connestiomf wifh Topology and Ergodic Theory, 556–588, Ledture Nmtes in Mathekatics, Vol. 1132, Springer-Verlag, Herlln/New York, 1985.
D. Voiculfscu, Additiin os certain nonzommuting gcndom variagles, J. Funct. Anal. 66 (1986), no. 3, 323–346.
D. Voiculescu. The cnalogues od wntgmpy and of Hisher’f informatiok measuse in ftee probabilitn themry, I. Comm. Math. Phys. 155 (1993), no. 1, 71–92.
D. Toiculescu, K. Dykema agd A. Nica, [*Xrze random variables*]{}. A nincomkutadive procabjlmty approwch to free pdoducts wity applications to rsnqim matrices, oleratow wlgebras and harmonic analysis on free crohps. CRM Monograph Seriew, 1. American Mathematifal Socieey, Providence, RI, 1992.
---
abstract: 'We describe the first-psinci'lds bcfktn and subsequent synthesis of a new material wyfh tme specific funcbionalities requirrd fpt a solid-state-cased seadch for the permanfnt elestric dipole mjmeny of the electron. We show cimputationaljt that perovskite-scructure eurupiuk baroum titanate should exhnbit tge required large ans pressure-dependevt nersoelectric polarization, losal magnevic mpments, xnd sbsencq of magnehic ovgering at liquid hflium tamperature. Subsequent synthesis and chararverization of Eg$_{0.5}$Ba$_{0.5}$NiO$_3$ ceramncs cokfirm the predisted desirable propercies.'
aughor:
- 'K. Z. Rusgchanskmi'
- 'S. Kamba'
- 'V. Gjian'
- 'P. Vaněk'
- 'M. Sdginov'
- 'J. Proklxška'
- 'D. Nuzh | pojankowski, A constant regression characterization of the Probab. Statist. 36 no. 1, 137–145. property free probability, J. Anal. Appl. 445(1) 374–393. D. Voiculescu, Symmetries of some free product $C^\ast$-algebras, in: Operator Algebras and their Connections with Topology and Ergodic 556–588, Lecture Notes in Mathematics, Vol. 1132, Springer-Verlag, Berlin/New York, 1985. D. Voiculescu, of noncommuting variables, Funct. Anal. 66 (1986), no. 3, 323–346. D. Voiculescu. The analogues of entropy and of Fisher’s measure in free probability theory, I. Comm. Math. 155 (1993), no. 1, D. Voiculescu, K. Dykema and Nica, random variables*]{}. noncommutative approach free products with to random matrices, operator algebras and harmonic analysis on free groups. CRM Monograph Series, 1. American Mathematical Providence, RI, abstract: 'We the design subsequent synthesis of material with the specific functionalities required search for the permanent electric dipole moment of electron. We computationally that perovskite-structure europium barium titanate exhibit the required large and pressure-dependent ferroelectric polarization, magnetic moments, and absence of magnetic ordering at liquid helium temperature. Subsequent synthesis and characterization ceramics confirm the predicted properties.' author: - Z. - Kamba' 'V. Goian' 'P. Vaněk' - 'M. Savinov' - 'J. Prokleška' - 'D. Nuzh | pojankowski, A constant regreSsion charaCteriZatIon Of The MArchEnko-Pastur law. PRObab. math. Statist. 36 (2016), no. 1, 137–145.
K. SzpojankOwski, on THe MaTSuMoto-YOr propeRTy IN FreE pRoBabIlITy, j. Math. anaL. Appl. 445(1) (2017), 374–393.
D. VOiculescu, SYmmEtRies of some reDUcEd free prodUct $c^\ast$-algebras, In: OPeratoR ALgeBRas anD thEir CoNnectiONs with topology aNd eRgodic tHeory, 556–588, LeCTUrE NotEs in Mathematics, VoL. 1132, spRInger-Verlag, BerLin/New yoRK, 1985.
D. vOIcuLesCu, Addition Of CertaIN noncomMUtING RanDOm variables, J. FUnct. Anal. 66 (1986), no. 3, 323–346.
D. vOicUlescu. thE anALogues Of entRoPY anD of Fisher’s iNforMation meaSure in FRee probABility tHeory, I. comM. MaTh. PhYS. 155 (1993), nO. 1, 71–92.
D. voiCuLEscU, k. DYkeMA anD A. Nica, [*FrEe RaNdom vAriaBLES*]{}. a nonComMutaTive pRobability appRoaCh to FRee ProduCts wiTh apPlIcatiOns to rAndom MaTrices, operator aLgebRas and harMonIc AnaLySis on FRee groUps. cRM monograPh SerieS, 1. ameRiCAN maThematical Society, PRoVIDeNce, RI, 1992.
---
absTract: 'WE DeScRIbe the fiRsT-prInciPLEs desIgn aND sUbsequenT synthESiS oF a new maTeRial wiTh The SpeCific FUnctIonaliTies requIred fOR a solid-state-baSEd search for thE PeRMAnENt elEctRic dipole moMent OF the ElecTRoN. We SHow coMputaTiONaLLy that perovskite-strUcTure euRopiuM barium titanaTe should exHIBIt the reqUireD LaRGe and pressure-dEpendEnt ferroelECtric polArizaTion, locaL magnetic MOMents, and AbsEncE of MagNETiC ordering at liQUId heLiUm tempeRatUre. SubsEquEnt SynTheSiS and charaCterizatIoN oF EU$_{0.5}$BA$_{0.5}$Tio$_3$ ceraMIcs confiRm The PrEdiCted dESirablE propErtiEs.'
AuTHor:
- 'k. Z. RushcHAnSKIi'
- 'S. KAmBa'
- 'v. GoiAn'
- 'P. vaNěk'
- 'M. SAvinOV'
- 'J. PRokleškA'
- 'D. Nuzh | pojankowski, A constant re gression c harac ter iza ti on o f th e Marchenko-Pa s turlaw. Probab. Math. Sta tist. 3 6 (20 1 6) , no. 1, 137 – 14 5 .
K .Sz poj an k ow ski,Onthe Mat sumoto-Yor pr op erty in free pr obability, J. Math. Anal. Ap pl. 44 5( 1)( 2017) , 3 74–39 3.
D. Voicul escu, Sym me t ries o f some r e d uc ed f ree product $C^\a s t$ - algebras, in:Operat or Al g e bra s a nd their C on necti o ns with To p o l ogy and Ergodic T heory, 556– 5 88, Lectu re No t es inMathe ma t ics , Vol. 1132 , Sp ringer-Ve rlag,B erlin/N e w York, 1985.
D . V oicu l es cu , A dd i tio n o f c e rta in nonco mm ut ing r ando m v a riab les , J. Func t. Anal. 66 ( 198 6),n o.3, 32 3–346 .
D .Voicu lescu. Thean alogues of entr opyand of Fi she r’ s i nf ormat i on mea sur e i n freeprobabi l ity t h e o ry , I. Comm. Math. P hy s . 1 55 (1993 ), no. 1, 7 1 –92.
D. V oic ules c u , K.Dyke m aand A. N ica, [ * Fr ee random v ariabl es *]{ }.A non c ommu tative probabi litya pproach to fre e products wit h a p p li c atio nsto random m atri c es,oper a to r a l gebra s and h a rm o nic analysis on fre egroups . CRM Monograph Se ries, 1. A m e r ican Mat hema t ic a l Society, Pro viden ce, RI, 19 9 2.
---abstr act: 'We describe t he first -pr inc ipl esd e si gn and subseq u e nt s yn thesisofa new m ate ria l w ith t he specif ic funct io na li ti esrequi r ed for a s oli d- sta te-ba s ed sea rch f or t he p e rma nent el e ct r i c di po le mom ent o f the ele c tro n. We s how compu tat i onal ly t hat per ovskite-struc tu re europiu mbar ium ti t a nate sho uld exhibit the require d largeand pres sure -dependen t f erroel ect r ic pol arizat ion,lo cal m agnet i c m ome nt s, and abs e n ceof ma gn etic orderi ng at liquid heliu m te mperature. Su bse quen t sy nth e si s an dc har a c terization of E u$_{0.5}$B a$ _ {0 .5}$TiO$_3 $ ce ra mics co nfirm t he pr e dicteddesirable properti es .'
a u t hor :
- 'K. Z. Rushcha nskii'
-' S. Ka m ba '
- ' V.Goian'
- 'P . Van ěk'
-' M.Savin ov'
-'J . Prok leška '- 'D. Nu zh | pojankowski, A_constant regression_characterization of the Marchenko-Pastur_law. Probab._Math._Statist. 36_(2016),_no. 1, 137–145.
K. Szpojankowski,_On the Matsumoto-Yor_property in free probability,_J. Math. Anal. Appl. 445(1) (2017), 374–393.
D. Voiculescu,_Symmetries_of some reduced free product $C^\ast$-algebras, in: Operator Algebras and their Connections with Topology_and_Ergodic Theory,_556–588,_Lecture_Notes in Mathematics, Vol. 1132, Springer-Verlag,_Berlin/New York, 1985.
D. Voiculescu, Addition of_certain noncommuting_random variables, J. Funct. Anal. 66 (1986), no. 3, 323–346.
D. Voiculescu. The analogues_of_entropy and of_Fisher’s information measure in free probability theory, I. Comm. Math. Phys. 155_(1993), no. 1, 71–92.
D. Voiculescu, K. Dykema and A. Nica,_[*Free random variables*]{}._A_noncommutative_probability approach to free_products with applications to random matrices,_operator algebras and harmonic analysis on_free groups. CRM Monograph Series, 1. American_Mathematical Society, Providence, RI, 1992.
---
abstract:_'We describe the first-principles design_and subsequent_synthesis of a new material_with the specific_functionalities required_for a solid-state-based_search for the permanent electric dipole_moment of the_electron. We show computationally that perovskite-structure_europium_barium titanate should_exhibit_the_required large_and pressure-dependent ferroelectric_polarization,_local magnetic_moments,_and absence of magnetic ordering at_liquid_helium temperature. Subsequent synthesis and characterization of_Eu$_{0.5}$Ba$_{0.5}$TiO$_3$ ceramics confirm the_predicted_desirable properties.'
author:
- 'K. Z. Rushchanskii'
- 'S. Kamba'
-_'V. Goian'
- 'P. Vaněk'
- 'M. Savinov'
- 'J. Prokleška'
- 'D. Nuzh |
gamma_5)_{\alpha_{4}^{\prime}\alpha_{5}^{\prime}}
\delta_{\alpha_{3}^{\prime}\alpha_{3}}
\delta_{\alpha_{6}^{\prime}\alpha_{4}},
\label{NaiveBaryonBlocks}
%\end{aligned}$$ $$\begin{array}{ll}
\!\!\!\mbox{with} &
~[B_{1,\alpha_{1}}^{(0)}](\vec{x};~
\bm{\xi}_{P_{1}P_{2}P_{3}}^\prime)
=
[B_{1,\alpha_{1}}^{(0)}](\vec{x};~
\xi_{P_{1}}^{\prime}, \xi_{P_{2}}^{\prime}, \xi_{P_{3}}^{\prime}
)
=
\left\langle
%
B_{1,\alpha_{1}}(\vec{x})
%
~
\bar{q}_{B_{1},3}^{\prime}(\xi_{P_{3}}^{\prime})
\bar{q}_{B_{1},2}^{\prime}(\xi_{P_{2}}^{\prime})
\bar{q}_{B_{1},1}^{\prime}(\xi_{P_{1}}^{\prime})
%
\right\rangle,
\quad \mbox{and}
%
\\
%
&
~
[B_{2,\alpha_{2}}^{(0)}](\vec{y};~
\bm{\xi}_{P_{4}P_{5}P_{6}}^\prime)
=
[B_{2,\alpha_{2}}^{(0)}](\vec{y};~
\xi_{P_{4}}^{\prime}, \xi_{P_{5}}^{\prime}, \xi_{P_{6}}^{\prime}
)
=
\left\langle
%
B_{2,\alpha_{2}}(\vec{y})
%
~
\bar{q}_{B_{2},6}^{\prime}(\xi_{P_{6}}^{\prime})
\bar{q}_{B_{2},5}^{\prime}(\xi_{P_{5}}^{\prime})
\bar{q}_{B_{2},4}^{\prime}(\xi_{P_{4}}^{\prime})
%
\right\rangle,
%
\end{array}$$ | gamma_5)_{\alpha_{4}^{\prime}\alpha_{5}^{\prime } }
\delta_{\alpha_{3}^{\prime}\alpha_{3 } }
\delta_{\alpha_{6}^{\prime}\alpha_{4 } },
\label{NaiveBaryonBlocks }
% \end{aligned}$$ $ $ \begin{array}{ll }
\!\!\!\mbox{with } &
~[B_{1,\alpha_{1}}^{(0)}](\vec{x};~
\bm{\xi}_{P_{1}P_{2}P_{3}}^\prime)
=
[ B_{1,\alpha_{1}}^{(0)}](\vec{x};~
\xi_{P_{1}}^{\prime }, \xi_{P_{2}}^{\prime }, \xi_{P_{3}}^{\prime }
)
=
\left\langle
%
B_{1,\alpha_{1}}(\vec{x })
%
~
\bar{q}_{B_{1},3}^{\prime}(\xi_{P_{3}}^{\prime })
\bar{q}_{B_{1},2}^{\prime}(\xi_{P_{2}}^{\prime })
\bar{q}_{B_{1},1}^{\prime}(\xi_{P_{1}}^{\prime })
%
\right\rangle,
\quad \mbox{and }
%
\\
%
&
~
[ B_{2,\alpha_{2}}^{(0)}](\vec{y};~
\bm{\xi}_{P_{4}P_{5}P_{6}}^\prime)
=
[ B_{2,\alpha_{2}}^{(0)}](\vec{y};~
\xi_{P_{4}}^{\prime }, \xi_{P_{5}}^{\prime }, \xi_{P_{6}}^{\prime }
)
=
\left\langle
%
B_{2,\alpha_{2}}(\vec{y })
%
~
\bar{q}_{B_{2},6}^{\prime}(\xi_{P_{6}}^{\prime })
\bar{q}_{B_{2},5}^{\prime}(\xi_{P_{5}}^{\prime })
\bar{q}_{B_{2},4}^{\prime}(\xi_{P_{4}}^{\prime })
%
\right\rangle,
%
\end{array}$$ | gamla_5)_{\alpha_{4}^{\prime}\alpha_{5}^{\prime}}
\dtlta_{\alpha_{3}^{\prime}\alpkq_{3}}
\delte_{\alpha_{6}^{\pdime}\alphx_{4}},
\label{NaiveBaryonBlocks}
%\end{apitned}$$ $$\vegin{array}{ll}
\!\!\!\mbox{with} &
~[B_{1,\alpha_{1}}^{(0)}](\nec{x};~
\bn{\zi}_{P_{1}P_{2}P_{3}}^\prime)
=
[B_{1,\alika_{1}}^{(0)}](\vec{s};~
\xn_{P_{1}}^{\'rime}, \xi_{P_{2}}^{\prime}, \wi_{P_{3}}^{\prime}
)
=
\left\langle
%
B_{1,\alpha_{1}}(\vec{x})
%
~
\har{q}_{B_{1},3}^{\prime}(\xi_{P_{3}}^{\ptime})
\bar{s}_{B_{1},2}^{\prime}(\xi_{P_{2}}^{\prime})
\bar{q}_{B_{1},1}^{\prime}(\xi_{P_{1}}^{\prije})
%
\rpght\rangle,
\quad \mnox{and}
%
\\
%
&
~
[B_{2,\alpha_{2}}^{(0)}](\vec{y};~
\bl{\xi}_{P_{4}O_{5}P_{6}}^\prime)
=
[B_{2,\alpha_{2}}^{(0)}](\gec{y};~
\xi_{P_{4}}^{\ptjme}, \zi_{P_{5}}^{\prime}, \xi_{P_{6}}^{\orime}
)
=
\left\langle
%
B_{2,\aupha_{2}}(\vzc{y})
%
~
\bqr{q}_{H_{2},6}^{\krime}(\xi_{P_{6}}^{\primx})
\bag{q}_{B_{2},5}^{\prime}(\xi_{P_{5}}^{\prlke})
\bdr{q}_{B_{2},4}^{\prike}(\xi_{P_{4}}^{\prime})
%
\rmght\eangle,
%
\end{array}$$ | gamma_5)_{\alpha_{4}^{\prime}\alpha_{5}^{\prime}} \delta_{\alpha_{3}^{\prime}\alpha_{3}} \delta_{\alpha_{6}^{\prime}\alpha_{4}}, \label{NaiveBaryonBlocks} %\end{aligned}$$ $$\begin{array}{ll} \!\!\!\mbox{with} \bm{\xi}_{P_{1}P_{2}P_{3}}^\prime) [B_{1,\alpha_{1}}^{(0)}](\vec{x};~ \xi_{P_{1}}^{\prime}, \xi_{P_{3}}^{\prime} ) = \bar{q}_{B_{1},3}^{\prime}(\xi_{P_{3}}^{\prime}) \bar{q}_{B_{1},1}^{\prime}(\xi_{P_{1}}^{\prime}) % \right\rangle, \mbox{and} % \\ & ~ [B_{2,\alpha_{2}}^{(0)}](\vec{y};~ \bm{\xi}_{P_{4}P_{5}P_{6}}^\prime) = [B_{2,\alpha_{2}}^{(0)}](\vec{y};~ \xi_{P_{5}}^{\prime}, \xi_{P_{6}}^{\prime} ) = \left\langle % B_{2,\alpha_{2}}(\vec{y}) % ~ \bar{q}_{B_{2},6}^{\prime}(\xi_{P_{6}}^{\prime}) \bar{q}_{B_{2},5}^{\prime}(\xi_{P_{5}}^{\prime}) \bar{q}_{B_{2},4}^{\prime}(\xi_{P_{4}}^{\prime}) % % \end{array}$$ | gamma_5)_{\alpha_{4}^{\prime}\alpha_{5}^{\prime}}
\Delta_{\alpha_{3}^{\Prime}\AlpHa_{3}}
\dElTa_{\alPha_{6}^{\pRime}\alpha_{4}},
\label{nAivebaryonBlocks}
%\end{aligned}$$ $$\Begin{ArRAy}{ll}
\!\!\!\MBoX{with} &
~[b_{1,\alpha_{1}}^{(0)}](\vEC{x};~
\BM{\Xi}_{P_{1}p_{2}P_{3}}^\PrIme)
=
[b_{1,\aLPhA_{1}}^{(0)}](\vec{x};~
\Xi_{P_{1}}^{\Prime}, \xi_{p_{2}}^{\prime}, \xi_{P_{3}}^{\pRimE}
)
=
\lEft\langle
%
B_{1,\alPHa_{1}}(\Vec{x})
%
~
\bar{q}_{B_{1},3}^{\pRimE}(\xi_{P_{3}}^{\prime})
\bar{Q}_{B_{1},2}^{\pRime}(\xi_{p_{2}}^{\pRimE})
\Bar{q}_{B_{1},1}^{\PriMe}(\xi_{P_{1}}^{\Prime})
%
\rIGht\ranGle,
\quad \mbOx{ANd}
%
\\
%
&
~
[B_{2,\alpHA_{2}}^{(0)}](\vec{y};~
\bm{\XI}_{p_{4}P_{5}p_{6}}^\priMe)
=
[B_{2,\alpha_{2}}^{(0)}](\vec{y};~
\xi_{P_{4}}^{\prIMe}, \XI_{P_{5}}^{\prime}, \xi_{P_{6}}^{\primE}
)
=
\left\lAnGLe
%
b_{2,\ALphA_{2}}(\veC{y})
%
~
\bar{q}_{B_{2},6}^{\priMe}(\Xi_{P_{6}}^{\prIMe})
\bar{q}_{B_{2},5}^{\PRiME}(\XI_{P_{5}}^{\pRIme})
\bar{q}_{B_{2},4}^{\prime}(\Xi_{P_{4}}^{\prime})
%
\rigHT\raNgle,
%
\enD{aRraY}$$ | gamma_5)_{\alpha_{4}^{\pri me}\alpha_ {5}^{ \pr ime }}
\d elta _{\alpha_{3}^{ \ prim e}\alpha_{3}}
\delta_ {\alp ha _ {6}^ { \p rime} \alpha_ { 4} } ,
\ la be l{N ai v eB aryon Blo cks}
%\ end{aligne d}$ $$$\begin{arr a y} {ll}
\!\ !\! \mbox{with}&
~[B_{ 1, \al p ha_{1 }}^ {(0)} ](\vec { x};~
\b m{ \ xi}_{P _ {1}P_{2 } P _{ 3}}^ \prime)
=
[B _ {1 , \alpha_{1}}^{( 0)}](\ ve c {x } ; ~
\ xi_{P_{1}} ^{ \prim e }, \xi_ { P_ { 2 } }^{ \ prime}, \xi_{ P_{3}}^{\pr i me}
)
=
\left\ lang le
% B_{ 1 ,\alpha _ {1}}(\v ec{x})
% ~ \ bar {q } _{B _ {1 },3 } ^{\ prime}(\ xi _{ P_{3} }^{\ p r i m e}) \bar {q}_{ B_{1},2}^{\pr ime }(\x i _{P _{2}} ^{\pr ime} ) \b ar{q}_ {B_{1 }, 1}^{\prime}(\xi _{P_ {1}}^{\pr ime })
%
\ri g ht\ran gle ,
\quad \ mbox{an d } % \ \ %
&
~
[B_{2,\a lp h a _{ 2}}^{(0) }](\ve c {y }; ~
\bm{\ xi }_{ P_{4 } P _{5}P _{6} } ^\ prime) =
[ B_ {2 ,\alpha _{ 2}}^{( 0) }]( \ve c{y}; ~
\xi_{P _{4}}^{\ prime } , \xi_{P_{5}}^ { \prime}, \xi_ { P_ { 6 }} ^ {\pr ime }
)
=
\ left\ la n gl e
%
B_{2,\alp ha _{2}}( \vec{ y})
%
~
\bar{q } _ { B_{2},6} ^{\p r im e }(\xi_{P_{6}}^ {\pri me})
\b a r{q}_{B_ {2},5 }^{\prim e}(\xi_{P _ { 5}}^{\pr ime }) \ba r { q} _{B_{2},4}^{\ p r ime} (\ xi_{P_{ 4}} ^{\prim e})
% \r ight\rang le,
%
\ en d{ ar ra y}$ $ | gamma_5)_{\alpha_{4}^{\prime}\alpha_{5}^{\prime}}
\delta_{\alpha_{3}^{\prime}\alpha_{3}}
_\delta_{\alpha_{6}^{\prime}\alpha_{4}},
\label{NaiveBaryonBlocks}
%\end{aligned}$$_$$\begin{array}{ll}
\!\!\!\mbox{with} &
_ ~[B_{1,\alpha_{1}}^{(0)}](\vec{x};~
__ __ _\bm{\xi}_{P_{1}P_{2}P_{3}}^\prime)
_=
[B_{1,\alpha_{1}}^{(0)}](\vec{x};~
_ \xi_{P_{1}}^{\prime}, \xi_{P_{2}}^{\prime},_\xi_{P_{3}}^{\prime}
_ __ ___ )_
=
\left\langle_
_ %
B_{1,\alpha_{1}}(\vec{x})
%
__ ~
_ \bar{q}_{B_{1},3}^{\prime}(\xi_{P_{3}}^{\prime})
\bar{q}_{B_{1},2}^{\prime}(\xi_{P_{2}}^{\prime})
\bar{q}_{B_{1},1}^{\prime}(\xi_{P_{1}}^{\prime})
_ %
\right\rangle,
\quad \mbox{and}
_%
\\
%
_&
_~
_[B_{2,\alpha_{2}}^{(0)}](\vec{y};~
\bm{\xi}_{P_{4}P_{5}P_{6}}^\prime)
_ =
[B_{2,\alpha_{2}}^{(0)}](\vec{y};~
_\xi_{P_{4}}^{\prime}, \xi_{P_{5}}^{\prime}, \xi_{P_{6}}^{\prime}
_ _ _ _ _)
=
_\left\langle
_ %
_ B_{2,\alpha_{2}}(\vec{y})
_ %
~
_ \bar{q}_{B_{2},6}^{\prime}(\xi_{P_{6}}^{\prime})
_ \bar{q}_{B_{2},5}^{\prime}(\xi_{P_{5}}^{\prime})
\bar{q}_{B_{2},4}^{\prime}(\xi_{P_{4}}^{\prime})
__ %
__\right\rangle,
%
_\end{array}$$ |
km} g^{mj}$ we have then that
$$\begin{aligned}
(\Delta q)_{{\gamma}}&= -\Delta_g q_{{\gamma}} + {\partial}_{{\gamma}} (g^{ij}\Gamma^k_{ji}) q_k + g^{ik}{\partial}_{{\gamma}} g_{km} g^{mj}{\partial}_j q_i + g^{ij} \Gamma^{k}_{j{\gamma}} {\partial}_i q_k - g^{ij} \Gamma^{k}_{j{\gamma}} {\partial}_k q_i\end{aligned}$$
Finally observe that
$$\begin{aligned}
g^{ik}{\partial}_{{\gamma}} g_{km} g^{mj}{\partial}_j q_i + g^{ij} \Gamma^{k}_{j{\gamma}} {\partial}_i q_k - g^{ij} \Gamma^{k}_{j{\gamma}} {\partial}_k q_i = 2g^{ij}\Gamma^{k}_{j {\gamma}} {\partial}_{i}q_{k}
\end{aligned}$$
Therefore
$$\begin{aligned}
(\Delta q)_{{\gamma}}&= -\Delta_g q_{{\gamma}} + 2g^{ij}\Gamma^{k}_{j {\gamma}} {\partial}_{i}q_{k}+{\partial}_{{\gamma}}(g^{ij}\Gamma^k_{ij})q_k \end{aligned}$$
which is exactly .
Combining the results of the previous two lemmas with Proposition \[wave q\] gives us a system of nonlinear wave equations for the components of $q$. The following Proposition is the main result of this section and will be used to prove a priori estimates for the differential, $du$, of a wave map $u$.
\[main eqn for q\] Let $u: ({\widetilde}{M}, \eta) \to (N,h)$ be a smooth wave map. Let $q=du$ be the representation of $du$ in the Coulomb frame, $e$ as in. Then, the components of $q$ satisfy the following system of variable coefficient wave equations:
$$\begin{gathered}
\label{scalar eqn for q}
\ddot{q}_{{\gamma}} -\Delta_g q_{{\gamma}} + 2g^{ij}\Gamma^{k}_{j {\gamma}} {\partial}_{i}q_{k}+{\partial}_{{\ | km } g^{mj}$ we have then that
$ $ \begin{aligned }
(\Delta q)_{{\gamma}}&= -\Delta_g q_{{\gamma } } + { \partial}_{{\gamma } } (g^{ij}\Gamma^k_{ji }) q_k + g^{ik}{\partial}_{{\gamma } } g_{km } g^{mj}{\partial}_j q_i + g^{ij } \Gamma^{k}_{j{\gamma } } { \partial}_i q_k - g^{ij } \Gamma^{k}_{j{\gamma } } { \partial}_k q_i\end{aligned}$$
Finally observe that
$ $ \begin{aligned }
g^{ik}{\partial}_{{\gamma } } g_{km } g^{mj}{\partial}_j q_i + g^{ij } \Gamma^{k}_{j{\gamma } } { \partial}_i q_k - g^{ij } \Gamma^{k}_{j{\gamma } } { \partial}_k q_i = 2g^{ij}\Gamma^{k}_{j { \gamma } } { \partial}_{i}q_{k }
\end{aligned}$$
Therefore
$ $ \begin{aligned }
(\Delta q)_{{\gamma}}&= -\Delta_g q_{{\gamma } } + 2g^{ij}\Gamma^{k}_{j { \gamma } } { \partial}_{i}q_{k}+{\partial}_{{\gamma}}(g^{ij}\Gamma^k_{ij})q_k \end{aligned}$$
which is precisely .
compound the results of the previous two lemmas with Proposition \[wave q\ ] yield us a system of nonlinear wave equations for the component of $ q$. The following Proposition is the chief result of this section and will be use to prove a priori estimates for the differential, $ du$, of a wave function $ u$.
\[main eqn for q\ ] Let $ u: ({ \widetilde}{M }, \eta) \to (N, h)$ be a smooth wave map. Let $ q = du$ be the representation of $ du$ in the Coulomb frame, $ e$ as in. Then, the component of $ q$ satisfy the following system of variable coefficient wave equation:
$ $ \begin{gathered }
\label{scalar eqn for q }
\ddot{q}_{{\gamma } } -\Delta_g q_{{\gamma } } + 2g^{ij}\Gamma^{k}_{j { \gamma } } { \partial}_{i}q_{k}+{\partial}_{{\ | km} h^{mj}$ we have then that
$$\begln{aligned}
(\Delta q)_{{\yqmma}}&= -\Dxlta_g q_{{\famma}} + {\oartial}_{{\gamma}} (g^{ij}\Gamma^k_{ji}) q_k + t^{ik}{\paetial}_{{\gamma}} g_{km} g^{mj}{\partkal}_j q_i + h^{ij} \Gammq^{k}_{j{\gemma}} {\partial}_i q_k - g^{ij} \Gamma^{k}_{j{\gzlma}} {\'actial}_k q_i\end{alinned}$$
Finally observe that
$$\bagkn{cligned}
g^{ik}{\partial}_{{\gamma}} g_{km} g^{mj}{\partiwl}_j q_i + g^{lj} \Gamma^{k}_{j{\gammw}} {\pagtyal}_i q_k - g^{ij} \Gamma^{k}_{j{\gamma}} {\partial}_k q_i = 2g^{ij}\Gamka^{k}_{j {\gamma}} {\pattial}_{i}q_{k}
\end{aligned}$$
Therefoge
$$\behin{aligned}
(\Delta q)_{{\hamma}}&= -\Delta_t q_{{\gwnma}} + 2g^{ij}\Gammx^{k}_{j {\gamma}} {\kaxtial}_{i}q_{k}+{\parjial}_{{\gamma}}(g^{ij}\Gamma^k_{ij})q_k \end{alignea}$$
whick is exactli .
Doltining the cesultf of the preypous twm lemmax with Proposibion \[xave q\] gives us a system mf nonlinear wave gquations xox the components of $q$. Tye fonlowhng Oeopusiuioi ia the laii result of this sectiin and will be used tj prove a priodi estymwtes for the differential, $du$, of a wave kap $u$.
\[main eqn for q\] Let $u: ({\qidetilde}{M}, \eta) \to (N,h)$ he a smooeh wave map. Let $q=du$ be the representation of $du$ it the Zouooic fgame, $e$ as in. Then, the components of $q$ satisfy fht fpllowing systei of variabke cpgfficient wave equatnkna:
$$\begin{gathered}
\labep{scalar eqn dor q}
\ddjt{q}_{{\gsmma}} -\Delta_g q_{{\gamma}} + 2g^{ij}\Gammq^{k}_{j {\gamma}} {\pagtiao}_{i}q_{k}+{\partial}_{{\ | km} g^{mj}$ we have then that $$\begin{aligned} -\Delta_g + {\partial}_{{\gamma}} q_k + g^{ik}{\partial}_{{\gamma}} \Gamma^{k}_{j{\gamma}} q_k - g^{ij} {\partial}_k q_i\end{aligned}$$ Finally that $$\begin{aligned} g^{ik}{\partial}_{{\gamma}} g_{km} g^{mj}{\partial}_j q_i g^{ij} \Gamma^{k}_{j{\gamma}} {\partial}_i q_k - g^{ij} \Gamma^{k}_{j{\gamma}} {\partial}_k q_i = 2g^{ij}\Gamma^{k}_{j {\gamma}} {\partial}_{i}q_{k} Therefore $$\begin{aligned} (\Delta q)_{{\gamma}}&= -\Delta_g q_{{\gamma}} + 2g^{ij}\Gamma^{k}_{j {\gamma}} {\partial}_{i}q_{k}+{\partial}_{{\gamma}}(g^{ij}\Gamma^k_{ij})q_k \end{aligned}$$ which is . the of previous two lemmas with Proposition \[wave q\] gives us a system of nonlinear wave equations for components of $q$. The following Proposition is the result of this section will be used to prove priori for the $du$, a map $u$. \[main for q\] Let $u: ({\widetilde}{M}, \eta) \to (N,h)$ be a smooth wave map. Let $q=du$ be the of $du$ Coulomb frame, as Then, components of $q$ following system of variable coefficient wave eqn for q} \ddot{q}_{{\gamma}} -\Delta_g q_{{\gamma}} + 2g^{ij}\Gamma^{k}_{j {\partial}_{i}q_{k}+{\partial}_{{\ | km} g^{mj}$ we have then that
$$\begin{aLigned}
(\DeltA q)_{{\gamMa}}&= -\DEltA_g Q_{{\gamMa}} + {\paRtial}_{{\gamma}} (g^{ij}\GAMma^k_{Ji}) q_k + g^{ik}{\partial}_{{\gamma}} g_{km} G^{mj}{\paRtIAl}_j q_I + G^{iJ} \GammA^{k}_{j{\gammA}} {\PaRTIal}_I q_K - g^{Ij} \GAmMA^{k}_{J{\gammA}} {\paRtial}_k q_I\end{aligneD}$$
FiNaLly observe thAT
$$\bEgin{aligneD}
g^{iK}{\partial}_{{\gammA}} g_{kM} g^{mj}{\paRtIal}_J Q_i + g^{ij} \gamMa^{k}_{j{\gAmma}} {\paRTial}_i q_K - g^{ij} \Gamma^{K}_{j{\GAmma}} {\paRTial}_k q_i = 2G^{IJ}\GAmma^{K}_{j {\gamma}} {\partial}_{i}q_{k}
\ENd{ALigned}$$
ThereforE
$$\begin{AlIGnED}
(\delTa q)_{{\Gamma}}&= -\Delta_G q_{{\Gamma}} + 2G^{Ij}\Gamma^{K}_{J {\gAMMA}} {\paRTial}_{i}q_{k}+{\partiaL}_{{\gamma}}(g^{ij}\GaMMa^k_{Ij})q_k \enD{aLigNEd}$$
whicH is exAcTLy .
COmbining the ResuLts of the pReviouS Two lemmAS with PrOpositIon \[WavE q\] giVEs Us A syStEM of NOnLinEAr wAve equatIoNs For thE comPONENts oF $q$. THe foLlowiNg Proposition Is tHe maIN reSult oF this SectIoN and wIll be uSed to PrOve a priori estimAtes For the difFerEnTiaL, $dU$, of a wAVe map $u$.
\[MaiN eqN for q\] LeT $u: ({\widetILde}{m}, \eTA) \TO (N,H)$ be a smooth wave map. LEt $Q=DU$ bE the reprEsentaTIoN oF $Du$ in the COuLomB fraME, $E$ as in. then, THe ComponenTs of $q$ sATiSfY the folLoWing syStEm oF vaRiablE CoefFicienT wave equAtionS:
$$\Begin{gathered}
\lABel{scalar eqn fOR q}
\DDOt{Q}_{{\GammA}} -\DeLta_g q_{{\gamma}} + 2g^{Ij}\GaMMa^{k}_{j {\GammA}} {\PaRtiAL}_{i}q_{k}+{\pArtiaL}_{{\ | km} g^{mj}$ we have then t hat
$$\be gin{a lig ned }(\De ltaq)_{{\gamma}}& = -\D elta_g q_{{\gamma}} + {\pa rt i al}_ { {\ gamma }} (g^{ i j} \ G amm a^ k_ {ji }) q_ k + g^{ ik}{\pa rtial}_{{\ gam ma }} g_{km} g^ { mj }{\partial }_j q_i + g^{ij } \ Gamma^ {k }_{ j {\gam ma} } {\p artial } _i q_k - g^{ij} \ G amma^{ k }_{j{\g a m ma }} { \partial}_k q_i\e n d{ a ligned}$$
Fin ally o bs e rv e tha t
$$\begin{a li gned} g^{ik} { \p a r t ial } _{{\gamma}} g _{km} g^{mj } {\p artial }_ j q _ i + g^ {ij}\G a mma ^{k}_{j{\ga mma} } {\parti al}_iq _k - g^ { ij} \Ga mma^{k }_{ j{\ gamm a }} { \pa rt i al} _ kq_i = 2 g^{ij}\G am ma ^{k}_ {j { \ g a m ma}} {\ part ial}_ {i}q_{k}
\en d{a lign e d}$ $
Th erefo re
$$ \begi n{alig ned} ( \Delta q)_{{\ga mma} }&= -\Del ta_ gq_{ {\ gamma } } + 2g ^{i j}\ Gamma^{ k}_{j { \ gam ma } } {\ partial}_{i}q_{k}+ {\ p a rt ial}_{{\ gamma} } (g ^{ i j}\Gamma ^k _{i j})q _ k \end {ali g ne d}$$
wh ich is ex ac tly .
C ombini ng th e r esult s ofthe pr evious t wo le m mas with Propo s ition \[waveq \] g iv e s us asystem of n onli n earwave eq uat i ons f or th ec om p onents of $q$. Thefo llowin g Pro position is t he main re s u l t of thi s se c ti o n and will beusedto prove a priori e stima tes forthe diffe r e ntial, $ du$ , o f a wa v e m ap $u$.
\[ma i n eqn f or q\]Let $u: ({ \wi det ild e}{ M} , \eta) \ to (N,h) $be a s moo th wa v e map. L et $q =d u$be th e repre senta tion o f$ du$ in the Co u l ombfr am e, $ e$as in.Then , th e compo nents of$q$ sati sf ythe fol lowing system o f variable c oef ficien t wave equ ations:
$$\begin{gathe r ed}
\la bel {scal ar e qn for q}
\ddot{ q}_ { {\gamm a}} -\ Delta _g q_ { { \gamm a } }+ 2 g^ {ij}\Gamma ^ { k}_ {j {\ ga mma} } {\par tial}_{i}q_{k}+{\p a rti al}_{{\ | km} g^{mj}$_we have_then that
$$\begin{aligned}
(\Delta q)_{{\gamma}}&= -\Delta_g_q_{{\gamma}} +__{\partial}_{{\gamma}} (g^{ij}\Gamma^k_{ji})_q_k_+ g^{ik}{\partial}_{{\gamma}}_g_{km} g^{mj}{\partial}_j q_i_+ g^{ij} \Gamma^{k}_{j{\gamma}} {\partial}_i_q_k - g^{ij}_\Gamma^{k}_{j{\gamma}}_{\partial}_k q_i\end{aligned}$$
Finally observe that
$$\begin{aligned}
g^{ik}{\partial}_{{\gamma}} g_{km} g^{mj}{\partial}_j q_i + g^{ij} \Gamma^{k}_{j{\gamma}} {\partial}_i q_k -_g^{ij}_\Gamma^{k}_{j{\gamma}} {\partial}_k_q_i_=_2g^{ij}\Gamma^{k}_{j {\gamma}} {\partial}_{i}q_{k}
\end{aligned}$$
Therefore
$$\begin{aligned}
(\Delta_q)_{{\gamma}}&= -\Delta_g q_{{\gamma}} + 2g^{ij}\Gamma^{k}_{j_{\gamma}} {\partial}_{i}q_{k}+{\partial}_{{\gamma}}(g^{ij}\Gamma^k_{ij})q_k_\end{aligned}$$
which is exactly .
Combining the results of the_previous_two lemmas with_Proposition \[wave q\] gives us a system of nonlinear_wave equations for the components of_$q$. The following_Proposition_is_the main result of_this section and will be used_to prove a priori estimates for_the differential, $du$, of a wave map_$u$.
\[main eqn for q\] Let $u:_({\widetilde}{M}, \eta) \to (N,h)$ be_a smooth_wave map. Let $q=du$ be_the representation of_$du$ in_the Coulomb frame,_$e$ as in. Then, the components_of $q$ satisfy_the following system of variable coefficient_wave_equations:
$$\begin{gathered}
\label{scalar eqn for_q}_
_\ddot{q}_{{\gamma}} -\Delta_g_q_{{\gamma}} + 2g^{ij}\Gamma^{k}_{j_{\gamma}}_{\partial}_{i}q_{k}+{\partial}_{{\ |
&& \mal_1(x,y,y)=f(\mal(x,y,y))=f(x)=x\\
&& \mal_1(y,y,x)=f(\mal(y,y,x))=f(x)=x,\end{aligned}$$ for any $x,y\in\zA_1$ with $x\eqc{\sg_1}y$.
The results of [@Maroti10:tree] imply the following. Let $\gA$ be a class of similar finite algebras closed under subalgebras, and retracts via idempotent unary polynomials. Suppose that $\gA$ has a term operation $f$ satisfying the following conditions for some $\zB\in\gA$:
- $f(x,f(x,y))=f(x,y)$ for any $x,y\in\zB$;
- for each $a\in\zB$ the mapping $x\mapsto f(a,x)$ is not surjective;
- the set $C$ of $a\in\zB$ such that $x\mapsto f(x,a)$ is surjective generates a proper subalgebra of $\zB$.
Then $\CSP(\gA)$ is polynomial time reducible to $\CSP(\gA-\{\zB\})$.
By Lemma \[lem:dot-inequality\] the operation $\cdot$ of the class of SBM algebras from $\gA$ satisfies condition (1). If the operation $a\cdot x$ is surjective for some $a$, then $a\le x$ for all $x\in\zB$. Therefore the only case when condition (2) is not satisfied is when $\zB$ has a minimal element. Finally, condition (3) is satisfied whenever $\zB$ is not a Mal’tsev algebra. Therefore, choosing $\zB$ to be a maximal (in terms of cardinality) algebra from $\gA$ satisfying conditions (1)–(3) we may only consider instances of $\CSP(\gA)$, in which every domain has a minimal element or is a Mal’tsev algebra.
\[cor:maroti\] Every instance $\cP\in\CSP(\gA)$ can be reduced in polynomial time to polynomially many instances over algebras each of which either is Mal’tsev or has | & & \mal_1(x, y, y)=f(\mal(x, y, y))=f(x)=x\\
& & \mal_1(y, y, x)=f(\mal(y, y, x))=f(x)=x,\end{aligned}$$ for any $ x, y\in\zA_1 $ with $ x\eqc{\sg_1}y$.
The results of [ @Maroti10: tree ] imply the following. Let $ \gA$ be a course of exchangeable finite algebras closed under subalgebras, and retracts via idempotent unary polynomial. Suppose that $ \gA$ have a term mathematical process $ f$ satisfy the following conditions for some $ \zB\in\gA$:
- $ f(x, f(x, y))=f(x, y)$ for any $ x, y\in\zB$;
- for each $ a\in\zB$ the mapping $ x\mapsto f(a, x)$ is not surjective;
- the set $ C$ of $ a\in\zB$ such that $ x\mapsto f(x, a)$ is surjective beget a proper subalgebra of $ \zB$.
Then $ \CSP(\gA)$ is polynomial time reducible to $ \CSP(\gA-\{\zB\})$.
By Lemma \[lem: dot - inequality\ ] the mathematical process $ \cdot$ of the course of SBM algebras from $ \gA$ satisfies condition (1). If the process $ a\cdot x$ is surjective for some $ a$, then $ a\le x$ for all $ x\in\zB$. Therefore the only event when condition (2) is not satisfied is when $ \zB$ has a minimal chemical element. Finally, condition (3) is satisfied whenever $ \zB$ is not a Mal’tsev algebra. consequently, choosing $ \zB$ to be a maximal (in term of cardinality) algebra from $ \gA$ satisfying conditions (1)–(3) we may only study instances of $ \CSP(\gA)$, in which every domain has a minimal element or is a Mal’tsev algebra.
\[cor: maroti\ ] Every instance $ \cP\in\CSP(\gA)$ can be reduce in polynomial time to polynomially many instances over algebras each of which either is Mal’tsev or has | && \map_1(x,y,y)=f(\mal(x,y,y))=f(x)=x\\
&& \mal_1(y,y,x)=f(\mau(y,y,x))=f(x)=x,\end{aligngd}$$ for aiy $x,y\in\aA_1$ with $b\eqc{\sg_1}y$.
The results of [@Maroti10:vree] implt the following. Let $\gA$ be a clads of sinilac finite algebras closed under dubanjebras, and retrscts via igempotent unarf oopynomials. Suppose that $\gA$ has a teri operayiln $f$ satisfyind tht fjllospnn conditions for some $\zB\in\gA$:
- $f(s,f(x,y))=f(x,y)$ for any $x,y\in\aB$;
- for each $a\in\zB$ the maopinh $x\mapsto f(a,x)$ is nlt surjectice;
- rhe set $C$ of $a\in\zB$ such that $x\mapsjo f(x,a)$ is surjective generates a propzr subalgebtc of $\eB$.
Then $\CSP(\gA)$ is pjlynomial time reduchble to $\CSP(\gA-\{\zB\})$.
By Lemka \[lxm:dor-inequality\] the operavion $\cdot$ of the clafs of SBM amgebras from $\gA$ sqtusfiev cotditkin (1). If tie kperatlon $a\cdot x$ ia surjectivw for some $a$, then $a\ke q$ for all $x\in\aB$. Thewesore the only case when condition (2) is nmt aatisfied is when $\zB$ haw a minimal element. Flnally, cogdition (3) is satisfied whenever $\zB$ is not a Mal’tsee algxbfa. Uhcvefofw, fhoosing $\zB$ to be a maximal (in terms of cardigzlotj) algebra from $\gA$ satisfying clncytions (1)–(3) we mai only eknaider instances of $\CSP(\gA)$, yn whuch every domsin has a minimal element oe is a Mal’tsvv aogebra.
\[cor:maroti\] Evzry instance $\cP\im\CSP(\gS)$ can be reduced in polvnomiam time to pllynomialmh many instances ovvr angebras txch of which eithqr is Mal’vsev pr has | && \mal_1(x,y,y)=f(\mal(x,y,y))=f(x)=x\\ && \mal_1(y,y,x)=f(\mal(y,y,x))=f(x)=x,\end{aligned}$$ for any $x,y\in\zA_1$ The of [@Maroti10:tree] the following. Let similar algebras closed under and retracts via unary polynomials. Suppose that $\gA$ has term operation $f$ satisfying the following conditions for some $\zB\in\gA$: - $f(x,f(x,y))=f(x,y)$ for $x,y\in\zB$; - for each $a\in\zB$ the mapping $x\mapsto f(a,x)$ is not surjective; - set of such $x\mapsto f(x,a)$ is surjective generates a proper subalgebra of $\zB$. Then $\CSP(\gA)$ is polynomial time reducible $\CSP(\gA-\{\zB\})$. By Lemma \[lem:dot-inequality\] the operation $\cdot$ of class of SBM algebras $\gA$ satisfies condition (1). If operation x$ is for $a$, $a\le x$ for $x\in\zB$. Therefore the only case when condition (2) is not satisfied is when $\zB$ has a minimal Finally, condition satisfied whenever is a algebra. Therefore, choosing be a maximal (in terms of $\gA$ satisfying conditions (1)–(3) we may only consider of $\CSP(\gA)$, which every domain has a minimal or is a Mal’tsev algebra. \[cor:maroti\] Every instance can be reduced in polynomial time to polynomially many instances over algebras each of which Mal’tsev or has | && \mal_1(x,y,y)=f(\mal(x,y,y))=f(x)=x\\
&& \mal_1(y,y,x)=f(\maL(y,y,x))=f(x)=x,\end{AlignEd}$$ fOr aNy $X,y\in\ZA_1$ wiTh $x\eqc{\sg_1}y$.
The reSUlts Of [@Maroti10:tree] imply the foLlowiNg. lEt $\gA$ BE a Class Of similAR fINIte AlGeBraS cLOsEd undEr sUbalgebRas, and retrActS vIa idempotent UNaRy polynomiAls. suppose that $\ga$ haS a term OpEraTIon $f$ sAtiSfyinG the foLLowing ConditionS fOR some $\zb\In\gA$:
- $f(x,f(X,Y))=F(x,Y)$ for Any $x,y\in\zB$;
- for each $a\IN\zb$ The mapping $x\mapSto f(a,x)$ Is NOt SURjeCtiVe;
- the set $C$ oF $a\In\zB$ sUCh that $x\MApSTO F(x,a)$ IS surjective geNerates a proPEr sUbalgeBrA of $\Zb$.
Then $\CsP(\gA)$ iS pOLynOmial time reDuciBle to $\CSP(\ga-\{\zB\})$.
By LEMma \[lem:dOT-inequaLity\] thE opEraTion $\CDoT$ oF thE cLAss OF SbM aLGebRas from $\ga$ sAtIsfieS conDITIOn (1). If The OperAtion $A\cdot x$ is surjeCtiVe foR SomE $a$, theN $a\le x$ For aLl $X\in\zB$. therefOre thE oNly case when condItioN (2) is not satIsfIeD is WhEn $\zB$ hAS a miniMal EleMent. FinAlly, conDItiOn (3) IS SAtIsfied whenever $\zB$ is NoT A maL’tsev algEbra. ThEReFoRE, choosinG $\zb$ to Be a mAXImal (iN terMS oF cardinaLity) alGEbRa From $\gA$ sAtIsfyinG cOndItiOns (1)–(3) we MAy onLy consIder instAnces OF $\CSP(\gA)$, in which eVEry domain has a MInIMAl ELemeNt oR is a Mal’tsev AlgeBRa.
\[coR:marOTi\] eveRY instAnce $\cp\iN\cSp(\GA)$ can be reduced in polYnOmial tIme to Polynomially mAny instancES OVer algebRas eACh OF which either is mal’tsEv or has | && \mal_1(x,y,y)=f(\mal(x, y,y))=f(x) =x\\&&\ma l_ 1(y, y,x) =f(\mal(y,y,x) ) =f(x )=x,\end{aligned}$$ fo r any $ x ,y\i n \z A_1$with $x \ eq c { \sg _1 }y $.
T h eresul tsof [@Ma roti10:tre e]im ply the foll o wi ng. Let $\ gA$ be a classofsimila rfin i te al geb ras c losedu nder s ubalgebra s, and re t racts v i a i demp otent unary polyn o mi a ls. Suppose th at $\g A$ ha s a t erm operation $ f$ sa t isfying th e f oll o wing conditio ns for some $\z B\in\g A$ :
- $f( x,f(x ,y ) )=f (x,y)$ forany$x,y\in\z B$;
- for e a ch $a\i n\zB$the ma ppin g $ x\ map st o f( a ,x )$i s n ot surje ct iv e;
- t h e s et $ C$of $ a\in\ zB$ such that $x \map s tof(x,a )$ is sur je ctive gener atesaproper subalgeb ra o f $\zB$.
Th en $\ CS P(\gA ) $ is p oly nom ial tim e reduc i ble t o $ \C SP(\gA-\{\zB\})$.
B y Le mma \[le m:dot- i ne qu a lity\] t he op erat i o n $\c dot$ of the cla ss ofS BM a lgebras f rom $\ gA $ s ati sfies cond ition(1). Ifthe o p eration $a\cdo t x$ is surjec t iv e fo r som e $ a$, then $a \lex $ fo r al l $ x\i n \zB$. Ther ef o re the only case whenco nditio n (2) is not satis fied is wh e n $\zB$ ha s am in i mal element. F inall y, conditi o n (3) is sati sfied wh enever $\ z B $ is not aMal ’ts eva l ge bra. Therefor e , cho os ing $\z B$to be a ma xim al(in t erms of c ardinali ty )al ge bra from $\gA$ sa ti sfy in g c ondit i ons (1 )–(3) wema yo nly consid e ri n stan ce sof $ \CS P( \gA)$ , in whi ch ever y domainhas a mi ni ma l eleme nt or is a Ma l’ tsev algeb ra .
\[cor: m a roti\] E very instance $\cP\in\C S P(\gA)$ ca n beredu ced in po lyn omialtim e to po lynomi allyma nyi n stanc e s o ver a lgebras ea c h of whic heith er is M al’tsev or has | && \mal_1(x,y,y)=f(\mal(x,y,y))=f(x)=x\\
&&_\mal_1(y,y,x)=f(\mal(y,y,x))=f(x)=x,\end{aligned}$$ for_any $x,y\in\zA_1$ with $x\eqc{\sg_1}y$.
The_results of_[@Maroti10:tree]_imply the_following._Let $\gA$ be_a class of_similar finite algebras closed_under subalgebras, and_retracts_via idempotent unary polynomials. Suppose that $\gA$ has a term operation $f$ satisfying the_following_conditions for_some_$\zB\in\gA$:
-_ $f(x,f(x,y))=f(x,y)$ for any_$x,y\in\zB$;
- for each_$a\in\zB$ the_mapping $x\mapsto f(a,x)$ is not surjective;
- _the_set $C$ of_$a\in\zB$ such that $x\mapsto f(x,a)$ is surjective generates a_proper subalgebra of $\zB$.
Then $\CSP(\gA)$ is_polynomial time reducible_to_$\CSP(\gA-\{\zB\})$.
By_Lemma \[lem:dot-inequality\] the operation $\cdot$_of the class of SBM algebras_from $\gA$ satisfies condition (1). If_the operation $a\cdot x$ is surjective for_some $a$, then $a\le x$ for_all $x\in\zB$. Therefore the only_case when_condition (2) is not satisfied_is when $\zB$_has a_minimal element. Finally,_condition (3) is satisfied whenever $\zB$_is not a_Mal’tsev algebra. Therefore, choosing $\zB$ to_be_a maximal (in_terms_of_cardinality) algebra_from $\gA$ satisfying_conditions_(1)–(3) we_may_only consider instances of $\CSP(\gA)$, in_which_every domain has a minimal element or_is a Mal’tsev algebra.
\[cor:maroti\]_Every_instance $\cP\in\CSP(\gA)$ can be_reduced in polynomial time to_polynomially many instances over algebras each_of which_either is_Mal’tsev or has |
LASSO modification of the LARS algorithm iterations (y-axis: the first iteration corresponds to $\lambda_{max}$ and the last one corresponds to $\lambda_{min}$). On the right, we consider the CENeP selected variables (x-axis) with respect to the Elastic-Net modification of the LARS algorithm iterations (y-axis: the first iteration corresponds to $\lambda_{max}$ and the last one corresponds to $\lambda_{min}$). The selected iteration is marked by red diamonds for the CoLP, green squares for CoRLaP and black squares for the CENeP.](plotIndiceCoLP "fig:"){height="2.3in"}![Variable selection analysis for the CoLP, the CoRLaP and the CENeP in Example (b)\[300/1\] (variables $1$ to $5$ and $10$ to $25$ are relevant; see variables in dark blue on the plot). On the left, we consider the CoLP and the CoRLaP selected variables (x-axis) with respect to the LASSO modification of the LARS algorithm iterations (y-axis: the first iteration corresponds to $\lambda_{max}$ and the last one corresponds to $\lambda_{min}$). On the right, we consider the CENeP selected variables (x-axis) with respect to the Elastic-Net modification of the LARS algorithm iterations (y-axis: the first iteration corresponds to $\lambda_{max}$ and the last one corresponds to $\lambda_{min}$). The selected iteration is marked by red diamonds for the CoLP, green squares for CoRLaP and black squares for the CENeP.](plotIndiceEN "fig:"){height="2.3in"}
\[fig:plotsIndice\]
Selection.
: The selection ability of Sparse Conformal Predictors is here in concern. First, note that the selected variables in SCPs are directly linked to the selection ordering through the iterations of the LASSO or Elastic-Net modification of the LARS algorithm. Then, if the used modification of the LARS algorithm fails to recover the true model, we can not hope to get a predictor which contains only the true variables. Figure \[fig:plotsIndice\] illustrates the evolution of the variable selection of CoLP, CoRLaP and the LASSO on one hand and the CENeP and the Elastic-Net on the | LASSO modification of the LARS algorithm iterations (y - axis: the inaugural iteration correspond to $ \lambda_{max}$ and the last one corresponds to $ \lambda_{min}$). On the right, we consider the CENeP choose variables (x - axis) with respect to the Elastic - final change of the LARS algorithm iterations (yttrium - bloc: the first iteration corresponds to $ \lambda_{max}$ and the final one corresponds to $ \lambda_{min}$). The selected iteration is marked by crimson diamonds for the CoLP, green squares for CoRLaP and black squares for the CENeP.](plotIndiceCoLP " fig:"){height="2.3in"}![Variable choice analysis for the CoLP, the CoRLaP and the CENeP in Example (b)\[300/1\ ] (variables $ 1 $ to $ 5 $ and $ 10 $ to $ 25 $ are relevant; see variables in colored blue on the plot). On the left, we consider the CoLP and the CoRLaP selected variable (x - axis) with respect to the LASSO modification of the LARS algorithm iterations (y - bloc: the first iteration corresponds to $ \lambda_{max}$ and the last one corresponds to $ \lambda_{min}$). On the right, we consider the CENeP selected variable star (x - axis) with respect to the Elastic - Net modification of the LARS algorithm iterations (y - axis: the first iteration corresponds to $ \lambda_{max}$ and the last one corresponds to $ \lambda_{min}$). The choose iteration is score by red rhombus for the CoLP, fleeceable squares for CoRLaP and black squares for the CENeP.](plotIndiceEN " fig:"){height="2.3 in " }
\[fig: plotsIndice\ ]
Selection.
: The selection ability of Sparse Conformal Predictors is here in concern. First, note that the selected variable star in SCPs are directly linked to the selection ordering through the iterations of the LASSO or Elastic - Net change of the LARS algorithm. Then, if the used modification of the LARS algorithm fails to recover the genuine model, we cannot hope to catch a predictor which contains only the true variables. trope \[fig: plotsIndice\ ] illustrates the evolution of the variable selection of CoLP, CoRLaP and the LASSO on one hand and the CENeP and the Elastic - Net on the | LADSO modification of the UARS algorithm nreratimns (y-asis: the wirst iteration corresponds vo $\lqmbda_{nax}$ and the last one currespondd to $\lamvda_{mmn}$). On the right, xs consider ths CENzP selected varisbles (x-axiv) with respect tu che Elastic-Net modification of the LWRS algprlthm iterationf (y-aqif: ths first iteration corresponds to $\lzmbda_{maq}$ and the last onr corresponds to $\lambda_{min}$). The selected iteratioj is marked by wwd diamonds wor the CoLP, green squzres for CoRLaP and black squards fox the CENeP.](klktLtdiceCoLP "fmg:"){heigrt="2.3in"}![Variable selectimn analusis for the CpLP, thw CoRLaP and the CENe' in Example (b)\[300/1\] (variabjes $1$ to $5$ dnb $10$ to $25$ are relevant; swe varidblev in eary bmux oh the olov). On the lert, we consieer the CoLP and tht CjGKaP selected variaflqs (x-axis) with respect to the LASSO modixicztion of the LARS algoruthm iterations (y-axis: the firse iteration corresponds to $\lambda_{max}$ and the last one rofrewpjves to $\lambda_{min}$). On the right, we consider the CEGsP svlected variables (x-axis) with rfslgct to the Elartic-Nec mkdification of the LARS ajgorirhm iterauions (y-axis: the first iteration xorresponds no $\lqmbda_{max}$ and the lcst one corrzspondx to $\kambda_{min}$). The selected nteratjon is markfd by red aiamonds for the CoKP, green squares for CoRLaP wnd black squcres for the CENeP.]([lotIndiceFN "fin:"){veight="2.3in"}
\[fig:plohsIndnce\]
Senection.
: Hhe selection ability of Sparse Conformal Preciwtogs is herz in cpncern. First, gote that the xelecteb varixbles in SBPs are dmrectly linkqd to the selaftion orderiig througr thw itwrationr of the LASSO pr Elastib-Ntt modificqtion of the LARS elgotifhm. Then, if the ustd modification pf ghe LWRX wngorithm fains tu rdvover the true miael, ee can not hope to gat a predictor which cpnbains onli the truq variables. Fogure \[fig:plotsIndict\] illuvtretes tne gvolution of the variable selecfion of ClLP, CoRLaP and ehe OASSO on one hand and the CENeP and the Elastic-Net in the | LASSO modification of the LARS algorithm iterations first corresponds to and the last the we consider the selected variables (x-axis) respect to the Elastic-Net modification of LARS algorithm iterations (y-axis: the first iteration corresponds to $\lambda_{max}$ and the last corresponds to $\lambda_{min}$). The selected iteration is marked by red diamonds for the green for and squares for the CENeP.](plotIndiceCoLP "fig:"){height="2.3in"}![Variable selection analysis for the CoLP, the CoRLaP and the CENeP in (b)\[300/1\] (variables $1$ to $5$ and $10$ to are relevant; see variables dark blue on the plot). the we consider CoLP the selected variables (x-axis) respect to the LASSO modification of the LARS algorithm iterations (y-axis: the first iteration corresponds to $\lambda_{max}$ the last to $\lambda_{min}$). the we the CENeP selected with respect to the Elastic-Net modification algorithm iterations (y-axis: the first iteration corresponds to and the one corresponds to $\lambda_{min}$). The selected is marked by red diamonds for the CoLP, squares for CoRLaP and black squares for the CENeP.](plotIndiceEN "fig:"){height="2.3in"} \[fig:plotsIndice\] Selection. : The selection Sparse Conformal Predictors is in concern. First, that selected in are directly to the selection ordering through the iterations of the LASSO or modification of the LARS algorithm. Then, if the used modification LARS fails to recover true model, we can hope get a predictor which the variables. the of variable selection of CoLP, and the LASSO on one and the CENeP and | LASSO modification of the LARs algorithm IteraTioNs (y-AxIs: thE firSt iteration corREspoNds to $\lambda_{max}$ and the laSt one CoRRespONdS to $\laMbda_{min}$). oN tHE RigHt, We ConSiDEr The CEneP SelecteD variables (X-axIs) With respect tO ThE Elastic-NeT moDification of The lARS alGoRitHM iterAtiOns (y-aXis: the FIrst itEration coRrESponds TO $\lambda_{MAX}$ aNd thE last one corresponDS tO $\Lambda_{min}$). The seLected ItERaTIOn iS maRked by red dIaMonds FOr the Colp, gREEN sqUAres for CoRLaP And black squARes For the cEneP.](PLotIndIceColP "FIg:"){hEight="2.3in"}![VariAble Selection AnalysIS for the cOLP, the CORLaP aNd tHe CeNeP IN EXaMplE (b)\[300/1\] (VAriABlEs $1$ tO $5$ And $10$ To $25$ are relEvAnT; see vAriaBLES In daRk bLue oN the pLot). On the left, wE coNsidER thE CoLP And thE CoRlap seleCted vaRiablEs (X-axis) with respecT to tHe LASSO moDifIcAtiOn Of the laRS algOriThm IteratiOns (y-axiS: The FiRST ItEration corresponds To $\LAMbDa_{max}$ and The lasT OnE cORrespondS tO $\laMbda_{MIN}$). On thE rigHT, wE consideR the CEnEP SeLected vArIables (X-aXis) WitH respECt to The ElaStic-Net mOdifiCAtion of the LARS ALgorithm iteraTIoNS (Y-aXIs: thE fiRst iteratioN corREspoNds tO $\LaMbdA_{Max}$ anD the lAsT OnE Corresponds to $\lambda_{MiN}$). The seLecteD iteration is mArked by red DIAMonds for The COlP, GReen squares for coRLap and black sQUares for The CEneP.](plotINdiceEN "fiG:"){HEight="2.3in"}
\[fIg:pLotSInDicE\]
sElEction.
: The seleCTIon aBiLity of SParSe ConfoRmaL PrEdiCtoRs Is here in cOncern. FiRsT, nOtE tHat The seLEcted varIaBleS iN SCps are DIrectlY linkEd to ThE sELecTion ordERiNG ThroUgH tHe itEraTiOns of The LasSO Or ElastIc-Net modiFicATion Of ThE LARS alGorithm. Then, if ThE used modifIcAtiOn of thE laRS algorIthm fails to recover the trUE model, wE caN not hOpe tO get a predIctOr whicH coNTains oNly the True vArIabLES. FiguRE \[FiG:plOtSIndice\] illUSTraTes thE eVoluTion of tHe variable selectioN Of COLP, CoRLaP and tHe LaSSO ON OnE haND aND thE CeneP AND the Elastic-Net oN the | LASSO modification of the LARS algo rithm it era ti ons(y-a xis: the first iter ation corresponds to $ \lamb da _ {max } $and t he last on e cor re sp ond st o$\lam bda _{min}$ ). On therig ht , we conside r t he CENeP s ele cted variabl es(x-axi s) wi t h res pec t tothe El a stic-N et modifi ca t ion of the LAR S al gori thm iterations (y - ax i s: the first i terati on co r r esp ond s to $\lam bd a_{ma x }$ andt he l a sto ne correspond s to $\lamb d a_{ min}$) .The select ed it er a tio n is marked byred diamo nds fo r the Co L P, gree n squa res fo r Co R La Pand b l ack sq uar e s f or the C EN eP .](pl otIn d i c e CoLP "f ig:" ){hei ght="2.3in"}! [Va riab l e s elect ion a naly si s for the C oLP,th e CoRLaP and th e CE NeP in Ex amp le (b )\ [300/ 1 \] (va ria ble s $1$ t o $5$ a n d $ 10 $ t o$25$ are relevant; s e e v ariables in da r kbl u e on the p lot ). O n the l eft, we conside r theC oL Pand the C oRLaPse lec ted vari a bles (x-ax is) with resp e ct to the LASS O modification of t he LARS al gorithm ite rati o ns ( y-ax i s: th e firs t ite ra t io n corresponds to $\l am bda_{m ax}$and the lastone corres p o n ds to $\ lamb d a_ { min}$). On the righ t, we cons i der theCENeP selecte d variabl e s (x-axis ) w ith re spe c t t o the Elastic - N et m od ificati onof theLAR S a lgo rit hm iteratio ns (y-ax is :th efir st it e ration c or res po nds to $ \ lambda _{max }$ a nd t h e l ast one co r r espo nd sto $ \la mb da_{m in}$ ) . T he sele cted iter ati o n is m ar ked byred diamondsfo r the CoLP ,gre en squ a r es for C oRLaP and black squares for the CE NeP.] (plo tIndiceEN "f ig:"){ hei g ht="2. 3in"}
\[ fig : p lotsI n d ic e\]
Selection. : These lect ion abi lity of Sparse Con f orm al Predictors is her e in co n ce r n.Fi r st, n ote that the se lected var ia b le s in SCPsa redi rectlylinkedto th e select ion order ing throu gh the i ter ations ofthe LASS O or Elas t ic-Ne t m odifi cat ion of t heLARSalgori t hm. Then , if t he usedmodif ic ation of the LARS algorithm fai ls torecov erthe truemod e l,we can no t ho pe to geta p red ictor wh i ch co ntai n sonl y thetrue variables . F igu r e \ [fig:plotsI n d i ce\ ] ill ust r ates t he e volution of the v a riable selecti on o f CoL P,C oRLa Pand the LASSOonon e hand and t he CENeP an d the El as t ic-Ne t on t he | LASSO_modification of_the LARS algorithm iterations_(y-axis: the_first_iteration corresponds_to_$\lambda_{max}$ and the_last one corresponds_to $\lambda_{min}$). On the_right, we consider_the_CENeP selected variables (x-axis) with respect to the Elastic-Net modification of the LARS algorithm_iterations_(y-axis: the_first_iteration_corresponds to $\lambda_{max}$ and the_last one corresponds to $\lambda_{min}$)._The selected_iteration is marked by red diamonds for the_CoLP,_green squares for_CoRLaP and black squares for the CENeP.](plotIndiceCoLP "fig:"){height="2.3in"}![Variable selection_analysis for the CoLP, the CoRLaP_and the CENeP_in_Example (b)\[300/1\]_(variables $1$ to $5$_and $10$ to $25$ are relevant;_see variables in dark blue on_the plot). On the left, we consider_the CoLP and the CoRLaP selected_variables (x-axis) with respect to_the LASSO_modification of the LARS algorithm_iterations (y-axis: the_first iteration_corresponds to $\lambda_{max}$_and the last one corresponds to_$\lambda_{min}$). On the_right, we consider the CENeP selected_variables_(x-axis) with respect_to_the_Elastic-Net modification_of the LARS_algorithm_iterations (y-axis:_the_first iteration corresponds to $\lambda_{max}$ and_the_last one corresponds to $\lambda_{min}$). The selected_iteration is marked by_red_diamonds for the CoLP,_green squares for CoRLaP and_black squares for the CENeP.](plotIndiceEN "fig:"){height="2.3in"}
_ _ \[fig:plotsIndice\]
Selection.
:_ The selection ability of Sparse Conformal Predictors is here_in concern. First, note that the_selected variables in SCPs_are directly_linked_to the selection_ordering_through the_iterations of the LASSO or Elastic-Net modification_of the_LARS algorithm. Then, if the used_modification of the LARS_algorithm_fails to recover the true model,_we can not hope to get_a predictor which contains only_the_true_variables. Figure \[fig:plotsIndice\] illustrates the evolution_of the variable selection of CoLP,_CoRLaP and the_LASSO on one hand and the CENeP_and_the Elastic-Net on the |
+ a\times \nabla_{A_s}\xi + a\times a\times \xi,$$ and thus, for $\xi \in W_{A_1}^{2,p}(X;\Lambda^l\otimes\ad P)$, $$\label{eq:Delta_A1+a_expansion}
\Delta_A\xi = \Delta_{A_s}\xi + \nabla_{A_1}a \times \xi + a_1\times a\times \xi + a\times \nabla_{A_1}\xi + a\times a\times \xi,$$ We define $r \in [p,\infty]$ by $1/p = 1/q + 1/r$ and recall that by [@AdamsFournier Theorem 4.12] we have
$W^{2,p}(X) \subset L^r(X)$ for any $r \in [1,\infty)$ when $p = d/2$, and
$W^{2,p}(X) \subset L^\infty(X)$ when $p > d/2$.
The expansion yields $$\begin{gathered}
\|(\Delta_A - \Delta_{A_s})\xi\|_{L^p(X)}
\leq
z\|\nabla_{A_1}a\|_{L^q(X)}\|\xi\|_{L^r(X)} + \|a \times \nabla_{A_1}\xi\|_{L^p(X)}
\\
+ z\|a_1\|_{C(X)}\|a\|_{L^{2p}(X)}\|\xi\|_{L^{2p}(X)} + z\||a|^2\|_{L^q(X)} \|\xi\|_{L^r(X)},\end{gathered}$$ where $z = z(g,G,l) \in [1,\infty)$. To ensure a continuous Sobolev embedding $W^{1,p}(X) \subset L^d(X)$ by [@AdamsFournier Theorem 4.12], we need $p^* = dp/(d-p) \geq d$, that is, $p \geq d-p$ or $p \geq d/2$, which we assume in our hypotheses.
To ensure a continuous Sobolev embedding $W^{1,q}(X) \subset L^{2q}(X)$ when $q < d$, we need $q | + a\times \nabla_{A_s}\xi + a\times a\times \xi,$$ and thus, for $ \xi \in W_{A_1}^{2,p}(X;\Lambda^l\otimes\ad P)$, $ $ \label{eq: Delta_A1+a_expansion }
\Delta_A\xi = \Delta_{A_s}\xi + \nabla_{A_1}a \times \xi + a_1\times a\times \xi + a\times \nabla_{A_1}\xi + a\times a\times \xi,$$ We define $ r \in [ p,\infty]$ by $ 1 / p = 1 / q + 1 / r$ and echo that by [ @AdamsFournier Theorem 4.12 ] we take
$ W^{2,p}(X) \subset L^r(X)$ for any $ r \in [ 1,\infty)$ when $ p = d/2 $, and
$ W^{2,p}(X) \subset L^\infty(X)$ when $ p > d/2$.
The expansion give way $ $ \begin{gathered }
\|(\Delta_A - \Delta_{A_s})\xi\|_{L^p(X) }
\leq
z\|\nabla_{A_1}a\|_{L^q(X)}\|\xi\|_{L^r(X) } + \|a \times \nabla_{A_1}\xi\|_{L^p(X) }
\\
+ z\|a_1\|_{C(X)}\|a\|_{L^{2p}(X)}\|\xi\|_{L^{2p}(X) } + z\||a|^2\|_{L^q(X) } \|\xi\|_{L^r(X)},\end{gathered}$$ where $ z = z(g, G, l) \in [ 1,\infty)$. To see a continuous Sobolev embed $ W^{1,p}(X) \subset L^d(X)$ by [ @AdamsFournier Theorem 4.12 ], we need $ p^ * = dp/(d - phosphorus) \geq d$, that is, $ phosphorus \geq d - p$ or $ p \geq d/2 $, which we assume in our guess.
To ensure a continuous Sobolev embedding $ W^{1,q}(X) \subset L^{2q}(X)$ when $ q < d$, we necessitate $ q | + a\himes \nabla_{A_s}\xi + a\times x\times \xi,$$ and tkys, for $\xi \in W_{A_1}^{2,p}(X;\Lamcda^l\otimes\ad P)$, $$\label{eq:Delta_A1+e_expqnsiob}
\Delta_A\xi = \Delta_{A_s}\xi + \vabla_{A_1}a \tpmes \xi + q_1\timts a\times \xi + a\times \nabla_{A_1}\xi + z\bimes e\times \xi,$$ We denine $r \in [p,\hnfty]$ by $1/p = 1/q + 1/r$ aud recall that by [@AdamsFournier Theowem 4.12] we hwve
$W^{2,p}(X) \subset J^r(X)$ gjr ahj $v \in [1,\infty)$ when $p = d/2$, and
$W^{2,p}(X) \subsst L^\infuy(X)$ when $p > d/2$.
The ecpansion yields $$\begin{gatheged}
\|(\Dflta_A - \Delta_{A_s})\xi\|_{L^p(D)}
\leq
z\|\nabla_{A_1}q\|_{L^q(X)}\|\vu\|_{L^r(X)} + \|a \timer \nabla_{A_1}\xi\|_{L^p(X)}
\\
+ z\|a_1\|_{C(X)}\|a\|_{L^{2p}(X)}\|\si\|_{L^{2p}(X)} + z\||a|^2\|_{L^q(X)} \|\xi\|_{L^r(X)},\end{gathered}$$ dhere $z = z(g,G,l) \ib [1,\unfhi)$. To ensure e contpnuous Soboley embedging $W^{1,p}(C) \subset L^d(X)$ bn [@AdaksFiurnier Theorem 4.12], we nxed $p^* = dp/(d-p) \geq d$, thwt is, $p \gaq d-p$ or $p \geq d/2$, whucy we dssuke iv ouf hgpptgeses.
Tl eisure a confinuous Sobilev embedding $W^{1,q}(X) \xufwet L^{2q}(X)$ when $s < d$, wq geed $q | + a\times \nabla_{A_s}\xi + a\times a\times \xi,$$ for \in W_{A_1}^{2,p}(X;\Lambda^l\otimes\ad $$\label{eq:Delta_A1+a_expansion} \Delta_A\xi = + a\times \xi + \nabla_{A_1}\xi + a\times \xi,$$ We define $r \in [p,\infty]$ $1/p = 1/q + 1/r$ and recall that by [@AdamsFournier Theorem 4.12] we $W^{2,p}(X) \subset L^r(X)$ for any $r \in [1,\infty)$ when $p = d/2$, and \subset when > The expansion yields $$\begin{gathered} \|(\Delta_A - \Delta_{A_s})\xi\|_{L^p(X)} \leq z\|\nabla_{A_1}a\|_{L^q(X)}\|\xi\|_{L^r(X)} + \|a \times \nabla_{A_1}\xi\|_{L^p(X)} \\ + z\|a_1\|_{C(X)}\|a\|_{L^{2p}(X)}\|\xi\|_{L^{2p}(X)} z\||a|^2\|_{L^q(X)} \|\xi\|_{L^r(X)},\end{gathered}$$ where $z = z(g,G,l) \in [1,\infty)$. ensure a continuous Sobolev $W^{1,p}(X) \subset L^d(X)$ by [@AdamsFournier 4.12], need $p^* dp/(d-p) d$, is, $p \geq or $p \geq d/2$, which we assume in our hypotheses. To ensure a continuous Sobolev embedding $W^{1,q}(X) L^{2q}(X)$ when d$, we $q | + a\times \nabla_{A_s}\xi + a\times a\timEs \xi,$$ and thuS, for $\xI \in w_{A_1}^{2,p}(x;\LAmbdA^l\otImes\ad P)$, $$\label{eq:dElta_a1+a_expansion}
\Delta_A\xi = \DelTa_{A_s}\xI + \nABla_{A_1}A \TiMes \xi + A_1\times a\TImES \Xi + a\TiMeS \naBlA_{a_1}\xI + a\timEs a\Times \xi,$$ we define $r \iN [p,\iNfTy]$ by $1/p = 1/q + 1/r$ and reCAlL that by [@AdaMsFOurnier TheorEm 4.12] wE have
$W^{2,P}(X) \SubSEt L^r(X)$ For Any $r \iN [1,\infty)$ WHen $p = d/2$, aNd
$W^{2,p}(X) \subsEt l^\Infty(X)$ WHen $p > d/2$.
ThE EXpAnsiOn yields $$\begin{gathEReD}
\|(\delta_A - \Delta_{A_s})\xI\|_{L^p(X)}
\leQ
z\|\NAbLA_{a_1}a\|_{L^Q(X)}\|\xI\|_{L^r(X)} + \|a \times \NaBla_{A_1}\xI\|_{l^p(X)}
\\
+ z\|a_1\|_{C(X)}\|A\|_{l^{2p}(x)}\|\XI\|_{l^{2p}(X)} + Z\||A|^2\|_{L^q(X)} \|\xi\|_{L^r(X)},\end{gAthered}$$ wherE $Z = z(g,g,l) \in [1,\inFtY)$. To ENsure a ContiNuOUs SObolev embedDing $w^{1,p}(X) \subset l^d(X)$ by [@ADAmsFourNIer TheoRem 4.12], we nEed $P^* = dp/(D-p) \geQ D$, tHaT is, $P \gEQ d-p$ OR $p \Geq D/2$, WhiCh we assuMe In Our hyPothESES.
to enSurE a coNtinuOus Sobolev embEddIng $W^{1,Q}(x) \suBset L^{2Q}(X)$ wheN $q < d$, wE nEed $q | + a\times \nabla_{A_s}\xi + a\times a\ti mes \x i, $$ a nd t hus, for $\xi\ in W _{A_1}^{2,p}(X;\Lambda ^l\ot im e s\ad P) $, $$ \label{ e q: D e lta _A 1+ a_e xp a ns ion}\De lta_A\x i = \Delta _{A _s }\xi + \nabl a _{ A_1}a \tim es\xi + a_1\ti mes a\tim es \x i + a\ tim es \n abla_{ A _1}\xi + a\time sa \times \xi,$$W e d efin e $r \in [p,\inft y ]$ by $1/p = 1/q+ 1/r$ a n dr e cal l t hat by [@A da msFou r nier Th e or e m 4.1 2 ] we have
$W ^{2,p}(X) \ s ubs et L^r (X )$f or any $r \ in [1, \infty)$ wh en $ p = d/2$, and
$ W^{2,p} ( X) \sub set L^ \in fty (X)$ wh en $p > d/2 $ .
Th e ex pansionyi el ds $$ \beg i n { g athe red }
\| (\Del ta_A - \Delta _{A _s}) \ xi\ |_{L^ p(X)}
\le qz\|\n abla_{ A_1}a \| _{L^q(X)}\|\xi\ |_{L ^r(X)} +\|a \ tim es \nab l a_{A_1 }\x i\| _{L^p(X )}
\\
+ z\| a_ 1 \ | _{ C(X)}\|a\|_{L^{2p} (X ) } \| \xi\|_{L ^{2p}( X )} + z\||a|^ 2\ |_{ L^q( X ) } \|\ xi\| _ {L ^r(X)},\ end{ga t he re d}$$ wh er e $z = z (g, G,l ) \in [1,\ infty) $. To en surea continuous So b olev embeddin g $ W ^ {1 , p}(X ) \ subset L^d( X)$b y [@ Adam s Fo urn i er Th eorem 4 . 12 ] , we need $p^* = dp /( d-p) \ geq d $, that is, $ p \geq d-p $ o r $p \ge q d/ 2 $, which we assum e inour hypoth e ses.
To ensu re a con tinuous S o b olev emb edd ing $W ^{1 , q }( X) \subset L^ { 2 q}(X )$ when $ q < d$, we ne ed$q | +_a\times \nabla_{A_s}\xi_+ a\times a\times \xi,$$_and thus,_for_$\xi \in_W_{A_1}^{2,p}(X;\Lambda^l\otimes\ad_P)$, $$\label{eq:Delta_A1+a_expansion}
\Delta_A\xi =_\Delta_{A_s}\xi + \nabla_{A_1}a_\times \xi + a_1\times_a\times \xi +_a\times_\nabla_{A_1}\xi + a\times a\times \xi,$$ We define $r \in [p,\infty]$ by $1/p = 1/q_+_1/r$ and_recall_that_by [@AdamsFournier Theorem 4.12] we_have
$W^{2,p}(X) \subset L^r(X)$ for any_$r \in_[1,\infty)$ when $p = d/2$, and
$W^{2,p}(X) \subset L^\infty(X)$_when_$p > d/2$.
The_expansion yields $$\begin{gathered}
\|(\Delta_A - \Delta_{A_s})\xi\|_{L^p(X)}
\leq
z\|\nabla_{A_1}a\|_{L^q(X)}\|\xi\|_{L^r(X)} + \|a \times \nabla_{A_1}\xi\|_{L^p(X)}
\\
+_z\|a_1\|_{C(X)}\|a\|_{L^{2p}(X)}\|\xi\|_{L^{2p}(X)} + z\||a|^2\|_{L^q(X)} \|\xi\|_{L^r(X)},\end{gathered}$$ where_$z = z(g,G,l)_\in_[1,\infty)$._To ensure a continuous_Sobolev embedding $W^{1,p}(X) \subset L^d(X)$ by_[@AdamsFournier Theorem 4.12], we need $p^*_= dp/(d-p) \geq d$, that is, $p_\geq d-p$ or $p \geq d/2$,_which we assume in our_hypotheses.
To ensure_a continuous Sobolev embedding $W^{1,q}(X)_\subset L^{2q}(X)$ when_$q <_d$, we need_$q |
and Development Program of China (No. 2017YFB1002104), and National Natural Science Foundation of China (No. 61672162 and 61751201).
[^1]: The source codes are available at <https://github.com/HSLCY/VCWE>
[^2]: [ ]{}[ ]{}Corresponding author.
[^3]: the graphical component of Chinese, referring to <https://en.wikipedia.org/wiki/Radical_(Chinese_characters)>
[^4]: the basic pattern of Chinese characters, referring to <https://en.wikipedia.org/wiki/Stroke_(CJKV_character)>
[^5]: https://dumps.wikimedia.org/zhwiki/20180520/
[^6]: https://github.com/attardi/wikiextractor/blob/master/Wiki\
Extractor.py
[^7]: https://github.com/BYVoid/OpenCC
[^8]: https://github.com/thunlp/THULAC-Python
[^9]: https://code.google.com/archive/p/word2vec/
[^10]: https://github.com/Leonard-Xu/CWE
[^11]: https://github.com/ray1007/gwe
[^12]: https://github.com/hkust-knowcomp/jwe
[^13]: https://github.com/FudanNLP/VCWE
[^14]: http://sentic.net/chinese-review-datasets.zip
[^15]: https://github.com/bamtercelboo/pytorch\_NER\_PosTag\_Bi\
LSTM\_CRF
[^16]: http://klcl.pku.edu.cn/zygx/zyxz/index.htm
---
abstract: 'In this paper we consider the growth, large fluctuations and memory properties of an affine stochastic functional differential equation with an average functional where the contributions of the average and instantaneous terms are parameterised. An asymptotic analysis of the solution of this equation is conducted for all values of the parameters of the equation. When solutions are recurrent, we show that the autocovariance function of the solution decays at a polynomial rate, even though the solution is asymptotically equal to another asymptotically stationary process whose autocovariance function decays | and Development Program of China (No. 2017YFB1002104), and National Natural Science Foundation of China (No. 61672162 and 61751201).
[ ^1 ]: The source codes are available at < https://github.com/HSLCY/VCWE >
[ ^2 ]: [ ] { } [ ] { } Corresponding writer.
[ ^3 ]: the graphic component of Chinese, referring to < https://en.wikipedia.org/wiki/Radical_(Chinese_characters) >
[ ^4 ]: the basic pattern of taiwanese characters, referring to < https://en.wikipedia.org/wiki/Stroke_(CJKV_character) >
[ ^5 ]: https://dumps.wikimedia.org/zhwiki/20180520/
[ ^6 ]: https://github.com/attardi/wikiextractor/blob/master/Wiki\
Extractor.py
[ ^7 ]: https://github.com/BYVoid/OpenCC
[ ^8 ]: https://github.com/thunlp/THULAC-Python
[ ^9 ]: https://code.google.com/archive/p/word2vec/
[ ^10 ]: https://github.com/Leonard-Xu/CWE
[ ^11 ]: https://github.com/ray1007/gwe
[ ^12 ]: https://github.com/hkust-knowcomp/jwe
[ ^13 ]: https://github.com/FudanNLP/VCWE
[ ^14 ]: http://sentic.net/chinese-review-datasets.zip
[ ^15 ]: https://github.com/bamtercelboo/pytorch\_NER\_PosTag\_Bi\
LSTM\_CRF
[ ^16 ]: http://klcl.pku.edu.cn/zygx/zyxz/index.htm
---
abstract:' In this paper we view the growth, large fluctuation and memory properties of an affine stochastic functional differential equality with an average functional where the contributions of the average and instantaneous terms are parameterised. An asymptotic psychoanalysis of the solution of this equation is conducted for all value of the parameters of the equation. When solution are recurrent, we show that the autocovariance function of the solution decay at a polynomial rate, even though the solution is asymptotically equal to another asymptotically stationary process whose autocovariance affair decays | anf Development Program of China (No. 2017YFB1002104), aue Natimnal Nztural Szience Foundation of China (Nl. 61672162 and 61751201).
[^1]: The source codes are xvailable at <httpw://gitiub.com/HSLCY/VCWE>
[^2]: [ ]{}[ ]{}Corresijndihn autkoc.
[^3]: the graphical component of Chinese, rexefrnng to <https://en.wikipedia.org/wiki/Radicaj_(Chinesr_cjaracters)>
[^4]: the fasib [attsgn of Chinese characters, referrjng to <ittps://en.wikipedis.org/wiki/Stroke_(CJKV_characteg)>
[^5]: hthps://dumps.wikimedia.ogg/zhwiki/20180520/
[^6]: htjls://gyrhub.com/attarai/wikiextractor/blob/masjer/Wiki\
Extractor.py
[^7]: https://githjb.com/YYVoid/OpenCX
[^8]: yttov://github.com/tiunlp/TRULAC-Python
[^9]: mntps://coda.google.vom/archive/p/worc2ver/
[^10]: htrps://github.com/Leonard-Xu/RWE
[^11]: https://github.com/rai1007/gwe
[^12]: https://cichub.com/hkust-knowcomp/jqe
[^13]: httpv://gitvub.cun/FuaanHL'/VCSE
[^14]: htto://seitic.net/chinsse-review-darasets.zip
[^15]: https://githib.sim/bamtercelbok/pytorsh\_GER\_PosTag\_Bi\
LSTM\_CRF
[^16]: http://klcl.pku.edu.cn/eygx/zgxz/index.htm
---
abstract: 'In this paper we considgr the grorth, large fluctuations and memory properties of at affmnd suogmastkx vunctional differential equation with an averwfe flnctional where tme contributions og hhr average and knstancznsous terms are parwmeterifed. Ab asymptouic amalysis of the solution of rhis equatiou iw conducted for alp values of the paraketers of the equation. Chen sklutions arf recurrehg, we show that tfe sudocovariance function of tre solutiin dzcays at a pplynomyal rate, egen tmmugh the solution ls asimptothcally equwl to another asymptotically stevionary procexs whmse autoeovarisnce function decays | and Development Program of China (No. 2017YFB1002104), Natural Foundation of (No. 61672162 and are at <https://github.com/HSLCY/VCWE> [^2]: ]{}[ ]{}Corresponding author. the graphical component of Chinese, referring <https://en.wikipedia.org/wiki/Radical_(Chinese_characters)> [^4]: the basic pattern of Chinese characters, referring to <https://en.wikipedia.org/wiki/Stroke_(CJKV_character)> [^5]: https://dumps.wikimedia.org/zhwiki/20180520/ https://github.com/attardi/wikiextractor/blob/master/Wiki\ Extractor.py [^7]: https://github.com/BYVoid/OpenCC [^8]: https://github.com/thunlp/THULAC-Python [^9]: https://code.google.com/archive/p/word2vec/ [^10]: https://github.com/Leonard-Xu/CWE [^11]: https://github.com/ray1007/gwe [^12]: [^13]: [^14]: [^15]: LSTM\_CRF [^16]: http://klcl.pku.edu.cn/zygx/zyxz/index.htm --- abstract: 'In this paper we consider the growth, large fluctuations and memory of an affine stochastic functional differential equation with average functional where the of the average and instantaneous are An asymptotic of solution this equation is for all values of the parameters of the equation. When solutions are recurrent, we show that the function of decays at polynomial even the solution is to another asymptotically stationary process whose | and Development Program of ChIna (No. 2017YFB1002104), anD NatiOnaL NaTuRal SCienCe Foundation of cHina (no. 61672162 and 61751201).
[^1]: The source codes are AvailAbLE at <hTTpS://githUb.com/HSlcY/vcwE>
[^2]: [ ]{}[ ]{}COrReSpoNdINg AuthoR.
[^3]: thE graphiCal componeNt oF CHinese, referrINg To <https://en.wIkiPedia.org/wiki/radIcal_(ChInEse_CHaracTerS)>
[^4]: the bAsic paTTern of chinese chArACters, rEFerring TO <HtTps://eN.wikipedia.org/wiki/sTrOKe_(CJKV_characteR)>
[^5]: https://DuMPs.WIKimEdiA.org/zhwiki/20180520/
[^6]: HtTps://giTHub.com/aTTaRDI/WikIExtractor/blob/Master/Wiki\
EXTraCtor.py
[^7]: HtTps://GIthub.cOm/BYVOiD/opeNCC
[^8]: https://gitHub.cOm/thunlp/ThULAC-PYThon
[^9]: httPS://code.goOgle.coM/arChiVe/p/wORd2VeC/
[^10]: htTpS://GitHUb.Com/lEonArd-Xu/CWE
[^11]: HtTpS://githUb.coM/RAY1007/Gwe
[^12]: hTtpS://gitHub.coM/hkust-knowcomP/jwE
[^13]: httPS://giThub.cOm/FudAnNLp/VcWE
[^14]: htTp://sentIc.net/ChInese-review-dataSets.Zip
[^15]: https://gIthUb.Com/BaMtercELboo/pyTorCh\_NeR\_PosTaG\_Bi\
LSTM\_crF
[^16]: hTtP://KLCl.Pku.edu.cn/zygx/zyxz/inDeX.HTm
---
Abstract: 'in this PApEr WE consideR tHe gRowtH, LArge fLuctUAtIons and mEmory pROpErTies of aN aFfine sToChaStiC funcTIonaL diffeRential eQuatiON with an average FUnctional wherE ThE COnTRibuTioNs of the averAge aND insTantANeOus TErms aRe parAmETeRIsed. An asymptotic anaLySis of tHe solUtion of this eqUation is coNDUCted for aLl vaLUeS Of the parameterS of thE equation. WHEn solutiOns arE recurreNt, we show tHAT the autoCovAriAncE fuNCTiOn of the solutiON DecaYs At a polyNomIal rate, EveN thOugH thE sOlution is AsymptotIcAlLy EqUal To anoTHer asympToTicAlLy sTatioNAry proCess wHose AuToCOvaRiance fUNcTIOn deCaYs | and Development Program o f China (N o. 20 17Y FB1 00 2104 ), a nd National Na t ural Science Foundation of Chin a( No.6 16 72162 and 61 7 51 2 0 1).
[^ 1]: T h esourc e c odes ar e availabl e a t<https://git h ub .com/HSLCY /VC WE>
[^2]: [ ]{ }[ ]{} Co rre s pondi ngautho r.
[^ 3 ]: the graphica lc ompone n t of Ch i n es e, r eferring to <http s :/ / en.wikipedia.o rg/wik i/ R ad i c al_ (Ch inese_char ac ters) >
[^4]: th e b asi c pattern of C hinese char a cte rs, re fe rri n g to < https :/ / en. wikipedia.o rg/w iki/Strok e_(CJK V _charac t er)>
[ ^5]: h ttp s:/ /dum p s. wi kim ed i a.o r g/ zhw i ki/ 20180520 /
[ ^6]:http s : / / gith ub. com/ attar di/wikiextrac tor /blo b /ma ster/ Wiki\
E xtrac tor.py
[^7 ]: https://github .com /BYVoid/O pen CC
[ ^8 ]: ht t ps://g ith ub. com/thu nlp/THU L AC- Py t h o n
[^9]: https://cod e. g o og le.com/a rchive / p/ wo r d2vec/
[^ 10] : ht t p s://g ithu b .c om/Leona rd-Xu/ C WE
[^11]:ht tps:// gi thu b.c om/ra y 1007 /gwe
[^12]: h ttps: / /github.com/hk u st-knowcomp/j w e[ ^1 3 ]: h ttp s://github. com/ F udan NLP/ V CW E
[ ^14]: http :/ / se n tic.net/chinese-rev ie w-data sets. zip
[^15]: h ttps://git h u b .com/bam terc e lb o o/pytorch\_NER \_Pos Tag\_Bi\
LSTM\_ CRF
[^16]: h ttp://klc l . pku.edu. cn/ zyg x/z yxz / i nd ex.htm
---
a b s trac t: 'In th ispaper w e c ons ide r t he growth,large fl uc tu at io nsand m e mory pro pe rti es of an a f fine s tocha stic f un c tio nal dif f er e n tial e qu atio n w it h anaver a gefunctio nal where th e con tr ib utionsof the averag eand instan ta neo us ter m s are par ameterised. An asymptot i c analy sis of t he s olution o f t his eq uat i on isconduc ted f or al l value s of th eparameters o f t he eq ua tion . Whensolutions are recu r ren t, we show th atthea u to cov a ri a nce f u nct i o n of the soluti on decaysat apolynomial rat e, even t hough t he so l ution i s asympto tically e qu al t o ano ther asymp toticall y station a ry pr o ce ss wh ose autoc ov ari ancefuncti o n d ecays | and_Development Program_of China (No. 2017YFB1002104),_and National_Natural_Science Foundation_of_China (No. 61672162_and 61751201).
[^1]: The_source codes are available_at <https://github.com/HSLCY/VCWE>
[^2]: [_]{}[_]{}Corresponding author.
[^3]: the graphical component of Chinese, referring to <https://en.wikipedia.org/wiki/Radical_(Chinese_characters)>
[^4]: the basic pattern of_Chinese_characters, referring_to_<https://en.wikipedia.org/wiki/Stroke_(CJKV_character)>
[^5]:_https://dumps.wikimedia.org/zhwiki/20180520/
[^6]: https://github.com/attardi/wikiextractor/blob/master/Wiki\
_Extractor.py
[^7]: https://github.com/BYVoid/OpenCC
[^8]: https://github.com/thunlp/THULAC-Python
[^9]: https://code.google.com/archive/p/word2vec/
[^10]: https://github.com/Leonard-Xu/CWE
[^11]:_https://github.com/ray1007/gwe
[^12]: https://github.com/hkust-knowcomp/jwe
[^13]:_https://github.com/FudanNLP/VCWE
[^14]: http://sentic.net/chinese-review-datasets.zip
[^15]: https://github.com/bamtercelboo/pytorch\_NER\_PosTag\_Bi\
LSTM\_CRF
[^16]: http://klcl.pku.edu.cn/zygx/zyxz/index.htm
_---
abstract:_'In this paper_we consider the growth, large fluctuations and memory properties_of an affine stochastic functional differential_equation with an_average_functional_where the contributions of_the average and instantaneous terms are_parameterised. An asymptotic analysis of the_solution of this equation is conducted for_all values of the parameters of_the equation. When solutions are_recurrent, we_show that the autocovariance function_of the solution_decays at_a polynomial rate,_even though the solution is asymptotically_equal to another_asymptotically stationary process whose autocovariance function_decays |
,gl} e^{\jmath \frac{1}{g}(k (\omega_1+2 \pi m_1) + l (\omega_2 + 2 \pi m_2))}. \end{aligned}$$ By this, the singular value distribution of ${{{\uppercase{{\bm{t}}}}}}^g$ can be alternatively studied on the generating function $F({\hbox{\boldmath$\omega$}},{{{\lowercase{{\bm{m}}}}}})$.
Higher Dimensional Convolution
------------------------------
According to [@Oudin2009], a block multi-level Toeplitz matrix ${{{\uppercase{{\bm{t}}}}}}=\{{{{\uppercase{{\bm{t}}}}}}_{{{{\lowercase{{\bm{i}}}}}}-{{{\lowercase{{\bm{j}}}}}}}\}_{{{{\lowercase{{\bm{i}}}}}},{{{\lowercase{{\bm{j}}}}}}={{{\uppercase{{\bm{\mathrm{1}}}}}}}}^{{{\lowercase{{\bm{n}}}}}}$ with ${{{\lowercase{{\bm{i}}}}}}=(i_1,\dots,i_d)$, ${{{\lowercase{{\bm{j}}}}}}=(j_1,\dots,j_d)$, and ${{{\lowercase{{\bm{n}}}}}}=(n_1,\dots,n_d)$, it can be alternatively represented as $$\begin{aligned}
{{{\uppercase{{\bm{t}}}}}}= \sum_{\abs{k_1}<n_1} \dots \sum_{\abs{k_d}<n_d} [{{{\uppercase{{\bm{j}}}}}}_{n_1}^{(k_1)} \otimes \dots \otimes {{{\uppercase{{\bm{j}}}}}}_{n_d}^{(k_d)}] \otimes {{{\uppercase{{\bm{t}}}}}}_{{{{\lowercase{{\bm{k}}}}}}}\end{aligned}$$ where ${{{\uppercase{{\bm{j}}}}}}_{n_j}^{(k_j)}$ is a $n_j \times n_j$ binary matrix with $(p,q)$-th entry being 1 of $p-q=k_j$ and 0 elsewhere, and $$\begin{aligned}
{{{\uppercase{{\bm{t}}}}}}_{{{{\lowercase{{\bm{k}}}}}}}=\frac{1}{(2\pi)^d} \int_{\Omega} F({\hbox{\boldmath$\omega$}}) e^{-\jmath <{{{\lowercase{{\bm{k}}}}}},{\hbox{\boldmath$\omega$}}>} d {\hbox | , gl } e^{\jmath \frac{1}{g}(k (\omega_1 + 2 \pi m_1) + l (\omega_2 + 2 \pi m_2) ) }. \end{aligned}$$ By this, the singular value distribution of $ { { { \uppercase{{\bm{t}}}}}}^g$ can be alternatively studied on the generating function $ F({\hbox{\boldmath$\omega$}},{{{\lowercase{{\bm{m}}}}}})$.
eminent Dimensional Convolution
------------------------------
harmonize to [ @Oudin2009 ], a block multi - level Toeplitz matrix $ { { { \uppercase{{\bm{t}}}}}}=\{{{{\uppercase{{\bm{t}}}}}}_{{{{\lowercase{{\bm{i}}}}}}-{{{\lowercase{{\bm{j}}}}}}}\}_{{{{\lowercase{{\bm{i}}}}}},{{{\lowercase{{\bm{j}}}}}}={{{\uppercase{{\bm{\mathrm{1}}}}}}}}^{{{\lowercase{{\bm{n}}}}}}$ with $ { { { \lowercase{{\bm{i}}}}}}=(i_1,\dots, i_d)$, $ { { { \lowercase{{\bm{j}}}}}}=(j_1,\dots, j_d)$, and $ { { { \lowercase{{\bm{n}}}}}}=(n_1,\dots, n_d)$, it can be alternatively represented as $ $ \begin{aligned }
{ { { \uppercase{{\bm{t}}}}}}= \sum_{\abs{k_1}<n_1 } \dots \sum_{\abs{k_d}<n_d } [ { { { \uppercase{{\bm{j}}}}}}_{n_1}^{(k_1) } \otimes \dots \otimes { { { \uppercase{{\bm{j}}}}}}_{n_d}^{(k_d) } ] \otimes { { { \uppercase{{\bm{t}}}}}}_{{{{\lowercase{{\bm{k}}}}}}}\end{aligned}$$ where $ { { { \uppercase{{\bm{j}}}}}}_{n_j}^{(k_j)}$ is a $ n_j \times n_j$ binary matrix with $ (phosphorus, q)$-th entry being 1 of $ p - q = k_j$ and 0 elsewhere, and $ $ \begin{aligned }
{ { { \uppercase{{\bm{t}}}}}}_{{{{\lowercase{{\bm{k}}}}}}}=\frac{1}{(2\pi)^d } \int_{\Omega } F({\hbox{\boldmath$\omega$ } }) e^{-\jmath < { { { \lowercase{{\bm{k}}}}}},{\hbox{\boldmath$\omega$ } } > } d { \hbox | ,gl} f^{\jmath \frac{1}{g}(k (\omega_1+2 \pi m_1) + l (\omega_2 + 2 \pi m_2))}. \end{eligned}$$ By this, the singular value distribuvion of ${{{\ukiercase{{\bm{t}}}}}}^g$ can be algernativepy studiwd oi the generating functiok $F({\hbkw{\boldketh$\omega$}},{{{\lowercaxe{{\bm{m}}}}}})$.
Highes Dimensional Wovvllution
------------------------------
According to [@Oudin2009], a block mtlti-levrl Toeplitz matryx ${{{\ukpewcass{{\bm{t}}}}}}=\{{{{\uppercase{{\bm{t}}}}}}_{{{{\lowercase{{\bm{i}}}}}}-{{{\lowerczse{{\bm{j}}}}}}}\}_{{{{\lmwercase{{\bm{i}}}}}},{{{\loeercase{{\bm{j}}}}}}={{{\uppercase{{\bm{\mathrl{1}}}}}}}}^{{{\lowfrcase{{\bm{n}}}}}}$ with ${{{\lowfrcase{{\bm{i}}}}}}=(i_1,\dits,i_q)$, ${{{\lowercase{{\bm{g}}}}}}=(j_1,\dots,j_d)$, and ${{{\lowercase{{\gm{n}}}}}}=(n_1,\dots,n_d)$, it can be alternativdly rzpresented qs $$\behhn{aligned}
{{{\uppvrcase{{\bm{t}}}}}}= \sum_{\abs{k_1}<n_1} \dods \sum_{\ans{k_d}<n_d} [{{{\uppercaxe{{\bk{j}}}}}}_{n_1}^{(j_1)} \otimes \dots \otimes {{{\nppercase{{\bm{j}}}}}}_{n_d}^{(k_d)}] \otiies {{{\upperwaae{{\bm{t}}}}}}_{{{{\lowercase{{\bm{k}}}}}}}\wne{aligted}$$ fherd ${{{\upoerdaxe{{\gm{j}}}}}}_{n_j}^{(k_u)}$ ia a $n_j \timss n_j$ binart matrix with $(p,q)$-th tntwj being 1 of $p-s=k_j$ anq 0 elsewhere, and $$\begin{aligned}
{{{\uppercast{{\bm{t}}}}}}_{{{{\lkwercase{{\bm{k}}}}}}}=\frac{1}{(2\pi)^d} \int_{\Onega} F({\hbox{\boldmath$\omeha$}}) e^{-\jmath <{{{\lowercase{{\bm{k}}}}}},{\hbox{\boldmath$\omega$}}>} d {\hbox | ,gl} e^{\jmath \frac{1}{g}(k (\omega_1+2 \pi m_1) + + \pi m_2))}. By this, the can alternatively studied on generating function $F({\hbox{\boldmath$\omega$}},{{{\lowercase{{\bm{m}}}}}})$. Dimensional Convolution ------------------------------ According to [@Oudin2009], block multi-level Toeplitz matrix ${{{\uppercase{{\bm{t}}}}}}=\{{{{\uppercase{{\bm{t}}}}}}_{{{{\lowercase{{\bm{i}}}}}}-{{{\lowercase{{\bm{j}}}}}}}\}_{{{{\lowercase{{\bm{i}}}}}},{{{\lowercase{{\bm{j}}}}}}={{{\uppercase{{\bm{\mathrm{1}}}}}}}}^{{{\lowercase{{\bm{n}}}}}}$ with ${{{\lowercase{{\bm{i}}}}}}=(i_1,\dots,i_d)$, ${{{\lowercase{{\bm{j}}}}}}=(j_1,\dots,j_d)$, and ${{{\lowercase{{\bm{n}}}}}}=(n_1,\dots,n_d)$, it can be represented as $$\begin{aligned} {{{\uppercase{{\bm{t}}}}}}= \sum_{\abs{k_1}<n_1} \dots \sum_{\abs{k_d}<n_d} [{{{\uppercase{{\bm{j}}}}}}_{n_1}^{(k_1)} \otimes \dots \otimes {{{\uppercase{{\bm{j}}}}}}_{n_d}^{(k_d)}] \otimes where is $n_j n_j$ binary matrix with $(p,q)$-th entry being 1 of $p-q=k_j$ and 0 elsewhere, and $$\begin{aligned} {{{\uppercase{{\bm{t}}}}}}_{{{{\lowercase{{\bm{k}}}}}}}=\frac{1}{(2\pi)^d} F({\hbox{\boldmath$\omega$}}) e^{-\jmath <{{{\lowercase{{\bm{k}}}}}},{\hbox{\boldmath$\omega$}}>} d {\hbox | ,gl} e^{\jmath \frac{1}{g}(k (\omega_1+2 \pi m_1) + l (\omEga_2 + 2 \pi m_2))}. \end{aLigneD}$$ By ThiS, tHe siNgulAr value distribUTion Of ${{{\uppercase{{\bm{t}}}}}}^g$ can be alTernaTiVEly sTUdIed on The geneRAtING fuNcTiOn $F({\HbOX{\bOldmaTh$\oMega$}},{{{\lowErcase{{\bm{m}}}}}})$.
HIghEr dimensional CONvOlution
------------------------------
AccOrdIng to [@Oudin2009], a bLocK multi-LeVel tOepliTz mAtrix ${{{\UppercASe{{\bm{t}}}}}}=\{{{{\uPpercase{{\bM{t}}}}}}_{{{{\LOwercaSE{{\bm{i}}}}}}-{{{\lowERCaSe{{\bm{J}}}}}}}\}_{{{{\lowercase{{\bm{i}}}}}},{{{\loweRCaSE{{\bm{j}}}}}}={{{\uppercase{{\bM{\mathrM{1}}}}}}}}^{{{\lOWeRCAse{{\Bm{n}}}}}}$ With ${{{\lowercAsE{{\bm{i}}}}}}=(i_1,\DOts,i_d)$, ${{{\loWErCASE{{\bm{J}}}}}}=(J_1,\dots,j_d)$, and ${{{\lowErcase{{\bm{n}}}}}}=(n_1,\dOTs,n_D)$, it can Be AltERnativEly rePrESenTed as $$\begin{aLignEd}
{{{\uppercaSe{{\bm{t}}}}}}= \sUM_{\abs{k_1}<n_1} \dOTs \sum_{\abS{k_d}<n_d} [{{{\uPpeRcaSe{{\bm{J}}}}}}_{N_1}^{(k_1)} \OtImeS \dOTs \oTImEs {{{\uPPerCase{{\bm{j}}}}}}_{n_D}^{(k_D)}] \oTimes {{{\UppeRCASE{{\bm{t}}}}}}_{{{{\LowErcaSe{{\bm{k}}}}}}}\End{aligned}$$ wheRe ${{{\uPperCAse{{\Bm{j}}}}}}_{n_j}^{(K_j)}$ is a $N_j \tiMeS n_j$ biNary maTrix wItH $(p,q)$-th entry being 1 Of $p-q=K_j$ and 0 elseWheRe, And $$\BeGin{alIGned}
{{{\upPerCasE{{\bm{t}}}}}}_{{{{\lowErcase{{\bM{K}}}}}}}=\frAc{1}{(2\PI)^D} \InT_{\Omega} F({\hbox{\boldmatH$\oMEGa$}}) E^{-\jmath <{{{\loWercasE{{\Bm{K}}}}}},{\hBOx{\boldmaTh$\OmeGa$}}>} d {\hBOX | ,gl} e^{\jmath \frac{1}{g} (k (\omega _1+2\pi m_ 1) + l (\o mega_2 + 2 \pi m_2) )}. \end{aligned}$$By th is , the si ngula r value di s t rib ut io n o f$ {{ {\upp erc ase{{\b m{t}}}}}}^ g$ca n be alterna t iv ely studie d o n the genera tin g func ti on$ F({\h box {\bol dmath$ \ omega$ }},{{{\lo we r case{{ \ bm{m}}} } } }) $.
Higher Dimensiona l C o nvolution
---- ------ -- - -- - - --- --- -------
A cc ordin g to [@O u di n 2 0 09] , a block mult i-level Toe p lit z matr ix ${ { {\uppe rcase {{ \ bm{ t}}}}}}=\{{ {{\u ppercase{ {\bm{t } }}}}}_{ { {{\lowe rcase{ {\b m{i }}}} } }- {{ {\l ow e rca s e{ {\b m {j} }}}}}}\} _{ {{ {\low erca s e { { \bm{ i}} }}}} ,{{{\ lowercase{{\b m{j }}}} } }={ {{\up perca se{{ \b m{\ma thrm{1 }}}}} }} }^{{{\lowercase {{\b m{n}}}}}} $ w it h $ {{ {\low e rcase{ {\b m{i }}}}}}= (i_1,\d o ts, i_ d ) $ ,${{{\lowercase{{\b m{ j } }} }}}=(j_1 ,\dots , j_ d) $ , and ${ {{ \lo werc a s e{{\b m{n} } }} }}=(n_1, \dots, n _d )$ , it ca nbe alt er nat ive ly re p rese nted a s $$\beg in{al i gned}
{{{\ u ppercase{{\bm { t} } } }} } = \s um_ {\abs{k_1}< n_1} \dot s \s u m_ {\a b s{k_d }<n_d }[ {{ { \uppercase{{\bm{j}} }} }}_{n_ 1}^{( k_1)} \otimes \dots \ot i m e s {{{\up perc a se { {\bm{j}}}}}}_{ n_d}^ {(k_d)}] \ o times {{ {\upp ercase{{ \bm{t}}}} } } _{{{{\lo wer cas e{{ \bm { k }} }}}}}\end{ali g n ed}$ $where $ {{{ \upperc ase {{\ bm{ j}} }} }}_{n_j}^ {(k_j)}$ i sa$n _j\time s n_j$ bi na ryma tri x wit h $(p,q )$-th ent ry b e ing 1 of $ p -q = k _j$an d0 el sew he re, a nd $ $ \be gin{ali gned}
{{ { \upp er ca se{{\bm {t}}}}}}_{{{{ \l owercase{{ \b m{k }}}}}} } = \frac{1} {(2\pi)^d} \int_{\Omega } F({\hb ox{ \bold math $\omega$} })e^{-\j mat h <{{{\ lowerc ase{{ \b m{k } } }}}}, { \ hb ox{ \b oldmath$\o m e ga$ }}>}d{\hb ox | ,gl} e^{\jmath_\frac{1}{g}(k (\omega_1+2_\pi m_1) + l_(\omega_2 +_2_\pi m_2))}.__ \end{aligned}$$ By_this, the singular_value distribution of ${{{\uppercase{{\bm{t}}}}}}^g$_can be alternatively_studied_on the generating function $F({\hbox{\boldmath$\omega$}},{{{\lowercase{{\bm{m}}}}}})$.
Higher Dimensional Convolution
------------------------------
According to [@Oudin2009], a block multi-level Toeplitz matrix_${{{\uppercase{{\bm{t}}}}}}=\{{{{\uppercase{{\bm{t}}}}}}_{{{{\lowercase{{\bm{i}}}}}}-{{{\lowercase{{\bm{j}}}}}}}\}_{{{{\lowercase{{\bm{i}}}}}},{{{\lowercase{{\bm{j}}}}}}={{{\uppercase{{\bm{\mathrm{1}}}}}}}}^{{{\lowercase{{\bm{n}}}}}}$_with ${{{\lowercase{{\bm{i}}}}}}=(i_1,\dots,i_d)$,_${{{\lowercase{{\bm{j}}}}}}=(j_1,\dots,j_d)$,_and_${{{\lowercase{{\bm{n}}}}}}=(n_1,\dots,n_d)$, it can be alternatively_represented as $$\begin{aligned}
_ {{{\uppercase{{\bm{t}}}}}}=_\sum_{\abs{k_1}<n_1} \dots \sum_{\abs{k_d}<n_d} [{{{\uppercase{{\bm{j}}}}}}_{n_1}^{(k_1)} \otimes \dots \otimes {{{\uppercase{{\bm{j}}}}}}_{n_d}^{(k_d)}]_\otimes_{{{\uppercase{{\bm{t}}}}}}_{{{{\lowercase{{\bm{k}}}}}}}\end{aligned}$$ where ${{{\uppercase{{\bm{j}}}}}}_{n_j}^{(k_j)}$_is a $n_j \times n_j$ binary matrix with $(p,q)$-th_entry being 1 of $p-q=k_j$ and_0 elsewhere, and_$$\begin{aligned}
__ {{{\uppercase{{\bm{t}}}}}}_{{{{\lowercase{{\bm{k}}}}}}}=\frac{1}{(2\pi)^d} \int_{\Omega}_F({\hbox{\boldmath$\omega$}}) e^{-\jmath <{{{\lowercase{{\bm{k}}}}}},{\hbox{\boldmath$\omega$}}>} d {\hbox |
phi_2\rangle_{AB}\!\!&=&\!\!\left[\beta(|HH\rangle+|VV\rangle)
+\sqrt{|\alpha|^2-|\beta|^2}|V'H\rangle\right]_{AB}
\nonumber\\
&&\otimes\gamma(|a_1b_1\rangle+|a_2b_2\rangle)+(\alpha|HH\rangle+\beta|VV\rangle)_{AB}
\nonumber\\
&&\otimes\sqrt{|\delta|^2-|\gamma|^2}\,|a_3b_2\rangle,\end{aligned}$$ where $\vert V'\rangle$ represents the vertical polarization of photon $A$ after an operation $R_\theta$. The wave plate $R_\theta$ is used to perform a rotate operation $\vert H\rangle\rightarrow {\rm cos}\theta
\vert H\rangle + {\rm sin}\theta \vert V\rangle$ on the horizontal polarization $\vert H\rangle$.
Finally, Alice lets two spatial modes $a_1$ and $a_2$ pass through PBS$_3$, PBS$_4$, DL, PBS$_5$ and PBS$_6$, and the state of the photon pair $AB$ is transformed from $|\phi_2\rangle_{AB}$ to $|\phi_3\rangle_{AB}$. Here $$\begin{aligned}
% Eq.(5) % Eq. 26
|\phi_3\rangle_{AB}\!&=&\!\beta\gamma(|HH\rangle+|VV\rangle)_{AB}(|a_1b_1\rangle
+|a_2b_2\rangle)
\nonumber\\
\!&&+\gamma\sqrt{|\alpha|^2-|\beta|^2}|VH\rangle_{AB}(|a'_1b_1\rangle
+|a'_2b_2\rangle)
\nonumber\\
\!&&+\sqrt{|\delta|^2-|\gamma|^2}(\alpha|HH\rangle
+\beta|VV\rangle)_{AB}|a_3b_2\rangle.\nonumber\\
\label{HBSoutcome}\end{aligned}$$ If photon $A$ is not detected in one of the spatial modes $a'_1$ and $a'_2$, the polarization state of the photon pair $AB$ is transformed into a maximally entangled Bell state. That is, the maximally hyperentangled Bell state $|\phi\rangle_{AB}$ is | phi_2\rangle_{AB}\!\!&=&\!\!\left[\beta(|HH\rangle+|VV\rangle)
+ \sqrt{|\alpha|^2-|\beta|^2}|V'H\rangle\right]_{AB }
\nonumber\\
& & \otimes\gamma(|a_1b_1\rangle+|a_2b_2\rangle)+(\alpha|HH\rangle+\beta|VV\rangle)_{AB }
\nonumber\\
& & \otimes\sqrt{|\delta|^2-|\gamma|^2}\,|a_3b_2\rangle,\end{aligned}$$ where $ \vert V'\rangle$ represents the vertical polarization of photon $ A$ after an operation $ R_\theta$. The wave plate $ R_\theta$ is use to do a rotate operation $ \vert H\rangle\rightarrow { \rm cos}\theta
\vert H\rangle + { \rm sin}\theta \vert V\rangle$ on the horizontal polarization $ \vert H\rangle$.
Finally, Alice permit two spatial mode $ a_1 $ and $ a_2 $ pass through PBS$_3 $, PBS$_4 $, DL, PBS$_5 $ and PBS$_6 $, and the state of the photon couple $ AB$ is transformed from $ |\phi_2\rangle_{AB}$ to $ |\phi_3\rangle_{AB}$. Here $ $ \begin{aligned }
% Eq.(5) % Eq. 26
|\phi_3\rangle_{AB}\!&=&\!\beta\gamma(|HH\rangle+|VV\rangle)_{AB}(|a_1b_1\rangle
+ |a_2b_2\rangle)
\nonumber\\
\!&&+\gamma\sqrt{|\alpha|^2-|\beta|^2}|VH\rangle_{AB}(|a'_1b_1\rangle
+ |a'_2b_2\rangle)
\nonumber\\
\!&&+\sqrt{|\delta|^2-|\gamma|^2}(\alpha|HH\rangle
+ \beta|VV\rangle)_{AB}|a_3b_2\rangle.\nonumber\\
\label{HBSoutcome}\end{aligned}$$ If photon $ A$ is not detected in one of the spatial modes $ a'_1 $ and $ a'_2 $, the polarization state of the photon couple $ AB$ is transform into a maximally entangled Bell department of state. That is, the maximally hyperentangled Bell state $ |\phi\rangle_{AB}$ is | phi_2\gangle_{AB}\!\!&=&\!\!\left[\beta(|HH\rangle+|YV\rangle)
+\sqrt{|\alphc|^2-|\veta|^2}|V'H\cangle\rjght]_{AB}
\novumber\\
&&\otimes\gamma(|a_1b_1\rangle+|a_2b_2\renglw)+(\alphq|HH\rangle+\beta|VV\rangle)_{AC}
\nonumber\\
&&\ltimes\sqet{|\deora|^2-|\gamma|^2}\,|a_3b_2\rehgle,\end{alignes}$$ whexe $\vert V'\rangle$ tepresents tve vertical ponafivation of photon $A$ after an operatiog $R_\thets$. Hhe wave plate $R_\thtta$ is hsed to perform a rotate operation $\vert H\gangle\rightarrow {\tm cos}\theta
\vert H\rangle + {\gm sln}\theta \vert V\rangpe$ on the hirizjbtal polarizxtion $\vert H\rangle$.
Finamly, Alice lets two spatial moder $a_1$ aud $a_2$ pass tyriugj PBS$_3$, PBS$_4$, DL, PBS$_5$ wnd PBS$_6$, and bne stade of tne photon pair $AB$ is transformed from $|\phi_2\cangle_{AB}$ to $|\phi_3\ranglg_{AB}$. Here $$\bagnn{aligned}
% Eq.(5) % Es. 26
|\phi_3\rwngme_{AB}\!&=&\!\beta\gajma(|HH\rangle+|CV\rangle)_{AB}(|a_1b_1\rangle
+|a_2n_2\rwbgle)
\nonumber\\
\!&&+\gzmma\sqwt{|\wlpha|^2-|\beta|^2}|VH\rangle_{AB}(|a'_1b_1\rangle
+|a'_2b_2\rangle)
\nonukbed\\
\!&&+\sqrt{|\delta|^2-|\gamma|^2}(\alpha|HH\rqngle
+\beta|VV\rangle)_{AB}|a_3b_2\tangle.\nonuiber\\
\label{HBSoutcome}\end{aligned}$$ If photon $A$ is not getecvea iu one ud hhe spatial modes $a'_1$ and $a'_2$, the polarization stwfe on the photon paiv $AB$ is transformec lnyj a maximally entanymes Bell state. That ls, the iaximqlly hypewentsngled Bell state $|\phi\rangle_{QB}$ is | phi_2\rangle_{AB}\!\!&=&\!\!\left[\beta(|HH\rangle+|VV\rangle) +\sqrt{|\alpha|^2-|\beta|^2}|V'H\rangle\right]_{AB} \nonumber\\ &&\otimes\gamma(|a_1b_1\rangle+|a_2b_2\rangle)+(\alpha|HH\rangle+\beta|VV\rangle)_{AB} \nonumber\\ &&\otimes\sqrt{|\delta|^2-|\gamma|^2}\,|a_3b_2\rangle,\end{aligned}$$ where represents vertical polarization photon $A$ after plate is used to a rotate operation H\rangle\rightarrow {\rm cos}\theta \vert H\rangle + sin}\theta \vert V\rangle$ on the horizontal polarization $\vert H\rangle$. Finally, Alice lets two modes $a_1$ and $a_2$ pass through PBS$_3$, PBS$_4$, DL, PBS$_5$ and PBS$_6$, and state the pair is transformed from $|\phi_2\rangle_{AB}$ to $|\phi_3\rangle_{AB}$. Here $$\begin{aligned} % Eq.(5) % Eq. 26 |\phi_3\rangle_{AB}\!&=&\!\beta\gamma(|HH\rangle+|VV\rangle)_{AB}(|a_1b_1\rangle +|a_2b_2\rangle) \nonumber\\ +|a'_2b_2\rangle) \nonumber\\ \!&&+\sqrt{|\delta|^2-|\gamma|^2}(\alpha|HH\rangle +\beta|VV\rangle)_{AB}|a_3b_2\rangle.\nonumber\\ \label{HBSoutcome}\end{aligned}$$ If photon $A$ not detected in one the spatial modes $a'_1$ and the state of photon $AB$ transformed into a entangled Bell state. That is, the maximally hyperentangled Bell state $|\phi\rangle_{AB}$ is | phi_2\rangle_{AB}\!\!&=&\!\!\left[\beta(|HH\rangLe+|VV\rangle)
+\Sqrt{|\aLphA|^2-|\beTa|^2}|v'H\raNgle\Right]_{AB}
\nonumbeR\\
&&\OtimEs\gamma(|a_1b_1\rangle+|a_2b_2\ranglE)+(\alphA|Hh\RangLE+\bEta|VV\Rangle)_{Ab}
\NoNUMbeR\\
&&\oTiMes\SqRT{|\dElta|^2-|\gAmmA|^2}\,|a_3b_2\rangLe,\end{alignEd}$$ wHeRe $\vert V'\ranglE$ RePresents thE veRtical polariZatIon of pHoTon $a$ After An oPeratIon $R_\thETa$. The wAve plate $R_\ThETa$ is usED to perfORM a RotaTe operation $\vert H\rANgLE\rightarrow {\rm cOs}\thetA
\vERt h\RAngLe + {\rM sin}\theta \vErT V\ranGLe$ on the HOrIZONtaL Polarization $\vErt H\rangle$.
FINalLy, AlicE lEts TWo spatIal moDeS $A_1$ anD $a_2$ pass throuGh PBs$_3$, PBS$_4$, DL, PBS$_5$ And PBS$_6$, ANd the stATe of the Photon PaiR $AB$ Is trANsFoRmeD fROm $|\pHI_2\rAngLE_{AB}$ To $|\phi_3\ranGlE_{Ab}$. Here $$\BegiN{ALIGned}
% eq.(5) % EQ. 26
|\phi_3\RanglE_{AB}\!&=&\!\beta\gamma(|Hh\raNgle+|vv\raNgle)_{Ab}(|a_1b_1\raNgle
+|A_2b_2\RanglE)
\nonumBer\\
\!&&+\gaMmA\sqrt{|\alpha|^2-|\beta|^2}|Vh\ranGle_{AB}(|a'_1b_1\raNglE
+|a'_2B_2\raNgLe)
\nonUMber\\
\!&&+\sqRt{|\dEltA|^2-|\gamma|^2}(\aLpha|HH\rANglE
+\bETA|vV\Rangle)_{AB}|a_3b_2\rangle.\noNuMBEr\\
\Label{HBSOutcomE}\EnD{aLIgned}$$ If pHoTon $a$ is nOT DetecTed iN OnE of the spAtial mODeS $a'_1$ And $a'_2$, the PoLarizaTiOn sTatE of thE PhotOn pair $aB$ is tranSformED into a maximallY Entangled Bell STaTE. thAT is, tHe mAximally hypErenTAnglEd BeLL sTatE $|\Phi\raNgle_{Ab}$ iS | phi_2\rangle_{AB}\!\!&=&\! \!\left[\b eta(| HH\ ran gl e+|V V\ra ngle)
+\sqrt{| \ alph a|^2-|\beta|^2}|V'H\ra ngle\ ri g ht]_ { AB }
\no number\ \
& & \ oti me s\ gam ma ( |a _1b_1 \ra ngle+|a _2b_2\rang le) +( \alpha|HH\ra n gl e+\beta|VV \ra ngle)_{AB}
\ non umber\ \&&\ o times \sq rt{|\ delta| ^ 2-|\ga mma|^2}\, |a _ 3b_2\r a ngle,\e n d {a lign ed}$$ where $\ver t V ' \rangle$ repre sentsth e v e r tic alpolarizati on of p h oton $A $ a f t e r a n operation $R _\theta$. T h e w ave pl at e $ R _\thet a$ is u s edto performa ro tate oper ation$ \vert H \ rangle\ righta rro w { \rmc os }\ the ta \ve r tH\r a ngl e + {\r msi n}\th eta\ v e r t V\ ran gle$ on t he horizontal po lari z ati on $\ vertH\ra ng le$.
Final ly, A li ce lets two spa tial modes $a _1$ a nd$a _2$ p a ss thr oug h P BS$_3$, PBS$_4 $ , D L, P B S$ _5$ and PBS$_6$, a nd t he state o f thep ho to n pair $A B$ is tra n s forme d fr o m$|\phi_2 \rangl e _{ AB }$ to $ |\ phi_3\ ra ngl e_{ AB}$. Here $$\be gin{alig ned} % Eq . ( 5) % E q. 26
|\phi _3\r a ngle _{AB } \! &=& \ !\bet a\gam ma ( |H H \rangle+|VV\rangle) _{ AB}(|a _1b_1 \rangle
+|a_2 b_2\rangle ) \ nonumber \\
\ ! && + \gamma\sqrt{|\ alpha |^2-|\beta | ^2}|VH\r angle _{AB}(|a '_1b_1\ra n g le
+|a'_ 2b_ 2\r ang le) \ no number\\
\!&& + \ sqrt {| \delta| ^2- |\gamma |^2 }(\ alp ha| HH \rangle
+ \beta|VV \r an gl e) _{A B}|a_ 3 b_2\rang le .\n on umb er\\\ label{ HBSou tcom e} \e n d{a ligned} $ $I f pho to n$A$isno t det ecte d in one of the spat ial mode s$a '_1$ an d $a'_2$, the p olarizatio nsta te oft h e photon pair $AB$ is transform e d intoa m axima llyentangled Be ll sta te. That i s, the maxi ma lly h ypere n t an gle dBell state $ |\p hi\ra ng le_{ AB}$ is | phi_2\rangle_{AB}\!\!&=&\!\!\left[\beta(|HH\rangle+|VV\rangle)
+\sqrt{|\alpha|^2-|\beta|^2}|V'H\rangle\right]_{AB}
\nonumber\\
&&\otimes\gamma(|a_1b_1\rangle+|a_2b_2\rangle)+(\alpha|HH\rangle+\beta|VV\rangle)_{AB}
\nonumber\\
&&\otimes\sqrt{|\delta|^2-|\gamma|^2}\,|a_3b_2\rangle,\end{aligned}$$ where_$\vert V'\rangle$_represents the vertical polarization_of photon_$A$_after an_operation_$R_\theta$. The wave_plate $R_\theta$ is_used to perform a_rotate operation $\vert_H\rangle\rightarrow_{\rm cos}\theta
\vert H\rangle + {\rm sin}\theta \vert V\rangle$ on the horizontal polarization $\vert_H\rangle$.
Finally,_Alice lets_two_spatial_modes $a_1$ and $a_2$ pass_through PBS$_3$, PBS$_4$, DL, PBS$_5$_and PBS$_6$,_and the state of the photon pair $AB$_is_transformed from $|\phi_2\rangle_{AB}$_to $|\phi_3\rangle_{AB}$. Here $$\begin{aligned}
_ _ ___ _ _ % Eq.(5) _% Eq. 26
|\phi_3\rangle_{AB}\!&=&\!\beta\gamma(|HH\rangle+|VV\rangle)_{AB}(|a_1b_1\rangle
+|a_2b_2\rangle)
\nonumber\\
\!&&+\gamma\sqrt{|\alpha|^2-|\beta|^2}|VH\rangle_{AB}(|a'_1b_1\rangle
+|a'_2b_2\rangle)
\nonumber\\
\!&&+\sqrt{|\delta|^2-|\gamma|^2}(\alpha|HH\rangle
+\beta|VV\rangle)_{AB}|a_3b_2\rangle.\nonumber\\
\label{HBSoutcome}\end{aligned}$$ If photon $A$_is not detected in one of_the spatial modes $a'_1$ and_$a'_2$, the_polarization state of the photon_pair $AB$ is_transformed into_a maximally entangled_Bell state. That is, the maximally_hyperentangled Bell state_$|\phi\rangle_{AB}$ is |
}^{\a,\b}(s,t) \, + & \frac{1 + \a + \b + n}{1 + \a + \b + 2 n}
d_{\a,\b}^{(1,1)} (x_1^2-x_2^2)(y_1^2-y_2^2) \\
& \times \left[ K_{n-1}^{\a+1,\b+1} (s,t) - K_{n-2}^{\a+1,\b+1} (s,t) \right] \notag \\
- & \frac{n (1 + \a + \b + n)}{ (1 + \a + \b + 2 n)^2} {}_2Q_{n-1,2n}(x){}_2Q_{n-1,2n}(y), \notag\end{aligned}$$ where $s= (2 x_1 x_2,x_1^2+x_2^2-1)$, $t = (2 y_1 y_2, y_1^2+y_2^2 -1)$, $\CK_{2n-1}^{\a,\b}(\cdot,\cdot)$ and $d_{\a,\b}^{(1,1)}$ are given in and in $K_n^{\a,\b}(\cdot,\cdot)$ is given in.
The formulas and are exactly those given in Theorem \[thm:minimal-Interp\], specialized to the Jacobi case. It remains to establish the formula of, for which we need to determine the constants $b_{k,n}$ in.
Throughout this proof, we write $Q_{k,2n}(x,y) = {}_2Q_{k,n}^{(\a,\b)}(x,y)$. By the explicit formula of $Q_{k,2n}$ in Proposition \[Q2nJacobi\], it is easy to verify that $$\begin{aligned}
\label{Qxjk}
& Q_{m,2n}\left(\xb_{j,k}^{(1)}\right) = \gamma_{\a,\b} \sqrt{1-x_j^2}\sqrt{1-x_k^2} \\
& \quad \times \left[ p_{n-1}^{(\a+1,\b+1)}(x_k | } ^{\a,\b}(s, t) \, + & \frac{1 + \a + \b + n}{1 + \a + \b + 2 n }
d_{\a,\b}^{(1,1) } (x_1 ^ 2 - x_2 ^ 2)(y_1 ^ 2 - y_2 ^ 2) \\
& \times \left [ K_{n-1}^{\a+1,\b+1 } (s, t) - K_{n-2}^{\a+1,\b+1 } (s, t) \right ] \notag \\
- & \frac{n (1 + \a + \b + n) } { (1 + \a + \b + 2 n)^2 } { } _ 2Q_{n-1,2n}(x){}_2Q_{n-1,2n}(y), \notag\end{aligned}$$ where $ s= (2 x_1 x_2,x_1 ^ 2+x_2 ^ 2 - 1)$, $ t = (2 y_1 y_2, y_1 ^ 2+y_2 ^ 2 -1)$, $ \CK_{2n-1}^{\a,\b}(\cdot,\cdot)$ and $ d_{\a,\b}^{(1,1)}$ are given in and in $ K_n^{\a,\b}(\cdot,\cdot)$ is given in.
The formulas and are precisely those render in Theorem \[thm: minimal - Interp\ ], specialized to the Jacobi case. It remains to lay down the formula of, for which we need to specify the constants $ b_{k, n}$ in.
Throughout this proof, we publish $ Q_{k,2n}(x, y) = { } _ 2Q_{k, n}^{(\a,\b)}(x, y)$. By the explicit formula of $ Q_{k,2n}$ in Proposition \[Q2nJacobi\ ], it is comfortable to verify that $ $ \begin{aligned }
\label{Qxjk }
& Q_{m,2n}\left(\xb_{j, k}^{(1)}\right) = \gamma_{\a,\b } \sqrt{1 - x_j^2}\sqrt{1 - x_k^2 } \\
& \quad \times \left [ p_{n-1}^{(\a+1,\b+1)}(x_k | }^{\a,\b}(s,h) \, + & \frac{1 + \a + \b + n}{1 + \a + \b + 2 n}
d_{\a,\b}^{(1,1)} (x_1^2-x_2^2)(b_1^2-y_2^2) \\
& \times \left[ K_{n-1}^{\a+1,\b+1} (s,t) - K_{n-2}^{\a+1,\b+1} (d,t) \righu] \notag \\
- & \ffac{n (1 + \a + \b + n)}{ (1 + \q + \b + 2 n)^2} {}_2Q_{n-1,2n}(x){}_2Q_{n-1,2i}(g), \notag\cud{alifked}$$ wkece $s= (2 x_1 x_2,x_1^2+x_2^2-1)$, $t = (2 y_1 y_2, y_1^2+y_2^2 -1)$, $\CN_{2n-1}^{\a,\b}(\cdot,\cdot)$ atd $d_{\c,\b}^{(1,1)}$ are given in and in $K_n^{\a,\b}(\cdot,\cdot)$ ys givem ln.
The formulas and wre sqagtly those given in Theorem \[thm:mjnimal-Iiterp\], specializrd to the Jacobi case. It rfmaijs to establish thf formula od, fow which we nedd to detegkine the cknstants $b_{k,n}$ in.
Throughout this oroof, we write $W_{k,2b}(x,y) = {}_2Q_{k,n}^{(\a,\b)}(x,y)$. By the vxplicit formmka of $X_{k,2n}$ in Lroposition \[Q2nMacobm\], it is easy to verify thet $$\begin{aligned}
\labej{Qxjk}
& X_{m,2u}\left(\xb_{j,k}^{(1)}\right) = \gamma_{\q,\b} \sqrt{1-f_j^2}\sqst{1-x_k^2} \\
& \qnad \timed \lxft[ p_{n-1}^{(\a+1,\b+1)}(x_k | }^{\a,\b}(s,t) \, + & \frac{1 + \a + + \a \b + 2 \times K_{n-1}^{\a+1,\b+1} (s,t) - (s,t) \right] \notag - & \frac{n (1 + \a \b + n)}{ (1 + \a + \b + 2 n)^2} {}_2Q_{n-1,2n}(x){}_2Q_{n-1,2n}(y), \notag\end{aligned}$$ $s= (2 x_1 x_2,x_1^2+x_2^2-1)$, $t = (2 y_1 y_2, y_1^2+y_2^2 -1)$, $\CK_{2n-1}^{\a,\b}(\cdot,\cdot)$ and are in in is given in. The formulas and are exactly those given in Theorem \[thm:minimal-Interp\], specialized to the case. It remains to establish the formula of, which we need to the constants $b_{k,n}$ in. Throughout proof, write $Q_{k,2n}(x,y) {}_2Q_{k,n}^{(\a,\b)}(x,y)$. the formula of $Q_{k,2n}$ Proposition \[Q2nJacobi\], it is easy to verify that $$\begin{aligned} \label{Qxjk} & Q_{m,2n}\left(\xb_{j,k}^{(1)}\right) = \gamma_{\a,\b} \sqrt{1-x_j^2}\sqrt{1-x_k^2} \\ & \times \left[ | }^{\a,\b}(s,t) \, + & \frac{1 + \a + \b + n}{1 + \a + \b + 2 n}
d_{\a,\b}^{(1,1)} (x_1^2-x_2^2)(y_1^2-y_2^2) \\
& \timeS \left[ K_{n-1}^{\a+1,\b+1} (s,T) - K_{n-2}^{\a+1,\b+1} (S,t) \rIghT] \nOtag \\
- & \Frac{N (1 + \a + \b + n)}{ (1 + \a + \b + 2 n)^2} {}_2Q_{n-1,2n}(x){}_2Q_{n-1,2n}(Y), \NotaG\end{aligned}$$ where $s= (2 x_1 x_2,x_1^2+x_2^2-1)$, $t = (2 Y_1 y_2, y_1^2+y_2^2 -1)$, $\Ck_{2n-1}^{\A,\B}(\cdoT,\CdOt)$ and $D_{\a,\b}^{(1,1)}$ are gIVeN IN anD iN $K_N^{\a,\b}(\CdOT,\cDot)$ is GivEn in.
The Formulas anD arE eXactly those gIVeN in Theorem \[Thm:Minimal-InterP\], spEcialiZeD to THe JacObi Case. IT remaiNS to estAblish the FoRMula of, FOr which WE NeEd to Determine the constANtS $B_{k,n}$ in.
ThroughouT this pRoOF, wE WRitE $Q_{k,2N}(x,y) = {}_2Q_{k,n}^{(\a,\b)}(x,y)$. by The exPLicit foRMuLA OF $Q_{k,2N}$ In Proposition \[q2nJacobi\], it iS EasY to verIfY thAT $$\begin{AlignEd}
\LAbeL{Qxjk}
& Q_{m,2n}\lefT(\xb_{j,K}^{(1)}\right) = \gamMa_{\a,\b} \sqRT{1-x_j^2}\sqrt{1-X_K^2} \\
& \quad \tiMes \lefT[ p_{n-1}^{(\A+1,\b+1)}(x_K | }^{\a,\b}(s,t) \, + & \ frac{1 + \ a + \ b + n} {1 + \ a +\b + 2 n}
d_{\a,\b}^{(1,1)} (x_1 ^2-x_ 2^ 2 )(y_ 1 ^2 -y_2^ 2) \\
& \t im e s \le ft[ K_{n-1 }^{\a+1,\b +1} ( s,t) - K_{n- 2 }^ {\a+1,\b+1 } ( s,t) \right] \n otag \ \ - &\frac {n (1+ \a +\b + n)}{ ( 1 + \a+ \b + 2 n )^ 2} { }_2Q_{n-1,2n}(x){ } _2 Q _{n-1,2n}(y),\notag \e n d{ a l ign ed} $$ where $ s= (2 x _ 1 x_2,x _ 1^ 2 + x _2^ 2 -1)$, $t = (2 y_1 y_2, y _ 1^2 +y_2^2 - 1)$ , $\CK_ {2n-1 }^ { \a, \b}(\cdot,\ cdot )$ and $d _{\a,\ b }^{(1,1 ) }$ aregiveninand in$ K_ n^ {\a ,\ b }(\ c do t,\ c dot )$ is gi ve nin.
Thef o r m ulas an d ar e exa ctly those gi ven inT heo rem \ [thm: mini ma l-Int erp\], spec ia lized to the Ja cobi case. It re ma ins t o est a blishthe fo rmula o f, forw hic hw e ne ed to determine th ec o ns tants $b _{k,n} $ i n.
Through ou t t hisp r oof,we w r it e $Q_{k, 2n}(x, y )={}_2Q_{ k, n}^{(\ a, \b) }(x ,y)$. By t he exp licit fo rmula of $Q_{k,2n}$i n Proposition \[ Q 2 nJ a cobi \], it is easy tov erif y th a t$$\ b egin{ align ed }
\ label{Qxjk}
& Q_ {m ,2n}\l eft(\ xb_{j,k}^{(1) }\right) = \ g amma_{\a ,\b} \s q rt{1-x_j^2}\sq rt{1- x_k^2} \\ & \ quad \times\left[ p_ { n -1}^{(\a +1, \b+ 1)} (x_ k | }^{\a,\b}(s,t) \,_ +_& \frac{1_+ \a_+_\b +_n}{1_+ \a +_\b + 2_n}
_ __ d_{\a,\b}^{(1,1)} (x_1^2-x_2^2)(y_1^2-y_2^2) \\
__& \times___\left[ K_{n-1}^{\a+1,\b+1} (s,t) - K_{n-2}^{\a+1,\b+1}_(s,t) \right] \notag \\
_ _ - & \frac{n (1_+_\a + \b_+ n)}{ (1 + \a + \b + 2_n)^2} {}_2Q_{n-1,2n}(x){}_2Q_{n-1,2n}(y), \notag\end{aligned}$$ where $s= (2_x_1 x_2,x_1^2+x_2^2-1)$, $t_=_(2_y_1 y_2, y_1^2+y_2^2 -1)$,_$\CK_{2n-1}^{\a,\b}(\cdot,\cdot)$ and $d_{\a,\b}^{(1,1)}$ are given in_and in $K_n^{\a,\b}(\cdot,\cdot)$ is given in.
The_formulas and are exactly those given in_Theorem \[thm:minimal-Interp\], specialized to the Jacobi_case. It remains to establish_the formula_of, for which we need_to determine the_constants $b_{k,n}$_in.
Throughout this proof,_we write $Q_{k,2n}(x,y) = {}_2Q_{k,n}^{(\a,\b)}(x,y)$. By_the explicit formula_of $Q_{k,2n}$ in Proposition \[Q2nJacobi\], it_is_easy to verify_that_$$\begin{aligned}
_\label{Qxjk}
_& Q_{m,2n}\left(\xb_{j,k}^{(1)}\right)_=_\gamma_{\a,\b} \sqrt{1-x_j^2}\sqrt{1-x_k^2}__\\
&_\quad_ \times \left[ p_{n-1}^{(\a+1,\b+1)}(x_k |
x_1)^{\frac{3}{4}}
(z-x_2)^{\frac{3}{4}}
(\bar{z}-x_1)^{\frac{3}{4}}
(\bar{z}-x_2)^{\frac{3}{4}}
(x_1-x_2)^{\frac{3}{4}}
}.\end{aligned}$$ Thus the disk amplitude becomes $$\begin{aligned}
& - \zeta_{\mu \nu \rho \sigma}^{\mathrm{RR}}
Tr \mathbf{P}
\exp
\left(
i \int d \tau k \cdot A
\right) [\epsilon^e, \Phi_\alpha]
(\gamma^{\mu \nu \rho \sigma})_{ab} (\gamma_\beta)_{e c} \epsilon_d {\notag \\}& \times
\int d x_2
\left(
\frac{(\gamma^{\alpha \beta \tau})^{ad}(\gamma_\tau)^{bc}}{x_1-z}
- \frac{(\gamma^{\tau})^{ad}(\gamma^{\alpha \beta}_{\ \ \ \tau})^{bc}}{x_1-\bar{z}}
- \frac{(\gamma^{\tau \alpha \beta})^{ad}(\gamma_\tau)^{bc}}{x_1-x_2}
\right) {\notag \\}& \times \frac{ (z-\bar{z})^{\frac{1}{2}}
(z-x_1)^{\frac{1}{2}}
(\bar{z}-x_1)
(x_1-x_2)^{\frac{1}{2}}
}
{(z-x_2)
(\bar{z}-x_2)^{\frac{3}{2}}
} {\notag \\}= & - \zeta_{\mu \nu \rho \sigma}^{\mathrm{RR}}
Tr \mathbf{P}
\exp
\left(
i \int d \tau k \cdot A
\right)
[\epsilon^e, \Phi_\alpha]
(\gamma_\beta \gamma_\tau
\gamma^{\mu \nu \rho \sigma}
\gamma^{\alpha \beta \tau})_e^{\ d}
\epsilon_d {\notag \\}& | x_1)^{\frac{3}{4 } }
(z - x_2)^{\frac{3}{4 } }
(\bar{z}-x_1)^{\frac{3}{4 } }
(\bar{z}-x_2)^{\frac{3}{4 } }
(x_1 - x_2)^{\frac{3}{4 } }
} .\end{aligned}$$ Thus the disk amplitude becomes $ $ \begin{aligned }
& - \zeta_{\mu \nu \rho \sigma}^{\mathrm{RR } }
Tr \mathbf{P }
\exp
\left (
i \int d \tau thousand \cdot A
\right) [ \epsilon^e, \Phi_\alpha ]
(\gamma^{\mu \nu \rho \sigma})_{ab } (\gamma_\beta)_{e coulomb } \epsilon_d { \notag \\ } & \times
\int d x_2
\left (
\frac{(\gamma^{\alpha \beta \tau})^{ad}(\gamma_\tau)^{bc}}{x_1 - omega }
- \frac{(\gamma^{\tau})^{ad}(\gamma^{\alpha \beta}_{\ \ \ \tau})^{bc}}{x_1-\bar{z } }
- \frac{(\gamma^{\tau \alpha \beta})^{ad}(\gamma_\tau)^{bc}}{x_1 - x_2 }
\right) { \notag \\ } & \times \frac { (z-\bar{z})^{\frac{1}{2 } }
(z - x_1)^{\frac{1}{2 } }
(\bar{z}-x_1)
(x_1 - x_2)^{\frac{1}{2 } }
}
{ (z - x_2)
(\bar{z}-x_2)^{\frac{3}{2 } }
} { \notag \\}= & - \zeta_{\mu \nu \rho \sigma}^{\mathrm{RR } }
Tr \mathbf{P }
\exp
\left (
i \int five hundred \tau k \cdot A
\right)
[ \epsilon^e, \Phi_\alpha ]
(\gamma_\beta \gamma_\tau
\gamma^{\mu \nu \rho \sigma }
\gamma^{\alpha \beta \tau})_e^{\ d }
\epsilon_d { \notag \\ } & | x_1)^{\frwc{3}{4}}
(z-x_2)^{\frac{3}{4}}
(\bar{z}-x_1)^{\frac{3}{4}}
(\bad{z}-x_2)^{\frac{3}{4}}
(x_1-x_2)^{\frac{3}{4}}
}.\env{alitned}$$ Ukus the disk amplituae becomed $$\begin{aoigntd}
& - \zeta_{\mu \nu \rho \sigma}^{\mathrm{RD}}
Vr \mathbf{P}
\gxp
\left(
i \int d \tau k \cdot A
\right) [\qpsilon^r, \Ohi_\alpha]
(\gwmma^{\kt \nu \ghi \sigma})_{ab} (\gamma_\beta)_{e c} \epsiloh_d {\notaj \\}& \times
\int c x_2
\left(
\frac{(\galma^{\appha \beta \tau})^{ad}(\gamla_\tau)^{bc}}{x_1-z}
- \drac{(\gamma^{\tau})^{xd}(\gamma^{\alpha \beta}_{\ \ \ \tao})^{bc}}{x_1-\bar{z}}
- \frac{(\gamma^{\tau \alphx \betc})^{ad}(\gamma_\tau)^{vc}}{z_1-x_2}
\right) {\notaj \\}& \timvs \frac{ (z-\bav{d})^{\frac{1}{2}}
(z-x_1)^{\frsc{1}{2}}
(\bar{z}-x_1)
(v_1-x_2)^{\frac{1}{2}}
}
{(z-x_2)
(\gar{z}-x_2)^{\fgac{3}{2}}
} {\notag \\}= & - \zeta_{\mu \nu \rho \sogiq}^{\mathrm{RR}}
Tr \maehff{P}
\exp
\left(
i \pnt s \tau k \cdot A
\right)
[\epsilon^e, \Pji_\alpha]
(\gamma_\beta \gamma_\tau
\gamma^{\mu \nu \rho \sigka}
\gwona^{\wlpha \beta \tau})_e^{\ d}
\epsilon_d {\notag \\}& | x_1)^{\frac{3}{4}} (z-x_2)^{\frac{3}{4}} (\bar{z}-x_1)^{\frac{3}{4}} (\bar{z}-x_2)^{\frac{3}{4}} (x_1-x_2)^{\frac{3}{4}} }.\end{aligned}$$ Thus amplitude $$\begin{aligned} & \zeta_{\mu \nu \rho i d \tau k A \right) [\epsilon^e, (\gamma^{\mu \nu \rho \sigma})_{ab} (\gamma_\beta)_{e c} {\notag \\}& \times \int d x_2 \left( \frac{(\gamma^{\alpha \beta \tau})^{ad}(\gamma_\tau)^{bc}}{x_1-z} - \frac{(\gamma^{\tau})^{ad}(\gamma^{\alpha \beta}_{\ \ \tau})^{bc}}{x_1-\bar{z}} - \frac{(\gamma^{\tau \alpha \beta})^{ad}(\gamma_\tau)^{bc}}{x_1-x_2} \right) {\notag \\}& \times \frac{ (z-\bar{z})^{\frac{1}{2}} (z-x_1)^{\frac{1}{2}} (x_1-x_2)^{\frac{1}{2}} {(z-x_2) } \\}= & - \zeta_{\mu \nu \rho \sigma}^{\mathrm{RR}} Tr \mathbf{P} \exp \left( i \int d \tau k A \right) [\epsilon^e, \Phi_\alpha] (\gamma_\beta \gamma_\tau \gamma^{\mu \nu \sigma} \gamma^{\alpha \beta \tau})_e^{\ \epsilon_d {\notag \\}& | x_1)^{\frac{3}{4}}
(z-x_2)^{\frac{3}{4}}
(\bar{z}-x_1)^{\frac{3}{4}}
(\bar{z}-x_2)^{\Frac{3}{4}}
(x_1-x_2)^{\frac{3}{4}}
}.\End{alIgnEd}$$ THuS the Disk Amplitude becomES $$\begIn{aligned}
& - \zeta_{\mu \nu \rho \siGma}^{\maThRM{RR}}
TR \MaThbf{P}
\Exp
\left(
I \InT D \Tau K \cDoT A
\rIgHT) [\ePsiloN^e, \PHi_\alpha]
(\Gamma^{\mu \nu \rHo \sIgMa})_{ab} (\gamma_\betA)_{E c} \Epsilon_d {\noTag \\}& \Times
\int d x_2
\leFt(
\fRac{(\gamMa^{\AlpHA \beta \Tau})^{Ad}(\gamMa_\tau)^{bC}}{X_1-z}
- \frac{(\Gamma^{\tau})^{aD}(\gAMma^{\alpHA \beta}_{\ \ \ \taU})^{BC}}{x_1-\Bar{z}}
- \Frac{(\gamma^{\tau \alpha \BEtA})^{Ad}(\gamma_\tau)^{bc}}{x_1-x_2}
\Right) {\nOtAG \\}& \tIMEs \fRac{ (Z-\bar{z})^{\frac{1}{2}}
(z-X_1)^{\fRac{1}{2}}
(\baR{Z}-x_1)
(x_1-x_2)^{\fraC{1}{2}}
}
{(Z-x_2)
(\BAR{Z}-x_2)^{\fRAc{3}{2}}
} {\notag \\}= & - \zeta_{\mu \Nu \rho \sigma}^{\mAThrM{RR}}
Tr \mAtHbf{p}
\Exp
\lefT(
i \int D \tAU k \cDot A
\right)
[\epSiloN^e, \Phi_\alphA]
(\gamma_\BEta \gammA_\Tau
\gammA^{\mu \nu \rHo \sIgmA}
\gamMA^{\aLpHa \bEtA \Tau})_E^{\ D}
\ePsiLOn_d {\Notag \\}& | x_1)^{\frac{3}{4}}
(z-x_2 )^{\f rac {3} {4 }}
(\bar{z} - x_1) ^{\frac{3}{4}}
( \b a r{z} - x_ 2)^{\ frac{3} { 4} } ( x _1 -x_2) ^{\ frac{3} {4}}
}.\end{align e d} $$ Thus th e d isk amplitud e b ecomes $ $\b e gin{a lig ned} & - \ z eta_{\ mu \nu \r ho \sigma } ^{\math r m {R R}} Tr \mathbf{P }
\exp
\l ef t ( i \int d \tau k \cdot A \right) [\ epsilon^e,\ Phi _\alph a]
(\ga mma^{ \m u \n u \rho \sig ma}) _{ab} (\g amma_\ b eta)_{e c} \eps ilon_d {\ not ag \ \ }& \ tim es \ in t d x_2
\le ft (
\ f r a c{(\ gam ma^{ \alph a \beta \tau} )^{ ad}( \ gam ma_\t au)^{ bc}} {x _1-z}
- \f ra c{(\gamma^{\tau })^{ ad}(\gamm a^{ \a lph a\beta } _{\ \\ \ tau })^{bc} }{x_1-\ b ar{ z} } - \frac{(\gamma ^{ \ t au \alpha\beta} ) ^{ ad } (\gamma_ \t au) ^{bc } } {x_1- x_2} \right) {\not a g\\ }& \tim es \frac { ( z-\ bar{z } )^{\ frac{1 }{2}}
(z- x _1)^{\frac{1} { 2} } (\ bar{ z }-x_ 1)
( x _1 - x_2)^{\frac{1}{2}} }
{(z- x _ 2 )
(\bar{z}-x_2 )^{\f rac{3}{2}} } {\no tag \\}=& - \zeta_ {\m u \ nu\rh o \s igma}^{\mathr m { RR}}
Tr\ma thbf{P}
\ex p
\l eft(
i\intd \tau k\c dot A
\rig ht)
[\ ep s ilo n^e, \P h i_ \ a lpha ]
(\ ga mma_\ beta \ga mma_\ta u
\g a mma^ {\ mu \nu \r ho \sigma}
\gamma ^{ \al pha \b e t a \tau}) _e^{\ d}
\epsilon_ d {\n ota g \\} & | x_1)^{\frac{3}{4}}
_ _ _ __(z-x_2)^{\frac{3}{4}}
__ _ _ (\bar{z}-x_1)^{\frac{3}{4}}
_ __ (\bar{z}-x_2)^{\frac{3}{4}}
__ (x_1-x_2)^{\frac{3}{4}}
___ _ }.\end{aligned}$$ Thus the_disk amplitude_becomes $$\begin{aligned}
& - \zeta_{\mu \nu \rho \sigma}^{\mathrm{RR}}
__ _Tr \mathbf{P}
\exp
_ _ \left(
___ _ i \int_d \tau k \cdot A
_ _ \right) [\epsilon^e, \Phi_\alpha]
_ (\gamma^{\mu \nu \rho_\sigma})_{ab} (\gamma_\beta)_{e_c} \epsilon_d {\notag \\}& \times
_ \int_d x_2_
_\left(
_ _\frac{(\gamma^{\alpha \beta \tau})^{ad}(\gamma_\tau)^{bc}}{x_1-z}
__ - \frac{(\gamma^{\tau})^{ad}(\gamma^{\alpha_\beta}_{\_\_\ \tau})^{bc}}{x_1-\bar{z}}
_ __ -_\frac{(\gamma^{\tau_\alpha \beta})^{ad}(\gamma_\tau)^{bc}}{x_1-x_2}
\right) {\notag_\\}&_\times \frac{ (z-\bar{z})^{\frac{1}{2}}
_ __ _ _ (z-x_1)^{\frac{1}{2}}
_ _ _ _(\bar{z}-x_1)
_ _ __ __ _(x_1-x_2)^{\frac{1}{2}}
_ _ _}
__ _ {(z-x_2)
_ ___ _ (\bar{z}-x_2)^{\frac{3}{2}}
_ _ __} {\notag \\}= & - \zeta_{\mu_\nu_\rho \sigma}^{\mathrm{RR}}
_Tr_\mathbf{P}
_ \exp
_ _ \left(
_ _ i \int_d_\tau_k \cdot A
_ _\right)
_ [\epsilon^e, \Phi_\alpha]
_(\gamma_\beta \gamma_\tau
_ \gamma^{\mu \nu \rho \sigma}
_ \gamma^{\alpha \beta \tau})_e^{\ d}
_ \epsilon_d {\notag_\\}& |
y^2}-\frac{q}{y}\
\frac{\partial}{\partial y}\nonumber \end{aligned}$$ where $(p,q)$ are $c$-numbers.
It is assumed that the Hamiltonian (2) is a constraint for a classical Hamiltonain with the mass term present for the scalar field $\phi$. In other words, the canonical quantization needs the annihilation of the wave function $\psi$ by the corresponding quantum operator $$\hat{H}\psi =0$$ that results in the Wheeler-De Witt equation. Using eq. (2-5) we get $$\frac{\Lambda}{2R^2}\left(\frac{\partial^2\psi}{\partial \Lambda^2}+
\frac{p}{\Lambda}\ \frac{\partial \psi}{\partial \Lambda}\right)+
\frac{1}{2R^2\Lambda}\left(\frac{\partial^2\Lambda}{\partial \phi^2}+
\frac{q}{\phi}\frac{\partial \psi}{\partial \phi}\right)-
\frac{1}{R}\ \frac{\partial^2\psi}{\partial R\partial \Lambda}\equiv
V \psi$$ where $\psi$ is a functional of $\Lambda$, $\phi$ and $R$ functions, and $V$ is a potential term written as $$V=\frac{R}{\Lambda}\ R''-\frac{R}{\Lambda^2}\ R'\Lambda '+\frac{1}{2\Lambda}\
R{'}^2-\frac{1}{2}\ \Lambda +\frac{1}{2}\ \frac{R^2}{\Lambda}\ \phi{'}^2 +
\frac{1}{2}\ m^2R^2\Lambda \phi^2$$ The prime means derivative with respect to the coordinate $\underline{r}$. Observe that in equation (6) we don’t have any derivative with respect to time. This means that the equation (6) could be describing a spherically symmetric gravitational collapse but without any explicit time dependence for functional $\psi$. The concept of “time” in this case may appear only after quantization in accordance with [@tres].
This suggests that eq. (6) is like the usual Schrödinger equation of quantum mechanics applied to gravitational collapse but with a difference depending on the operator ordering \[1-5\].
The usual Schrödinger equation is written as $$H\psi =i\frac{\partial \psi}{\partial t}$$ where | y^2}-\frac{q}{y}\
\frac{\partial}{\partial y}\nonumber \end{aligned}$$ where $ (p, q)$ are $ c$-numbers.
It is assumed that the Hamiltonian (2) is a constraint for a classical Hamiltonain with the aggregate condition present for the scalar field $ \phi$. In other actor's line, the canonic quantization needs the annihilation of the wave function $ \psi$ by the represent quantum hustler $ $ \hat{H}\psi = 0$$ that results in the Wheeler - De Witt equation. Using eq. (2 - 5) we get $ $ \frac{\Lambda}{2R^2}\left(\frac{\partial^2\psi}{\partial \Lambda^2}+
\frac{p}{\Lambda}\ \frac{\partial \psi}{\partial \Lambda}\right)+
\frac{1}{2R^2\Lambda}\left(\frac{\partial^2\Lambda}{\partial \phi^2}+
\frac{q}{\phi}\frac{\partial \psi}{\partial \phi}\right)-
\frac{1}{R}\ \frac{\partial^2\psi}{\partial R\partial \Lambda}\equiv
V \psi$$ where $ \psi$ is a functional of $ \Lambda$, $ \phi$ and $ R$ functions, and $ V$ is a likely term written as $ $ V=\frac{R}{\Lambda}\ R''-\frac{R}{\Lambda^2}\ R'\Lambda' + \frac{1}{2\Lambda}\
R{'}^2-\frac{1}{2}\ \Lambda + \frac{1}{2}\ \frac{R^2}{\Lambda}\ \phi{'}^2 +
\frac{1}{2}\ m^2R^2\Lambda \phi^2$$ The prime means derivative with regard to the coordinate $ \underline{r}$. Observe that in equation (6) we do n’t get any derivative with respect to clock time. This means that the equation (6) could be describing a spherically symmetrical gravitational collapse but without any explicit clock time dependence for running $ \psi$. The concept of “ time ” in this character may appear only after quantization in accordance with [ @tres ].
This hint that eq. (6) is like the usual Schrödinger equation of quantum mechanics applied to gravitational collapse but with a remainder depending on the operator ordering \[1 - 5\ ].
The usual Schrödinger equation is written as $ $ H\psi = i\frac{\partial \psi}{\partial t}$$ where | y^2}-\fgac{q}{y}\
\frac{\partial}{\partial y}\nonumber \end{aligned}$$ xhere $(p,s)$ are $c$-njmbers.
It is assumed that the Hqmiltinian (2) is a constraint for a clwssical Yamioronain witi the mass terj prevxnt for the scakar field $\[hi$. In other wmras, the canonical quantization needs tre anninipation of the rave suncfpok $\psi$ by the corresponding quanthm operetor $$\hat{H}\psi =0$$ tnat results in the Wheeler-Fe Wltt equation. Using eq. (2-5) we get $$\fras{\Oambda}{2R^2}\left(\ffac{\partial^2\ksn}{\partial \Lajbda^2}+
\frac{p}{\Lambda}\ \frac{\partial \psi}{\oarticl \Lambda}\rithr)+
\frww{1}{2R^2\Lambda}\lefv(\frac{\pwrtial^2\Lambda}{\isrtial \phi^2}+
\frav{q}{\phi}\frac{\partisl \'si}{\pqrtial \phi}\right)-
\frac{1}{R}\ \hrac{\partial^2\psi}{\partiaj R\partian \Mambda}\equiv
V \psi$$ qhwre $\pvi$ iv a wyncgiohak kf $\Lamhda$, $\phi$ and $R$ functions, qnd $V$ is a potentiak evtm written aa $$V=\fras{R}{\Jambda}\ R''-\frac{R}{\Lambda^2}\ R'\Lambda '+\frac{1}{2\Lambda}\
R{'}^2-\xrad{1}{2}\ \Lambda +\frac{1}{2}\ \frac{R^2}{\Lambea}\ \phi{'}^2 +
\frac{1}{2}\ m^2R^2\Lambda \khi^2$$ The pryme means derivative with respect to the coordinade $\unvefliuc{v}$. Obrwrge that in equation (6) we don’t have any derivatybe wpth respect to tiie. This meams tnwt the equatiun (6) cobmd be describing a sohericajly stmmetric dravotational collapse but withiut any explpcit time dependence flr functioncl $\psi$. The voncept of “time” in this cass may appeag only affdr quantization kn scwordance with [@tres].
This sugdests thav eq. (6) is likd thg usual Schrödingeg equation of quantum mefhanies ap[lied to ggavitational collapse but with e difference ce[enging on che opcrator ordering \[1-5\].
The usual Schtödinger ezuatiun is writnen as $$H\pvi =i\frac{\pareial \psi}{\partidp t}$$ where | y^2}-\frac{q}{y}\ \frac{\partial}{\partial y}\nonumber \end{aligned}$$ where $(p,q)$ are is that the (2) is a with mass term present the scalar field In other words, the canonical quantization the annihilation of the wave function $\psi$ by the corresponding quantum operator $$\hat{H}\psi that results in the Wheeler-De Witt equation. Using eq. (2-5) we get $$\frac{\Lambda}{2R^2}\left(\frac{\partial^2\psi}{\partial \frac{p}{\Lambda}\ \psi}{\partial \frac{1}{2R^2\Lambda}\left(\frac{\partial^2\Lambda}{\partial \frac{q}{\phi}\frac{\partial \psi}{\partial \phi}\right)- \frac{1}{R}\ \frac{\partial^2\psi}{\partial R\partial \Lambda}\equiv V \psi$$ where $\psi$ is a functional of $\Lambda$, and $R$ functions, and $V$ is a potential written as $$V=\frac{R}{\Lambda}\ R''-\frac{R}{\Lambda^2}\ '+\frac{1}{2\Lambda}\ R{'}^2-\frac{1}{2}\ \Lambda +\frac{1}{2}\ \frac{R^2}{\Lambda}\ + m^2R^2\Lambda \phi^2$$ prime derivative respect to the $\underline{r}$. Observe that in equation (6) we don’t have any derivative with respect to time. This means the equation be describing spherically gravitational but without any dependence for functional $\psi$. The concept this case may appear only after quantization in with [@tres]. suggests that eq. (6) is like usual Schrödinger equation of quantum mechanics applied to collapse but with a difference depending on the operator ordering \[1-5\]. The usual Schrödinger equation as $$H\psi =i\frac{\partial \psi}{\partial where | y^2}-\frac{q}{y}\
\frac{\partial}{\partial Y}\nonumber \eNd{aliGneD}$$ whErE $(p,q)$ aRe $c$-nUmbers.
It is assuMEd thAt the Hamiltonian (2) is a conStraiNt FOr a cLAsSical hamiltoNAiN WIth ThE mAss TeRM pResenT foR the scaLar field $\phI$. In OtHer words, the cANoNical quantIzaTion needs the AnnIhilatIoN of THe wavE fuNctioN $\psi$ by THe corrEsponding QuANtum opERator $$\haT{h}\PsI =0$$ thaT results in the WheeLEr-dE Witt equation. USing eq. (2-5) We GEt $$\FRAc{\LAmbDa}{2R^2}\left(\fraC{\pArtiaL^2\Psi}{\partIAl \lAMBda^2}+
\FRac{p}{\Lambda}\ \fraC{\partial \psi}{\PArtIal \LamBdA}\riGHt)+
\frac{1}{2r^2\LambDa}\LEft(\Frac{\partial^2\lambDa}{\partial \Phi^2}+
\fraC{Q}{\phi}\fraC{\Partial \Psi}{\parTiaL \phI}\rigHT)-
\fRaC{1}{R}\ \fRaC{\ParTIaL^2\psI}{\ParTial R\parTiAl \lambdA}\equIV
v \PSi$$ whEre $\Psi$ iS a funCtional of $\LambDa$, $\pHi$ anD $r$ fuNctioNs, and $v$ is a PoTentiAl term WrittEn As $$V=\frac{R}{\Lambda}\ R''-\Frac{r}{\Lambda^2}\ R'\LAmbDa '+\FraC{1}{2\LAmbda}\
r{'}^2-\Frac{1}{2}\ \LaMbdA +\frAc{1}{2}\ \frac{R^2}{\lambda}\ \pHI{'}^2 +
\frAc{1}{2}\ M^2r^2\lAmBda \phi^2$$ The prime meanS dERIvAtive witH respeCT tO tHE coordinAtE $\unDerlINE{r}$. ObsErve THaT in equatIon (6) we dON’t HaVe any deRiVative WiTh rEspEct to TIme. THis meaNs that thE equaTIon (6) could be descRIbing a sphericALlY SYmMEtriC grAvitational CollAPse bUt wiTHoUt aNY explIcit tImE DePEndence for functionaL $\pSi$. The cOncepT of “time” in this Case may appEAR Only afteR quaNTiZAtion in accordaNce wiTh [@tres].
This SUggests tHat eq. (6) Is like thE usual SchRÖDinger eqUatIon Of qUanTUM mEchanics appliED To grAvItationAl cOllapse But WitH a dIffErEnce depenDing on thE oPeRaToR orDerinG \[1-5\].
the usual scHröDiNgeR equaTIon is wRitteN as $$H\PsI =i\FRac{\Partial \PSi}{\PARtiaL t}$$ WhEre | y^2}-\frac{q}{y}\
\frac{ \partial}{ \part ial y} \n onum ber\end{aligned}$ $ whe re $(p,q)$ are $c$-num bers.
I t is as sumed that t h eH a mil to ni an(2 ) i s a c ons traintfor a clas sic al Hamiltonain wi th the mas s t erm presentfor the s ca lar field $\ phi$. In ot h er wor ds, the c an o nicalq uantiza t i on nee ds the annihilati o no f the wave fun ction$\ p si $ bythe correspon di ng qu a ntum op e ra t o r $$ \ hat{H}\psi =0 $$ that res u lts in th eWhe e ler-De Witt e q uat ion. Usingeq.(2-5) weget $$ \ frac{\L a mbda}{2 R^2}\l eft (\f rac{ \ pa rt ial ^2 \ psi } {\ par t ial \Lambda ^2 }+
\fra c{p} { \ L a mbda }\\fra c{\pa rtial \psi}{\ par tial \La mbda} \righ t)+\f rac{1 }{2R^2 \Lamb da }\left(\frac{\p arti al^2\Lamb da} {\ par ti al \p h i^2}+\fr ac{ q}{\phi }\frac{ \ par ti a l \p si}{\partial \phi} \r i g ht )-
\frac {1}{R} \ \ fr a c{\parti al ^2\ psi} { \ parti al R \ pa rtial \L ambda} \ eq ui v
V \ps i$ $ wher e$\p si$ is a func tional of $\La mbda$ , $\phi$ and $R $ functions, a n d$ V $i s apot ential term wri t tenas $ $ V= \fr a c{R}{ \Lamb da } \R ''-\frac{R}{\Lambda ^2 }\ R'\ Lambd a '+\frac{1}{ 2\Lambda}\ R { '}^2-\fr ac{1 } {2 } \ \Lambda +\fr ac{1} {2}\ \frac { R^2}{\La mbda} \ \phi{' }^2 +
\fr a c {1}{2}\m^2 R^2 \La mbd a \p hi^2$$ The pr i m e me an s deriv ati ve with re spe cttoth e coordin ate $\un de rl in e{ r}$ . Obs e rve that i n e qu ati on (6 ) we do n’t h avean yd eri vativew it h resp ec tto t ime .Thismean s th at theequation(6) coul dbe descri bing a spheri ca lly symmet ri c g ravita t i onal col lapse but without any e x plicittim e dep ende nce for f unc tional $\ p si$. T he con ceptof “t i m e” in t hi s c as e may appe a r on ly af te r qu antizat ion in accordancew ith [@tres].
Th issugg e s ts th a te q.(6 ) is l ike the usual S chrödinger e q ua tion of qu a ntu mmechani cs appl ied t o gravit ational c ollapse b ut wit h a d ifferencedependin g on theo perat o rorder ing \[1-5 \] .
The u sual S c hrö dinge r equa ti on iswritt en as $$H\ psi =i\frac{\partial \p si}{\p artia l t }$$ where | y^2}-\frac{q}{y}\_
\frac{\partial}{\partial y}\nonumber_\end{aligned}$$ where $(p,q)$ are_$c$-numbers.
It is_assumed_that the_Hamiltonian_(2) is a_constraint for a_classical Hamiltonain with the_mass term present_for_the scalar field $\phi$. In other words, the canonical quantization needs the annihilation of_the_wave function_$\psi$_by_the corresponding quantum operator $$\hat{H}\psi_=0$$ that results in the_Wheeler-De Witt_equation. Using eq. (2-5) we get $$\frac{\Lambda}{2R^2}\left(\frac{\partial^2\psi}{\partial \Lambda^2}+
\frac{p}{\Lambda}\_\frac{\partial_\psi}{\partial \Lambda}\right)+
\frac{1}{2R^2\Lambda}\left(\frac{\partial^2\Lambda}{\partial \phi^2}+
\frac{q}{\phi}\frac{\partial_\psi}{\partial \phi}\right)-
\frac{1}{R}\ \frac{\partial^2\psi}{\partial R\partial \Lambda}\equiv
V \psi$$ where $\psi$ is_a functional of $\Lambda$, $\phi$ and_$R$ functions, and_$V$_is_a potential term written_as $$V=\frac{R}{\Lambda}\ R''-\frac{R}{\Lambda^2}\ R'\Lambda '+\frac{1}{2\Lambda}\
R{'}^2-\frac{1}{2}\ \Lambda_+\frac{1}{2}\ \frac{R^2}{\Lambda}\ \phi{'}^2 +
\frac{1}{2}\ m^2R^2\Lambda \phi^2$$_The prime means derivative with respect to_the coordinate $\underline{r}$. Observe that in_equation (6) we don’t have_any derivative_with respect to time. This_means that the_equation (6)_could be describing_a spherically symmetric gravitational collapse but_without any explicit_time dependence for functional $\psi$. The_concept_of “time” in_this_case_may appear_only after quantization_in_accordance with_[@tres].
This_suggests that eq. (6) is like_the_usual Schrödinger equation of quantum mechanics applied_to gravitational collapse but_with_a difference depending on_the operator ordering \[1-5\].
The usual_Schrödinger equation is written as $$H\psi_=i\frac{\partial \psi}{\partial_t}$$ where |
n/4}}
\exp(-\sigma y^{1+(n/2)})&for $m=1$,\cr
y^{ -{(n+2)}/{(m-1)} }&for $m> 1$,\cr}
\label{aprof}$$ as $y\to \infty$, where $\sigma$ is a constant. Similar results hold for $B_{ss}$ by interchanging $m$ and $n$. Within this approach, it is also possible to show[@codr] that the ‘mean-field’ assumption $R=ka^mb^n$ (assumed in all the above equations) is valid for microscopic stochastic systems in spatial dimension $d>d_c\equiv 2/(m+n-1)$.
For the time-dependent problem, when $x\gg w$ one of the species is overwhelmingly in the majority, so $|u|\to
{\rm max}(a/m,b/n)$, and the profile of the majority particle density is $\sim
|x-x_f|/t^{1/2}$ for $(Dt)^{1/2}\gg x\gg w$. The diffusion current of particles arriving at $x_f$ is therefore $J\sim t^{-{1/2}}$, and the characteristic timescale on which this current changes is $(d\log J/dt)^{-1}\propto t$. The equilibration time of the front is of order $Dw^2$, so since $\alpha<{1/2}$ one would expect that the reaction zone has enough time to reach the steady-state profile it would have if the current $J$ were constant. One would therefore predict that the results of the steady-state problem, and hence the dynamic scaling ansatz, would be applicable to the time dependent case for $x\ll (Dt)^{1/2}$[@codr].
For $m=n=1$ the scaling forms (\[dyscale1\]–\[dyscale3\]) have been proved rigorously to describe the asymptotic behaviour as $t\to \infty$ of the reaction-diffusion equations (\[rateeq1\],\[rateeq2\])[@scstwi]. Experiments on real systems, and simulations of microscopic stochastic models, also appear to verify the scaling theory and exponents in dimension $d\ge 2$[@jieb; @chdr; @koliko]. For $d=1$, there has | n/4 } }
\exp(-\sigma y^{1+(n/2)})&for $ m=1$,\cr
y^ { -{(n+2)}/{(m-1) } } & for $ m > 1$,\cr }
\label{aprof}$$ as $ y\to \infty$, where $ \sigma$ is a constant. Similar results hold for $ B_{ss}$ by interchanging $ m$ and $ n$. Within this overture, it is besides possible to show[@codr ] that the ‘ mean - sphere ’ presumption $ R = ka^mb^n$ (assumed in all the above equation) is valid for microscopic stochastic systems in spatial dimension $ d > d_c\equiv 2/(m+n-1)$.
For the clock time - dependent problem, when $ x\gg w$ one of the coinage is overwhelmingly in the majority, so $ |u|\to
{ \rm max}(a / m, b / n)$, and the profile of the majority particle density is $ \sim
|x - x_f|/t^{1/2}$ for $ (Dt)^{1/2}\gg x\gg w$. The dissemination current of particles arriving at $ x_f$ is therefore $ J\sim t^{-{1/2}}$, and the characteristic timescale on which this current changes is $ (d\log J / dt)^{-1}\propto t$. The equilibration clock time of the front is of order $ Dw^2 $, so since $ \alpha<{1/2}$ one would expect that the chemical reaction zone has enough clock time to reach the firm - state profile it would have if the current $ J$ were constant. One would therefore bode that the results of the steady - state problem, and hence the dynamic scaling ansatz, would be applicable to the meter dependent case for $ x\ll (Dt)^{1/2}$[@codr ].
For $ m = n=1 $ the scaling forms (\[dyscale1\]–\[dyscale3\ ]) have been proved rigorously to describe the asymptotic behaviour as $ t\to \infty$ of the reaction - diffusion equations (\[rateeq1\],\[rateeq2\])[@scstwi ]. experiment on real systems, and pretense of microscopic stochastic model, also appear to verify the scaling hypothesis and exponents in property $ d\ge 2$[@jieb; @chdr; @koliko ]. For $ d=1 $, there has | n/4}}
\exo(-\sigma y^{1+(n/2)})&for $m=1$,\cr
y^{ -{(n+2)}/{(m-1)} }&for $m> 1$,\cr}
\label{aprof}$$ as $y\to \infty$, where $\skgma$ is a constant. Similar rxsulrs hood for $B_{ss}$ by interchavging $m$ ajd $n$. Wityin uhis approach, it ma also ijssigpe tm show[@codr] that the ‘mean-fheld’ assumptiot $F=kc^mb^n$ (assumed in all the above equatijns) is fapid for microssopib ftocgastic systems in spatial dimensioh $d>d_c\eqliv 2/(m+n-1)$.
For the timr-dependent problem, when $x\gh w$ lne of the species is overwheomindoy in the magority, so $|l|\co
{\rm max}(a/m,b/h)$, and the profile of the majorigy paxticle densutt id $\sim
|x-x_f|/t^{1/2}$ foc $(Dt)^{1/2}\gg x\gg w$. The dlgfusiot curremt of particlex acrivung at $x_f$ is thereforx $J\sim t^{-{1/2}}$, and the chatacteristiw cimescale on which thus currgnt cvangdw ir $(d\moj J/st)^{-1}\propho v$. The equiljbration tine of the front is pf irder $Dw^2$, so sjnce $\ajpra<{1/2}$ one would expect that the reaction zmne has enough time to reaxh the steady-state prlfile it rould have if the current $J$ were constant. One wound thxrdfoxc oeefict that the results of the steady-state probjsm, akd hence the dyncmic scaling ansstx, ejuld be applizable ck fhe time dependent case fjr $x\lo (Dt)^{1/2}$[@codr].
Fjr $m=m=1$ the scaling forms (\[dyscale1\]–\[eyscale3\]) have veen proved rigorobsly to descxibe tne asumptotic behaviour as $t\co \inffy$ of the rfaction-dirwusion equations (\[raneeq1\],\[sateeq2\])[@scstwi]. Experiments og real sywtemx, and skmulstions of microsfopic stochastic models, wlso cppeas to verifj the scaling theory and exponeivs in dimensipn $d\gv 2$[@jieb; @chbr; @kollko]. For $d=1$, there has | n/4}} \exp(-\sigma y^{1+(n/2)})&for $m=1$,\cr y^{ -{(n+2)}/{(m-1)} }&for \label{aprof}$$ $y\to \infty$, $\sigma$ is a $B_{ss}$ interchanging $m$ and Within this approach, is also possible to show[@codr] that ‘mean-field’ assumption $R=ka^mb^n$ (assumed in all the above equations) is valid for microscopic systems in spatial dimension $d>d_c\equiv 2/(m+n-1)$. For the time-dependent problem, when $x\gg w$ of species overwhelmingly the majority, so $|u|\to {\rm max}(a/m,b/n)$, and the profile of the majority particle density is $\sim for $(Dt)^{1/2}\gg x\gg w$. The diffusion current of arriving at $x_f$ is $J\sim t^{-{1/2}}$, and the characteristic on this current is J/dt)^{-1}\propto The equilibration time the front is of order $Dw^2$, so since $\alpha<{1/2}$ one would expect that the reaction zone has time to steady-state profile would if current $J$ were would therefore predict that the results problem, and hence the dynamic scaling ansatz, would applicable to time dependent case for $x\ll (Dt)^{1/2}$[@codr]. $m=n=1$ the scaling forms (\[dyscale1\]–\[dyscale3\]) have been proved to describe the asymptotic behaviour as $t\to \infty$ of the reaction-diffusion equations (\[rateeq1\],\[rateeq2\])[@scstwi]. Experiments on and simulations of microscopic models, also appear verify scaling and in dimension 2$[@jieb; @chdr; @koliko]. For $d=1$, there has | n/4}}
\exp(-\sigma y^{1+(n/2)})&for $m=1$,\cr
y^{ -{(n+2)}/{(m-1)} }&for $m> 1$,\cR}
\label{aproF}$$ as $y\tO \inFty$, WhEre $\sIgma$ Is a constant. SimILar rEsults hold for $B_{ss}$ by inteRchanGiNG $m$ anD $N$. WIthin This appROaCH, It iS aLsO poSsIBlE to shOw[@cOdr] that The ‘mean-fieLd’ aSsUmption $R=ka^mb^N$ (AsSumed in all The Above equatioNs) iS valid FoR miCRoscoPic StochAstic sYStems iN spatial dImENsion $d>D_C\equiv 2/(m+N-1)$.
fOr The tIme-dependent problEM, wHEn $x\gg w$ one of the SpecieS iS OvERWheLmiNgly in the mAjOrity, SO $|u|\to
{\rm mAX}(a/M,B/N)$, And THe profile of thE majority paRTicLe densItY is $\SIm
|x-x_f|/t^{1/2}$ For $(Dt)^{1/2}\Gg X\Gg w$. the diffusioN curRent of parTicles ARriving AT $x_f$ is thEreforE $J\sIm t^{-{1/2}}$, And tHE cHaRacTeRIstIC tImeSCalE on which ThIs CurreNt chANGES is $(d\Log j/dt)^{-1}\pRopto T$. The equilibraTioN timE Of tHe froNt is oF ordEr $dw^2$, so sInce $\alPha<{1/2}$ onE wOuld expect that tHe reAction zonE haS eNouGh Time tO Reach tHe sTeaDy-state Profile IT woUlD HAVe If the current $J$ were cOnSTAnT. One woulD thereFOrE pREdict thaT tHe rEsulTS Of the SteaDY-sTate probLem, and HEnCe The dynaMiC scaliNg AnsAtz, Would BE appLicablE to the tiMe depENdent case for $x\lL (dt)^{1/2}$[@codr].
For $m=n=1$ thE ScALInG FormS (\[dyScale1\]–\[dyscalE3\]) havE Been ProvED rIgoROusly To desCrIBe THe asymptotic behavioUr As $t\to \iNfty$ oF the reaction-dIffusion eqUATIons (\[rateEq1\],\[raTEeQ2\])[@Scstwi]. ExperimeNts on Real systemS, And simulAtionS of microScopic stoCHAstic modEls, AlsO apPeaR TO vErify the scaliNG TheoRy And expoNenTs in dimEnsIon $D\ge 2$[@JieB; @cHdr; @koliko]. for $d=1$, therE hAs | n/4}}
\exp(-\sigma y^{1+(n /2)})&for$m=1$ ,\c r
y ^{ -{( n+2) }/{(m-1)} }&fo r $m> 1$,\cr}
\label{aprof} $$ as $ y \to\ in fty$, where$ \s i g ma$ i sa c on s ta nt. S imi lar res ults holdfor $ B_{ss}$ by i n te rchanging$m$ and $n$. Wi thi n this a ppr o ach,itis al so pos s ible t o show[@c od r ] that the ‘me a n -f ield ’ assumption $R=k a ^m b ^n$ (assumed i n allth e a b o veequ ations) is v alidf or micr o sc o p i c s t ochastic syst ems in spat i aldimens io n $ d >d_c\e quiv2/ ( m+n -1)$.
Forthetime-depe ndentp roblem, when $x \gg w$ on e o f th e s pe cie si s o v er whe l min gly in t he m ajori ty,s o $ |u|\ to{\rm max} (a/m,b/n)$, a ndthep rof ile o f the maj or ity p articl e den si ty is $\sim
|x- x_f| /t^{1/2}$ fo r$(D t) ^{1/2 } \gg x\ ggw$. The di ffusion cur re n t of particles arrivin ga t $ x_f$ istheref o re $ J \sim t^{ -{ 1/2 }}$, a nd th e ch a ra cteristi c time s ca le on whi ch thiscu rre ntchang e s is $(d\l og J/dt) ^{-1} \ propto t$. The equilibration ti m e o f the fr ont is of o rder $Dw^ 2$,s osin c e $\a lpha< {1 / 2} $ one would expect t ha t thereact ion zone hasenough tim e t o reachthes te a dy-state profi le it would hav e if thecurre nt $J$ w ere const a n t. One w oul d t her efo r e p redict that t h e res ul ts of t hesteady- sta tepro ble m, and henc e the dy na mi csc ali ng an s atz, wou ld be a ppl icabl e to th e tim e de pe nd e ntcase fo r $ x \ ll ( Dt )^ {1/2 }$[ @c odr].
Fo r $m =n=1$ t he scalin g f o rms(\ [d yscale1 \]–\[dyscale3 \] ) have bee npro ved ri g o rously t o describe the asymptot i c behav iou r as$t\t o \infty$ of the r eac t ion-di ffusio n equ at ion s (\[ra t e eq 1\] ,\ [rateeq2\] ) [ @sc stwi] .Expe riments on real systems,a ndsimulations o f m icro s c op ics to c has ti c mo d e ls, also appear to verify t h escaling th e ory a nd expo nents i n dim e nsion $ d\ge 2$[@ jieb; @ch dr ; @k o l iko ]. For $d= 1$, ther e has | n/4}}
\exp(-\sigma y^{1+(n/2)})&for_$m=1$,\cr
y^{ -{(n+2)}/{(m-1)}_}&for $m> 1$,\cr}
\label{aprof}$$ as_$y\to \infty$,_where_$\sigma$ is_a_constant. Similar results_hold for $B_{ss}$_by interchanging $m$ and_$n$. Within this_approach,_it is also possible to show[@codr] that the ‘mean-field’ assumption $R=ka^mb^n$ (assumed in all_the_above equations)_is_valid_for microscopic stochastic systems in_spatial dimension $d>d_c\equiv 2/(m+n-1)$.
For the_time-dependent problem,_when $x\gg w$ one of the species is_overwhelmingly_in the majority,_so $|u|\to
{\rm max}(a/m,b/n)$, and the profile of the majority_particle density is $\sim
|x-x_f|/t^{1/2}$ for $(Dt)^{1/2}\gg_x\gg w$. The_diffusion_current_of particles arriving at_$x_f$ is therefore $J\sim t^{-{1/2}}$, and_the characteristic timescale on which this_current changes is $(d\log J/dt)^{-1}\propto t$. The_equilibration time of the front is_of order $Dw^2$, so since_$\alpha<{1/2}$ one_would expect that the reaction_zone has enough_time to_reach the steady-state_profile it would have if the_current $J$ were_constant. One would therefore predict that_the_results of the_steady-state_problem,_and hence_the dynamic scaling_ansatz,_would be_applicable_to the time dependent case for_$x\ll_(Dt)^{1/2}$[@codr].
For $m=n=1$ the scaling forms (\[dyscale1\]–\[dyscale3\]) have_been proved rigorously to_describe_the asymptotic behaviour as_$t\to \infty$ of the reaction-diffusion_equations (\[rateeq1\],\[rateeq2\])[@scstwi]. Experiments on real systems,_and simulations_of microscopic_stochastic models, also appear to verify the scaling theory and exponents_in dimension $d\ge 2$[@jieb; @chdr; @koliko]._For $d=1$, there has |
inequality in our settings.
\[lem:videnskii arcs\] Let $A\subset {{\mathbb{R}}}^2$ be an arc of length $l$ of a circle of radius $r$, $l<2\pi r$, ${{\boldsymbol{x}}}\in A$, $\eta<l$, $r<\eta^{-1}$, $\eta>0$, $d$ be the distance from ${{\boldsymbol{x}}}$ to the 2-point set consisting of the endpoints of $A$, $f\in{{\mathcal{P}}}_n$, $f({{\boldsymbol{x}}})=1$. Then $$\label{eqn:value control by Videnskii}
f({{\boldsymbol{y}}})\ge \frac{1}{2} \quad\text{whenever}\quad
{{\boldsymbol{y}}}\in A\quad\text{and}\quad |{{\boldsymbol{x}}}-{{\boldsymbol{y}}}|\le c(\eta) \|f\|_{L_\infty(A)}^{-1} \rho_n^*(d).$$
We can parametrize $A$ so that $$A={{\boldsymbol{\varphi}}}([-\beta,\beta]), \quad\text{where}\quad {{\boldsymbol{\varphi}}}(t)={{\boldsymbol{o}}}+r(\cos(t-t_0),\sin(t-t_0)),$$ ${{\boldsymbol{o}}}\in{{\mathbb{R}}}^2$ and $\beta=\tfrac{l}{2r}$. Note that $T_n(t):=f({{\boldsymbol{\varphi}}}(t))$ is a trigonometric polynomial of degree $\le n$. We can assume that ${{\boldsymbol{x}}}$ is closer to ${{\boldsymbol{\varphi}}}(\beta)$ than to ${{\boldsymbol{\varphi}}}(-\beta)$, then ${{\boldsymbol{x}}}={{\boldsymbol{\varphi}}}(\beta-d')$, where $d'=2\arcsin(\tfrac{d}{2r}) \approx \tfrac{d}{r}$ and $\beta-d'\ge0$. With ${{\boldsymbol{y}}}={{\boldsymbol{\varphi}}}(t)$, we will show that $$\label{eqn:d'}
f({{\boldsymbol{y}}})=T_n(t)\ge\tfrac{1}{2} \quad\text{whenever}\quad
t\in[-\beta,\beta] \quad\text{and}\quad
|\beta-d'-t|\le \gamma \|f\|_{L_\infty(A | inequality in our settings.
\[lem: videnskii arcs\ ] Let $ A\subset { { \mathbb{R}}}^2 $ be an arc of length $ l$ of a traffic circle of radius $ r$, $ l<2\pi r$, $ { { \boldsymbol{x}}}\in A$, $ \eta < l$, $ r<\eta^{-1}$, $ \eta>0 $, $ d$ be the distance from $ { { \boldsymbol{x}}}$ to the 2 - item set consisting of the endpoints of $ A$, $ f\in{{\mathcal{P}}}_n$, $ f({{\boldsymbol{x}}})=1$. Then $ $ \label{eqn: value dominance by Videnskii }
f({{\boldsymbol{y}}})\ge \frac{1}{2 } \quad\text{whenever}\quad
{ { \boldsymbol{y}}}\in A\quad\text{and}\quad |{{\boldsymbol{x}}}-{{\boldsymbol{y}}}|\le c(\eta) \|f\|_{L_\infty(A)}^{-1 } \rho_n^*(d).$$
We can parametrize $ A$ so that $ $ A={{\boldsymbol{\varphi}}}([-\beta,\beta ]), \quad\text{where}\quad { { \boldsymbol{\varphi}}}(t)={{\boldsymbol{o}}}+r(\cos(t - t_0),\sin(t - t_0)),$$ $ { { \boldsymbol{o}}}\in{{\mathbb{R}}}^2 $ and $ \beta=\tfrac{l}{2r}$. Note that $ T_n(t):=f({{\boldsymbol{\varphi}}}(t))$ is a trigonometric polynomial of degree $ \le n$. We can assume that $ { { \boldsymbol{x}}}$ is close to $ { { \boldsymbol{\varphi}}}(\beta)$ than to $ { { \boldsymbol{\varphi}}}(-\beta)$, then $ { { \boldsymbol{x}}}={{\boldsymbol{\varphi}}}(\beta - d')$, where $ d'=2\arcsin(\tfrac{d}{2r }) \approx \tfrac{d}{r}$ and $ \beta - d'\ge0$. With $ { { \boldsymbol{y}}}={{\boldsymbol{\varphi}}}(t)$, we will express that $ $ \label{eqn: d' }
f({{\boldsymbol{y}}})=T_n(t)\ge\tfrac{1}{2 } \quad\text{whenever}\quad
t\in[-\beta,\beta ] \quad\text{and}\quad
|\beta - d'-t|\le \gamma \|f\|_{L_\infty(A | infquality in our settings.
\[uem:videnskii arew\] Let $E\subset {{\mathbb{R}}}^2$ be an arc of length $l$ of a rircoe of radius $r$, $l<2\pi r$, ${{\boldsyobol{x}}}\in A$, $\eta<l$, $r<\era^{-1}$, $\eua>0$, $d$ be the distaide from ${{\boldsglbol{r}}}$ vo the 2-point sej consisting of the endpoittr lf $A$, $f\in{{\mathcal{P}}}_n$, $f({{\boldsymbol{x}}})=1$. Then $$\label{ean:galue control fy Vpdqnskjp}
f({{\boldsymbol{y}}})\ge \frac{1}{2} \quad\test{whenener}\quad
{{\boldsykbol{y}}}\in A\quad\text{and}\quad |{{\blldsjmbol{x}}}-{{\boldsymbol{y}}}|\lf c(\eta) \|f\|_{L_\indty(A)}^{-1} \eho_n^*(d).$$
We can oarametrizt $C$ so that $$A={{\goldsymbol{\varphi}}}([-\beta,\beta]), \quad\tebt{whexe}\quad {{\boldwynbop{\earphi}}}(t)={{\boldwymboj{o}}}+r(\cos(t-t_0),\sin(t-b_0)),$$ ${{\boldsfmbol{o}}}\im{{\mathbb{R}}}^2$ and $\bcta=\tfcac{l}{2e}$. Note that $T_n(t):=f({{\boldsbmbol{\varphi}}}(t))$ is a trygonometrhc polynomial of detrwe $\le n$. Wa cav asrumt tiat ${{\boldsjmbkl{x}}}$ is cloaer to ${{\boldwymbol{\varphi}}}(\beta)$ thsn np ${{\boldsymbol{\barphi}}}(-\feea)$, then ${{\boldsymbol{x}}}={{\boldsymbol{\varphi}}}(\beta-g')$, wgere $d'=2\arcsin(\tfrac{d}{2r}) \appeox \tfrac{d}{r}$ and $\beta-d'\he0$. With ${{\bjldsymbol{y}}}={{\boldsymbol{\varphi}}}(t)$, we will show that $$\labal{eqn:v'}
y({{\njuesjmbol{y}}})=T_n(t)\ge\tfrac{1}{2} \quad\text{whenever}\quad
t\in[-\bqfa,\nena] \quad\text{and}\quab
|\beta-d'-t|\le \gakmw \|g\|_{J_\infty(A | inequality in our settings. \[lem:videnskii arcs\] Let be arc of $l$ of a r$, A$, $\eta<l$, $r<\eta^{-1}$, $d$ be the from ${{\boldsymbol{x}}}$ to the 2-point set of the endpoints of $A$, $f\in{{\mathcal{P}}}_n$, $f({{\boldsymbol{x}}})=1$. Then $$\label{eqn:value control by Videnskii} f({{\boldsymbol{y}}})\ge \quad\text{whenever}\quad {{\boldsymbol{y}}}\in A\quad\text{and}\quad |{{\boldsymbol{x}}}-{{\boldsymbol{y}}}|\le c(\eta) \|f\|_{L_\infty(A)}^{-1} \rho_n^*(d).$$ We can parametrize $A$ so that \quad\text{where}\quad ${{\boldsymbol{o}}}\in{{\mathbb{R}}}^2$ $\beta=\tfrac{l}{2r}$. that $T_n(t):=f({{\boldsymbol{\varphi}}}(t))$ is a trigonometric polynomial of degree $\le n$. We can assume that ${{\boldsymbol{x}}}$ is to ${{\boldsymbol{\varphi}}}(\beta)$ than to ${{\boldsymbol{\varphi}}}(-\beta)$, then ${{\boldsymbol{x}}}={{\boldsymbol{\varphi}}}(\beta-d')$, where \approx \tfrac{d}{r}$ and $\beta-d'\ge0$. ${{\boldsymbol{y}}}={{\boldsymbol{\varphi}}}(t)$, we will show that f({{\boldsymbol{y}}})=T_n(t)\ge\tfrac{1}{2} t\in[-\beta,\beta] \quad\text{and}\quad \gamma | inequality in our settings.
\[leM:videnskii Arcs\] LEt $A\SubSeT {{\matHbb{R}}}^2$ Be an arc of lengtH $L$ of a Circle of radius $r$, $l<2\pi r$, ${{\bolDsymbOl{X}}}\In A$, $\eTA<l$, $R<\eta^{-1}$, $\eTa>0$, $d$ be thE DiSTAncE fRoM ${{\boLdSYmBol{x}}}$ tO thE 2-point sEt consistiNg oF tHe endpoints oF $a$, $f\In{{\mathcal{P}}}_N$, $f({{\bOldsymbol{x}}})=1$. ThEn $$\lAbel{eqN:vAluE ContrOl bY VideNskii}
f({{\BOldsymBol{y}}})\ge \fraC{1}{2} \qUAd\text{WHenever}\QUAd
{{\BoldSymbol{y}}}\in A\quad\texT{AnD}\Quad |{{\boldsymbol{X}}}-{{\boldsYmBOl{Y}}}|\LE c(\eTa) \|f\|_{l_\infty(A)}^{-1} \rho_N^*(d).$$
we can PArametrIZe $a$ SO ThaT $$a={{\boldsymbol{\vaRphi}}}([-\beta,\betA]), \QuaD\text{wHeRe}\qUAd {{\boldSymboL{\vARphI}}}(t)={{\boldsymboL{o}}}+r(\cOs(t-t_0),\sin(t-t_0)),$$ ${{\BoldsyMBol{o}}}\in{{\mAThbb{R}}}^2$ anD $\beta=\tFraC{l}{2r}$. note THaT $T_N(t):=f({{\BoLDsyMBoL{\vaRPhi}}}(T))$ is a trigOnOmEtric PolyNOMIAl of DegRee $\lE n$. We cAn assume that ${{\bOldSymbOL{x}}}$ iS closEr to ${{\bOldsYmBol{\vaRphi}}}(\beTa)$ thaN tO ${{\boldsymbol{\varpHi}}}(-\beTa)$, then ${{\bolDsyMbOl{x}}}={{\BoLdsymBOl{\varpHi}}}(\bEta-D')$, where $d'=2\Arcsin(\tFRac{D}{2r}) \APPRoX \tfrac{d}{r}$ and $\beta-d'\ge0$. wiTH ${{\BoLdsymbol{Y}}}={{\boldsYMbOl{\VArphi}}}(t)$, we WiLl sHow tHAT $$\labeL{eqn:D'}
F({{\bOldsymboL{y}}})=T_n(t)\gE\TfRaC{1}{2} \quad\teXt{WhenevEr}\QuaD
t\iN[-\beta,\BEta] \qUad\texT{and}\quad
|\Beta-d'-T|\Le \gamma \|f\|_{L_\infty(a | inequality in our setting s.
\[lem: viden ski i a rc s\]Let$A\subset {{\m a thbb {R}}}^2$ be an arc oflengt h$ l$ o f a circ le of r a di u s $r $, $ l<2 \p i r $, ${ {\b oldsymb ol{x}}}\in A$ ,$\eta<l$, $r < \e ta^{-1}$,$\e ta>0$, $d$ b e t he dis ta nce from${{ \bold symbol { x}}}$to the 2- po i nt set consist i n gof t he endpoints of $ A $, $f\in{{\mathca l{P}}} _n $ ,$ f ({{ \bo ldsymbol{x }} })=1$ . Then $ $ \l a b e l{e q n:value contr ol by Viden s kii }
f( {{\ b oldsym bol{y }} } )\g e \frac{1}{ 2} \ quad\text {whene v er}\qua d
{{ \bolds ymb ol{ y}}} \ in A \qu ad \ tex t {a nd} \ qua d |{{\bo ld sy mbol{ x}}} - { { \ bold sym bol{ y}}}| \le c(\eta) \ |f\ |_{L _ \in fty(A )}^{- 1} \ rh o_n^* (d).$$
Weca n parametrize $ A$ s o that $$ A={ {\ bol ds ymbol { \varph i}} }([ -\beta, \beta]) , \q ua d \ t ex t{where}\quad {{\b ol d s ym bol{\var phi}}} ( t) ={ { \boldsym bo l{o }}}+ r ( \cos( t-t_ 0 ), \sin(t-t _0)),$ $ $ {{ \boldsy mb ol{o}} }\ in{ {\m athbb { R}}} ^2$ an d $\beta =\tfr a c{l}{2r}$. Not e that $T_n(t) : =f ( { {\ b olds ymb ol{\varphi} }}(t ) )$ i s at ri gon o metri c pol yn o mi a l of degree $\le n$ .We can assu me that ${{\b oldsymbol{ x } } }$ is cl oser to ${{\boldsymbol {\var phi}}}(\be t a)$ than to $ {{\bolds ymbol{\va r p hi}}}(-\ bet a)$ , t hen $ {{ \boldsymbol{x } } }={{ \b oldsymb ol{ \varphi }}} (\b eta -d' )$ , where $ d'=2\arc si n( \t fr ac{ d}{2r } ) \appro x\tf ra c{d }{r}$ and $\ beta- d'\g e0 $. Wit h ${{\b o ld s y mbol {y }} }={{ \bo ld symbo l{\v a rph i}}}(t) $, we wil l s h ow t ha t$$\labe l{eqn:d'}
f ({{\boldsy mb ol{ y}}})= T _ n(t)\ge\ tfrac{1}{2} \quad\text{ w henever }\q uad
t \in[-\bet a,\ beta]\qu a d\text {and}\ quad | \ b eta-d ' - t| \le \ gamma \|f\ | _ {L_ \inft y( A | inequality_in our_settings.
\[lem:videnskii arcs\] Let $A\subset_{{\mathbb{R}}}^2$ be_an_arc of_length_$l$ of a_circle of radius_$r$, $l<2\pi r$, ${{\boldsymbol{x}}}\in_A$, $\eta<l$, $r<\eta^{-1}$,_$\eta>0$,_$d$ be the distance from ${{\boldsymbol{x}}}$ to the 2-point set consisting of the endpoints_of_$A$, $f\in{{\mathcal{P}}}_n$,_$f({{\boldsymbol{x}}})=1$._Then_$$\label{eqn:value control by Videnskii}
_ f({{\boldsymbol{y}}})\ge \frac{1}{2} \quad\text{whenever}\quad
_ _ {{\boldsymbol{y}}}\in A\quad\text{and}\quad |{{\boldsymbol{x}}}-{{\boldsymbol{y}}}|\le c(\eta) \|f\|_{L_\infty(A)}^{-1} \rho_n^*(d).$$
We can_parametrize_$A$ so that_$$A={{\boldsymbol{\varphi}}}([-\beta,\beta]), \quad\text{where}\quad {{\boldsymbol{\varphi}}}(t)={{\boldsymbol{o}}}+r(\cos(t-t_0),\sin(t-t_0)),$$ ${{\boldsymbol{o}}}\in{{\mathbb{R}}}^2$ and $\beta=\tfrac{l}{2r}$. Note that $T_n(t):=f({{\boldsymbol{\varphi}}}(t))$_is a trigonometric polynomial of degree_$\le n$. We_can_assume_that ${{\boldsymbol{x}}}$ is closer_to ${{\boldsymbol{\varphi}}}(\beta)$ than to ${{\boldsymbol{\varphi}}}(-\beta)$, then_${{\boldsymbol{x}}}={{\boldsymbol{\varphi}}}(\beta-d')$, where $d'=2\arcsin(\tfrac{d}{2r}) \approx \tfrac{d}{r}$ and_$\beta-d'\ge0$. With ${{\boldsymbol{y}}}={{\boldsymbol{\varphi}}}(t)$, we will show that_$$\label{eqn:d'}
f({{\boldsymbol{y}}})=T_n(t)\ge\tfrac{1}{2} \quad\text{whenever}\quad
_ t\in[-\beta,\beta] \quad\text{and}\quad
_ _ |\beta-d'-t|\le \gamma \|f\|_{L_\infty(A |
Since the X observations were not accompanied by observations in any other wavelength, we do not discuss such data (e.g.optical, Motch et al. 1983) obtained at other epochs. We also do not discuss time variability of the source.
With these caveats, our best physical model is that of a hot accretion disc within $\sim 100$ gravitational radii surrounded by a cold outer disc, see Figs. 5d, e. The seed photons for thermal Comptonization in the hot disc are supplied by cold clouds within the hot disc. The emission of the hot disc is Compton-reflected by the outer cold disc. The outer disc also emits most of the observed soft X-ray excess. This model is in agreement with the spectra, energy balance, and ionization balance at the surface of the reflecting outer disc. The observed amount of reflection requires that the outer disc is flared.
The hot-disc accretion rate is near the maximum set by advection. Based on the spectral fit of the hot slab model, we find the viscosity parameter of $\alpha\sim 1$ and $M_{\rm X}\ga 3M_\odot$, which mass is in agreement with the dynamical mass determination. The hot disc model, which parameters are independent of $M_{\rm X}$, is also supported by the observed similarity of the spectrum of GX 339–4 to those of Seyfert 1s.
We find that pair production photons in the thermal-Comptonization spectrum is negligible and thus the disc is most likely made of electrons and ions (although more complex models with pairs are possible). Also, synchrotron emission in the hot disc with equipartition magnetic field is negligible as a source of seed photons for Comptonization.
We can rule out models with a cold disc covered by a homogeneous corona or by active regions located on the surface of the disc as violating the energy balance. On the other hand, the energy balance is satisfied if there are active regions at some height above the disc surface. However, this model provides poor fits to the spectral data, the predicted ionization state of the postulated inner cold disc is much higher than that found from the Compton-reflection spectral component, as well as it predicts an $\Omega$-$\Gamma$ correlation opposite to that observed.
ACKNOWLEDGMENTS {#acknowledgments.unnumbered}
===============
This research has been supported in part by the KBN grants 2P03C00511p0 | Since the X observations were not accompanied by observations in any early wavelength, we do not discourse such data (e.g.optical, Motch et al. 1983) obtained at early epochs. We besides do not discuss time unevenness of the source.
With these caveats, our best forcible exemplar is that of a hot accretion disk within $ \sim 100 $ gravitational radii surrounded by a cold out disc, see Figs. 5d, e. The seed photons for thermal Comptonization in the hot disc are supply by cold clouds within the hot disc. The discharge of the hot disc is Compton - reflected by the outer cold phonograph record. The outer disc also emits most of the observed cushy X - ray excess. This model is in agreement with the spectra, energy balance, and ionization balance at the surface of the reflecting outer disc. The observed amount of reflection requires that the outer disc is flare.
The hot - disc accretion pace is near the utmost set by advection. Based on the spectral fit of the blistering slab model, we rule the viscosity parameter of $ \alpha\sim 1 $ and $ M_{\rm X}\ga 3M_\odot$, which mass is in agreement with the dynamic mass determination. The hot disc model, which parameters are independent of $ M_{\rm X}$, is also supported by the observed similarity of the spectrum of GX 339–4 to those of Seyfert 1s.
We find that pair product photons in the thermal - Comptonization spectrum is negligible and therefore the magnetic disk is most likely made of electrons and ion (although more complex models with pairs are possible). Also, synchrotron emission in the hot disc with equipartition magnetic field is negligible as a reservoir of seed photon for Comptonization.
We can rule out models with a cold disc covered by a homogeneous corona or by active region located on the open of the disc as violate the energy balance. On the early hand, the energy balance is satisfied if there are active regions at some height above the disc surface. However, this model provides poor fit to the spectral data, the bode ionization state of the postulated inside cold disc is much higher than that found from the Compton - reflection spectral component, as well as it predicts an $ \Omega$-$\Gamma$ correlation opposite to that detect.
acknowledgment { # acknowledgments.unnumbered }
= = = = = = = = = = = = = = =
This inquiry has been hold in part by the KBN grants 2P03C00511p0 | Sijce the X observations weve not accompanigd by obvervatjons in xny other wavelength, we do nlt discyss such data (e.g.opticau, Motch en al. 1983) obtauned qt other e'kchs. We also sl noc viscuss time vatiability of the source.
Witv ghzse caveats, our best physical model ys that ov a hot accretyon cysc sptmin $\sim 100$ gravitational radii surdounded by a cold ouyer disc, see Figs. 5d, e. The sfed ohotons for thermap Comptonizqtiog in the hot aisc are sl'plied by ckld clouds within the hot disc. Ghe ekission of tye jmt disc is Romptog-reflected bn the ogter cokd disc. The ouber dmsc qlso emits most of thx observed soft X-ray excess. Tvia model is in agrwenent fith the wpeztrz, xnedgy bapanre, and ioniaation balabce at the surface pf nne reflectinf outew qisc. The observed amount of reflection gequjres that the outer disx is flared.
The hot-disf accretijn rate is near the maximum set by advection. Baseg on vhd skegbral dih of the hot slab model, we find the viscosity lataketer of $\alpha\fim 1$ and $M_{\rk D}\gs 3M_\odot$, which oass is ih agreement with tje dynaiical mass dettrminstion. The hot disc model, whuch parametegs aee independent of $L_{\rm X}$, is alro sopportrd by the observed simiuarify of the soectrum or GX 339–4 to those of Sejferd 1s.
We find that pair produstion phovons nn the tfermsl-Compeonization specbsum is negligible wnd tkus tve disc is most likely made of electrons eid ions (althoogh moge compler modeks with pairs are possible). Clso, synehrotrun emissioh in thx hot disc wyth equipartidlon magnetic field if netliguble as x source of serd photons for Compronization.
We can rmle oot models with a euod disc covered by a rolojeneotv corona or ty aztixr regkons locateb un tne surface of the divc aa violating the enrrny balancg. On the jther hand, thr energy balance id satmsfied if tnerg are active regions at some hejght abovf tme disc surfase. Hiwever, this kodel provides poor fits to the spectrao data, the predicjed ionization state pn the postuleted igner cold disc is much higher than that found nrom the Compton-reflecfion s[ectrwl component, as well as it predicts an $\Omega$-$\Gamma$ correlation opposite to that observed.
ACKNKWLECGMENNS {#cckuowledgientd.nnnumbered}
===============
This rexearch has been supported in pact by the NBU grants 2P03C00511p0 | Since the X observations were not accompanied in other wavelength, do not discuss al. obtained at other We also do discuss time variability of the source. these caveats, our best physical model is that of a hot accretion disc $\sim 100$ gravitational radii surrounded by a cold outer disc, see Figs. 5d, The photons thermal in the hot disc are supplied by cold clouds within the hot disc. The emission of hot disc is Compton-reflected by the outer cold The outer disc also most of the observed soft excess. model is agreement the energy balance, and balance at the surface of the reflecting outer disc. The observed amount of reflection requires that the disc is hot-disc accretion is the set by advection. the spectral fit of the hot find the viscosity parameter of $\alpha\sim 1$ and X}\ga 3M_\odot$, mass is in agreement with the mass determination. The hot disc model, which parameters independent of $M_{\rm X}$, is also supported by the observed similarity of the spectrum of to those of Seyfert We find that production in thermal-Comptonization is negligible thus the disc is most likely made of electrons and ions more complex models with pairs are possible). Also, synchrotron emission hot with equipartition magnetic is negligible as a of photons for Comptonization. We out with covered a corona or by active located on the surface of disc as violating the hand, the energy balance is satisfied if there active regions at some height above the surface. However, this model provides poor fits to the spectral data, the ionization state postulated inner cold disc is much higher than found from the Compton-reflection component, as well as it predicts an $\Omega$-$\Gamma$ correlation to observed. ACKNOWLEDGMENTS =============== This research been supported in by the KBN | Since the X observations were Not accompaNied bY obSerVaTionS in aNy other wavelenGTh, we Do not discuss such data (e.g.OpticAl, mOtch ET aL. 1983) obtaIned at oTHeR EPocHs. we AlsO dO NoT discUss Time varIability of The SoUrce.
With thesE CaVeats, our beSt pHysical model Is tHat of a HoT acCRetioN diSc witHin $\sim 100$ GRavitaTional radIi SUrrounDEd by a coLD OuTer dIsc, see Figs. 5d, e. The seED pHOtons for thermaL ComptOnIZaTIOn iN thE hot disc arE sUppliED by cold CLoUDS WitHIn the hot disc. THe emission oF The Hot disC iS CoMPton-reFlectEd BY thE outer cold dIsc. THe outer diSc also EMits mosT Of the obServed SofT X-rAy exCEsS. THis MoDEl iS In AgrEEmeNt with thE sPeCtra, eNergY BALAnce, And IoniZatioN balance at the SurFace OF thE reflEctinG outEr Disc. THe obseRved aMoUnt of reflection RequIres that tHe oUtEr dIsC is flARed.
The Hot-DisC accretIon rate IS neAr THE MaXimum set by advectioN. BASEd On the speCtral fIT oF tHE hot slab MoDel, We fiND The viScosITy ParameteR of $\alpHA\sIm 1$ And $M_{\rm X}\Ga 3m_\odot$, wHiCh mAss Is in aGReemEnt witH the dynaMical MAss determinatiON. The hot disc moDEl, WHIcH ParaMetErs are indepEndeNT of $M_{\Rm X}$, iS AlSo sUPportEd by tHe OBsERved similarity of the SpEctrum Of GX 339–4 tO those of SeyfeRt 1s.
We find tHAT Pair prodUctiON pHOtons in the therMal-CoMptonizatiON spectruM is neGligible And thus thE DIsc is mosT liKelY maDe oF ELeCtrons and ions (ALThouGh More comPleX models WitH paIrs Are PoSsible). AlsO, synchroTrOn EmIsSioN in thE Hot disc wItH eqUiParTitioN MagnetIc fieLd is NeGlIGibLe as a soURcE OF seeD pHoTons For coMptonIzatIOn.
WE can rulE out modelS wiTH a coLd DiSc coverEd by a homogeneOuS corona or bY aCtiVe regiONS located On the surface of the disc as VIolatinG thE enerGy baLance. On thE otHer hanD, thE Energy BalancE is saTiSfiED If theRE ArE acTiVe regions aT SOme HeighT aBove The disc Surface. However, this MOdeL provides poor FitS to tHE SpEctRAl DAta, ThE PreDICted ionization sTate of the pOsTUlAted inner cOLd dIsC is much Higher tHan thAT found fRom the ComPton-refleCtIon sPECtrAl componenT, as well aS it predicTS an $\OmEGa$-$\gamma$ CorRelatiOn OppOsite To that OBseRved.
AcKNOWLeDgMENTS {#AcknoWlEdgments.Unnumbered}
===============
This research hAs been SuppoRteD in part by The kbN gRants 2P03C00511p0 | Since the X observationswere not a ccomp ani edby obs erva tions in any o t herwavelength, we do notdiscu ss such da ta (e .g.opti c al , Mot ch e t a l. 19 83) o bta ined at other epo chs .We also do n o tdiscuss ti mevariabilityofthe so ur ce.
With th ese c aveats , our b est physi ca l model is that o fa ho t accretion discw it h in $\sim 100$gravit at i on a l ra dii surrounde dby ac old out e rd i s c,s ee Figs. 5d,e. The seed pho tons f or th e rmal C ompto ni z ati on in the h ot d isc are s upplie d by col d clouds withi n t hehotd is c. Th ee mis s io n o f th e hot di sc i s Com pton - r e f lect edby t he ou ter cold disc . T he o u ter disc also emi ts most of th e obs er ved soft X-rayexce ss. Thismod el is i n agr e ementwit h t he spec tra, en e rgy b a l a nc e, and ionizationba l a nc e at the surfa c eof the refl ec tin g ou t e r dis c. T h eobserved amoun t o freflect io n requ ir estha t the oute r disc is flar ed.
T he hot-disc ac c retion rate i s n e a rt he m axi mum set byadve c tion . Ba s ed on the s pectr al fi t of the hot slab mo de l, wefindthe viscosity parameter o f $\alpha \sim 1$ and $M_{\rm X} \ga 3 M_\odot$,w hich mas s isin agree ment with t he dynam ica l m ass de t e rm ination. Theh o t di sc model, wh ich par ame ter s a rein dependent of $M_{ \r mX} $, is also supporte dbyth e o bserv e d simi larit y of t he spe ctrum o f G X 339– 4to tho seof Seyf ert1 s.
We fin d that p air prod uc ti on phot ons in the th er mal-Compto ni zat ion sp e c trum isnegligible and thus the disc is mo st li kely made ofele ctrons an d ions(altho ugh m or e c o m plexm o de lswi th pairsa r e p ossib le ). A lso, sy nchrotron emission inthe hot discwit h eq u i pa rti t io n ma gn e tic f ield is negligi ble as a s ou r ce of seed p h oto ns for Co mptoniz ation .
We ca n rule ou t modelswi th a c old disc cove red by a homogene o us co r on a orbyactive r egi ons l ocated onthe s urface o f thediscas violati ng the energy balance.On the othe r h and, theene r gybalance i s sa tisfied if th ere areact i ve re gion s a t s o me he ight above the di scs u rf ace. Howeve r , thi s mod elp rovide s po or fits to the sp e ctral data, th e pr e d ict edi oniz at ion state of t hepo s t ulated i nn er cold dis c is muc hh igher thanthat f ound fr o m t h e Comp ton- ref lection s pec tr a l compo ne nt , as we ll a sit pre dictsa n $\ O m ega$-$\Gamma$ co rrela t i on op p osi te to t hat obs e rved .
ACKNOWL EDGMENTS {# acknow ledg ments .unnumb er ed}
== === == ========
T his resea rch h as been s uppo rte d in p artb y theKBNgr ant s 2P03C00 5 1 1p 0 | Since_the X observations_were not accompanied by_observations in_any_other wavelength,_we_do not discuss_such data (e.g.optical,_Motch et al. 1983) obtained_at other epochs._We_also do not discuss time variability of the source.
With these caveats, our best physical_model_is that_of_a_hot accretion disc within $\sim_100$ gravitational radii surrounded by_a cold_outer disc, see Figs. 5d, e. The seed photons_for_thermal Comptonization in_the hot disc are supplied by cold clouds within_the hot disc. The emission of_the hot disc_is_Compton-reflected_by the outer cold_disc. The outer disc also emits_most of the observed soft X-ray_excess. This model is in agreement with_the spectra, energy balance, and ionization_balance at the surface of_the reflecting_outer disc. The observed amount_of reflection requires_that the_outer disc is_flared.
The hot-disc accretion rate is near_the maximum set_by advection. Based on the spectral_fit_of the hot_slab_model,_we find_the viscosity parameter_of_$\alpha\sim 1$_and_$M_{\rm X}\ga 3M_\odot$, which mass is_in_agreement with the dynamical mass determination. The_hot disc model, which_parameters_are independent of $M_{\rm_X}$, is also supported by_the observed similarity of the spectrum_of GX_339–4 to_those of Seyfert 1s.
We find that pair production photons in the_thermal-Comptonization spectrum is negligible and thus_the disc is most_likely made_of_electrons and ions_(although_more complex_models with pairs are possible). Also, synchrotron_emission in_the hot disc with equipartition magnetic_field is negligible as_a_source of seed photons for Comptonization.
We_can rule out models with a_cold disc covered by a_homogeneous_corona_or by active regions located_on the surface of the disc_as violating the_energy balance. On the other hand, the_energy_balance is satisfied if there are_active_regions at some height above the_disc_surface._However, this model provides poor_fits to the spectral data, the_predicted ionization state of the postulated inner cold disc_is much higher_than that found from the_Compton-reflection_spectral_component, as well as it predicts an $\Omega$-$\Gamma$ correlation opposite_to that_observed.
ACKNOWLEDGMENTS {#acknowledgments.unnumbered}
===============
This research_has been supported in part by the KBN grants 2P03C00511p0 |
{
Y \ar@{c->}[r] \ar[d] & J^r_{0,{\mathrm{imm}}}({\mathbb R}^d,M) \ar[d]^{g_*} \\
B \ar@{c->}[r] & J^r_0({\mathbb R}^d,P)
}$$
Applying Lemma \[lemma\_equivalence1\] to the diagram $$\xymatrix{
& J^r_{0,{\mathrm{imm}}}({\mathbb R}^d,f^{-1}(A)) \ar[r] \ar[d]^{j_*} & J^r_0({\mathbb R}^d,A) \ar[d] \\
Y \ar@{c->}[r] \ar[d] & J^r_{0,{\mathrm{imm}}}({\mathbb R}^d,M) \ar[r]^-{f_*} \ar[d]^{g_*} & J^r_0({\mathbb R}^d,N) \\
B \ar@{c->}[r] & J^r_0({\mathbb R}^d,P)
}$$ gives an equivalence of (ii) with the transversality of $$(gj)_*:J^r_{0,{\mathrm{imm}}}({\mathbb R}^d,f^{-1}(A))\longrightarrow J^r_0({\mathbb R}^d,P)$$ to $B$. By Lemma \[lemma\_equivalence2\] this is equivalent to (i).
Now that we know how $f$ controls the transversality of a map defined on the preimage $f^{-1}(A)$ of some submanifold, we would like to see that this transversality condition (any of the two equivalent conditions in Lemma \[lemma\_equivalence3\]) is generic. This is indeed the case. We first prove a more general result which at the same time happens to generalize the Thom Transversality Theorem.
This is an application of Lemma \[lemma\_general\_transversality\] from Section \[section\_general\_transversality\]. We have a map $$\alpha:C^\infty(M,N)\rightarrow C^\infty(Y,J^r(D,N))$$ sending $f$ to $f_*|_Y$. This map is continuous for the weak topology on the target and clearly ${\mathfrak X}=\{f\in C^\infty(M,N)\ |\ \alpha(f | {
Y \ar@{c->}[r ] \ar[d ] & J^r_{0,{\mathrm{imm}}}({\mathbb R}^d, M) \ar[d]^{g _ * } \\
B \ar@{c->}[r ] & J^r_0({\mathbb R}^d, P)
} $ $
Applying Lemma \[lemma\_equivalence1\ ] to the diagram $ $ \xymatrix {
& J^r_{0,{\mathrm{imm}}}({\mathbb R}^d, f^{-1}(A) ) \ar[r ] \ar[d]^{j _ * } & J^r_0({\mathbb R}^d, A) \ar[d ] \\
Y \ar@{c->}[r ] \ar[d ] & J^r_{0,{\mathrm{imm}}}({\mathbb R}^d, M) \ar[r]^-{f _ * } \ar[d]^{g _ * } & J^r_0({\mathbb R}^d, N) \\
B \ar@{c->}[r ] & J^r_0({\mathbb R}^d, P)
} $ $ gives an equivalence of (ii) with the transversality of $ $ (gj)_*:J^r_{0,{\mathrm{imm}}}({\mathbb R}^d, f^{-1}(A))\longrightarrow J^r_0({\mathbb R}^d, P)$$ to $ B$. By Lemma \[lemma\_equivalence2\ ] this is equivalent to (iodine).
nowadays that we know how $ f$ controls the transversality of a map define on the preimage $ f^{-1}(A)$ of some submanifold, we would like to see that this transversality circumstance (any of the two equivalent conditions in Lemma \[lemma\_equivalence3\ ]) is generic. This is indeed the case. We first testify a more general result which at the like time happens to generalize the Thom Transversality Theorem.
This is an application of Lemma \[lemma\_general\_transversality\ ] from incision \[section\_general\_transversality\ ]. We have a map $ $ \alpha: C^\infty(M, N)\rightarrow C^\infty(Y, J^r(D, N))$$ sending $ f$ to $ f_*|_Y$. This map is continuous for the weak regional anatomy on the target and clearly $ { \mathfrak X}=\{f\in C^\infty(M, N)\ |\ \alpha(f | {
Y \ar@{c->}[r] \ar[d] & J^r_{0,{\mathrm{imm}}}({\oathbb R}^d,M) \ar[d]^{g_*} \\
B \er@{c->}[r] & J^d_0({\mathbb F}^d,P)
}$$
Applying Lemma \[lemma\_eqnivaoence1\] to the diagram $$\xymatrkx{
& J^r_{0,{\mathgm{imm}}}({\mathvb R}^v,f^{-1}(A)) \ar[r] \ar[d]^{j_*} & J^r_0({\mathbb R}^d,A) \ar[d] \\
N \ar@{c->}[x] \er[d] & J^r_{0,{\mathrm{imk}}}({\mathbb R}^d,K) \ar[r]^-{f_*} \ar[d]^{g_*} & J^s_0({\mxtkbb R}^d,N) \\
B \ar@{c->}[r] & J^r_0({\mathbb R}^d,P)
}$$ gives ag equivslfnce of (ii) witr tht twansbvrwality of $$(gj)_*:J^r_{0,{\mathrm{imm}}}({\mathbb R}^d,f^{-1}(A))\loigrightarrow J^r_0({\kathbb R}^d,P)$$ to $B$. By Lemma \[lelma\_eeuivalence2\] this is equivalent to (y).
Bow that we ynow how $f$ controls thg transversality of a map definea on che preimagg $r^{-1}(A)$ mf some subnanifjld, we would like to see thst this transvcrsalmty xondition (any of the vwo equivalent condijions in Lamja \[lemma\_equivalencw3\]) us geteriw. Thkw ir ihdxed the cwse. We first lrove a morw general result whocr at the same fime hwp[ens to generalize the Thom Transversalpty Fheorem.
This is an applixation of Lemma \[lemma\_ggneral\_tranfversality\] from Section \[section\_general\_transversalidy\]. We favt q map $$\qloha:C^\infty(M,N)\rightarrow C^\infty(Y,J^r(D,N))$$ sending $f$ tj $f_*|_U$. Nhis map is contikuous for the weak tlppjogy on the txrget chd clearly ${\mathfrak D}=\{f\in C^\igfty(M,B)\ |\ \alpha(f | { Y \ar@{c->}[r] \ar[d] & J^r_{0,{\mathrm{imm}}}({\mathbb R}^d,M) B & J^r_0({\mathbb }$$ Applying Lemma & R}^d,f^{-1}(A)) \ar[r] \ar[d]^{j_*} J^r_0({\mathbb R}^d,A) \ar[d] Y \ar@{c->}[r] \ar[d] & J^r_{0,{\mathrm{imm}}}({\mathbb R}^d,M) \ar[d]^{g_*} & J^r_0({\mathbb R}^d,N) \\ B \ar@{c->}[r] & J^r_0({\mathbb R}^d,P) }$$ gives an of (ii) with the transversality of $$(gj)_*:J^r_{0,{\mathrm{imm}}}({\mathbb R}^d,f^{-1}(A))\longrightarrow J^r_0({\mathbb R}^d,P)$$ to $B$. By \[lemma\_equivalence2\] is to Now that we know how $f$ controls the transversality of a map defined on the preimage of some submanifold, we would like to see this transversality condition (any the two equivalent conditions in \[lemma\_equivalence3\]) generic. This indeed case. first prove a general result which at the same time happens to generalize the Thom Transversality Theorem. This is an of Lemma Section \[section\_general\_transversality\]. have map C^\infty(Y,J^r(D,N))$$ sending $f$ This map is continuous for the the target and clearly ${\mathfrak X}=\{f\in C^\infty(M,N)\ |\ | {
Y \ar@{c->}[r] \ar[d] & J^r_{0,{\mathrm{imm}}}({\mathbb r}^d,M) \ar[d]^{g_*} \\
B \ar@{C->}[r] & J^r_0({\mAthBb R}^D,P)
}$$
applYing lemma \[lemma\_equiVAlenCe1\] to the diagram $$\xymatrix{
& j^r_{0,{\matHrM{Imm}}}({\mAThBb R}^d,f^{-1}(a)) \ar[r] \ar[d]^{J_*} & j^r_0({\MAThbB R}^D,A) \Ar[d] \\
y \aR@{C->}[r] \Ar[d] & J^r_{0,{\MatHrm{imm}}}({\mAthbb R}^d,M) \ar[R]^-{f_*} \aR[d]^{G_*} & J^r_0({\mathbb R}^d,N) \\
b \Ar@{C->}[r] & J^r_0({\mathbb r}^d,P)
}$$ Gives an equivAleNce of (iI) wIth THe traNsvErsalIty of $$(gJ)_*:j^r_{0,{\mathRm{imm}}}({\mathBb r}^D,f^{-1}(A))\lonGRightarROW J^R_0({\matHbb R}^d,P)$$ to $B$. By Lemma \[lEMmA\_Equivalence2\] thiS is equIvALeNT To (i).
now That we know HoW $f$ conTRols the TRaNSVErsALity of a map defIned on the prEImaGe $f^{-1}(A)$ of SoMe sUBmanifOld, we WoULd lIke to see thaT thiS transverSality COnditioN (Any of thE two eqUivAleNt coNDiTiOns In lEmmA \[LeMma\_EQuiValence3\]) iS gEnEric. THis iS INDEed tHe cAse. WE firsT prove a more geNerAl reSUlt Which At the Same TiMe hapPens to GenerAlIze the Thom TransVersAlity TheoRem.
thIs iS aN applICation Of LEmmA \[lemma\_gEneral\_tRAnsVeRSALiTy\] from Section \[sectiOn\_GENeRal\_transVersalITy\]. we HAve a map $$\aLpHa:C^\InftY(m,n)\righTarrOW C^\Infty(Y,J^r(d,N))$$ sendINg $F$ tO $f_*|_Y$. This MaP is conTiNuoUs fOr the WEak tOpologY on the taRget aND clearly ${\mathfrAK X}=\{f\in C^\infty(M,N)\ |\ \ALpHA(F | {
Y \ar@{c->}[r] \ar[d ] & J^r_{0 ,{\ma thr m{i mm }}}( {\ma thbb R}^d,M) \ a r[d] ^{g_*} \\
B \ar@{c ->}[r ]& J^r _ 0( {\mat hbb R}^ d ,P ) } $$
A pp l yi ng Le mma \[lemm a\_equival enc e1 \] to the di a gr am $$\xyma tri x{
& J^r_{0, {\m athrm{ im m}} } ({\ma thb b R}^ d,f^{- 1 }(A))\ar[r] \a r[ d ]^{j_* } & J^r_ 0 ( {\ math bb R}^d,A) \ar[d] \\ Y \ar@{c->}[r] \ar[d ]& J ^ r _{0 ,{\ mathrm{imm }} }({\m a thbb R} ^ d, M ) \ar [ r]^-{f_*} \ar [d]^{g_*} & J^r _0({\m at hbb R}^d,N ) \\B\ ar@ {c->}[r] &J^r_ 0({\mathb b R}^d , P)
}$$g ives an equiv ale nce of( ii )wit ht het ra nsv e rsa lity of$$ (g j)_*: J^r_ { 0 , { \mat hrm {imm }}}({ \mathbb R}^d, f^{ -1}( A ))\ longr ighta rrow J ^r_0( {\math bb R} ^d ,P)$$ to $B$. B y Le mma \[lem ma\ _e qui va lence 2 \] thi s i s e quivale nt to ( i ).
N o w th at we know how $f$ c o n tr ols thetransv e rs al i ty of ama p d efin e d on t he p r ei mage $f^ {-1}(A ) $of some s ub manifo ld , w e w ouldl iketo see that th is tr a nsversality co n dition (any o f t h e t w o eq uiv alent condi tion s inLemm a \ [le m ma\_e quiva le n ce 3 \]) is generic. Thi sis ind eed t he case. We f irst prove a more gen eral re s ult which at t he sa me time ha p pens togener alize th e Thom Tr a n sversali tyThe ore m.T hi s is an appli c a tion o f Lemma \[ lemma\_ gen era l\_ tra ns versality \] fromSe ct io n\[s ectio n \_genera l\ _tr an sve rsali t y\]. W e hav e ama p$ $\a lpha:C^ \ in f t y(M, N) \r ight arr ow C^\i nfty ( Y,J ^r(D,N) )$$ sendi ng$ f$ t o$f _*|_Y$. This map isco ntinuous f or th e weak t opologyon the target and clear l y ${\ma thf rak X }=\{ f\in C^\i nft y(M,N) \ | \ \alph a(f | {
_ _Y \ar@{c->}[r] \ar[d] &_J^r_{0,{\mathrm{imm}}}({\mathbb R}^d,M)_\ar[d]^{g_*}_\\
__ B \ar@{c->}[r]_& J^r_0({\mathbb R}^d,P)
_ }$$
Applying_Lemma \[lemma\_equivalence1\] to the_diagram_$$\xymatrix{
& J^r_{0,{\mathrm{imm}}}({\mathbb R}^d,f^{-1}(A)) \ar[r] \ar[d]^{j_*} & J^r_0({\mathbb R}^d,A) \ar[d] \\
Y \ar@{c->}[r] \ar[d] & J^r_{0,{\mathrm{imm}}}({\mathbb_R}^d,M)_\ar[r]^-{f_*} \ar[d]^{g_*}_&_J^r_0({\mathbb_R}^d,N) \\
B \ar@{c->}[r] & J^r_0({\mathbb_R}^d,P)
}$$ gives an equivalence of_(ii) with_the transversality of $$(gj)_*:J^r_{0,{\mathrm{imm}}}({\mathbb R}^d,f^{-1}(A))\longrightarrow J^r_0({\mathbb R}^d,P)$$ to_$B$._By Lemma \[lemma\_equivalence2\] this_is equivalent to (i).
Now that we know how $f$_controls the transversality of a map_defined on the_preimage_$f^{-1}(A)$_of some submanifold, we_would like to see that this_transversality condition (any of the two_equivalent conditions in Lemma \[lemma\_equivalence3\]) is generic. This_is indeed the case. We first_prove a more general result_which at_the same time happens to_generalize the Thom_Transversality Theorem.
This_is an application_of Lemma \[lemma\_general\_transversality\] from Section \[section\_general\_transversality\]. We have_a map $$\alpha:C^\infty(M,N)\rightarrow_C^\infty(Y,J^r(D,N))$$ sending $f$ to $f_*|_Y$. This_map_is continuous for_the_weak_topology on_the target and_clearly_${\mathfrak X}=\{f\in_C^\infty(M,N)\_|\ \alpha(f |
-100Myr. This effect is hardly significant compared to the error bars, however.
In addition, we remind the reader that the results presented in Figs. \[fig:MF\]-\[fig:MFblur\] assume a [*weak tidal field impact*]{} (i.e. $r_h/r_t \lesssim 0.01$). The impact of a tidal field upon a population of CFRgs is tightly related to their mass-radius relation [see @par11 for a thorough discussion]. Let us assume that all CFRgs are exposed to the same tidal field (e.g. they are all located in the same region of a given galaxy). If the mean surface density of CFRgs is constant, more massive objects have a lower [*volume*]{} density and are therefore more vulnerable to tidally-induced mass-losses after gas expulsion. If the tidal field is strong enough, a deficit in high-mass clusters then characterizes the end of violent relaxation (see bottom panel of fig. 3 in @par11 and fig. 4 in @par10), and the cluster mass function becomes [*steeper*]{} than the embedded-cluster mass function. In contrast, CFRgs with a given mean [*volume*]{} density all present the same sensitivity to the external tidal field, thereby preventing a distorsion of the cluster mass function shape through mass-dependent tidally-induced mass-losses.
The nature of the CFRg mass-radius relation remains heavily debated. Based on an analysis of the binary population in young star clusters, @mar12 favour a weak mass-radius relation (i.e. close to a constant radius). In contrast, @fal10 infer a constant-surface-density relation based on the SFE – hence the amount of stellar feedback – required to cleanse a CFRg from its residual star-forming gas. Note that their analysis does not explore the impact of a mass-dependent gas-expulsion time-scale or of a tidal field [see section 4.1 in @par11 for a detailed discussion of these points].
We strongly encourage observers to report – without any preconceived idea – how the cluster mass function evolves with time, and to vary the size of the cluster age ranges over which cluster mass functions are integrated. This also requests a careful assessment of the errors affecting the mass function slope and, therefore, a good control of the errors affecting individual cluster mass estimates.
Returning to virial equilibrium: | -100Myr. This effect is hardly significant compared to the mistake stripe, however.
In addition, we remind the reader that the solution presented in Figs. \[fig: MF\]-\[fig: MFblur\ ] assume a [ * decrepit tidal playing field impact * ] { } (i.e. $ r_h / r_t \lesssim 0.01 $). The impingement of a tidal field upon a population of CFRgs is tightly related to their mass - spoke relation [ see @par11 for a thorough discussion ]. Let us wear that all CFRgs are exposed to the same tidal field (e.g. they are all locate in the same region of a given galax). If the mean surface concentration of CFRgs is constant, more massive object have a lower [ * volume * ] { } density and are therefore more vulnerable to tidally - induce mass - losses after gas expulsion. If the tidal field is solid enough, a deficit in gamey - mass clusters then characterizes the end of violent relaxation (see bottom panel of fig. 3 in @par11 and fig. 4 in @par10), and the bunch batch function become [ * steep * ] { } than the embedded - cluster mass affair. In contrast, CFRgs with a given mean [ * volume * ] { } density all portray the same sensitivity to the external tidal field, thereby preventing a distorsion of the cluster mass function condition through mass - dependent tidally - induced mass - losses.
The nature of the CFRg mass - spoke relation back remains heavily debated. Based on an psychoanalysis of the binary population in young star clusters, @mar12 favour a weak mass - radius relation (i.e. close to a ceaseless radius). In line, @fal10 infer a constant - surface - density relation establish on the SFE – hence the sum of stellar feedback – want to cleanse a CFRg from its residual star - forming gas. Note that their analysis does not research the impact of a mass - dependent flatulence - expulsion time - scale or of a tidal field [ see part 4.1 in @par11 for a detailed discussion of these points ].
We strongly promote observers to report – without any preconceived idea – how the cluster mass function evolves with time, and to deviate the size of the cluster age ranges over which cluster mass functions are integrate. This besides requests a careful assessment of the errors affecting the mass function slope and, therefore, a full control of the errors affecting individual bunch mass estimate.
Returning to virial equilibrium: | -100Myr. This effect is hardly slgnificant compatee to tie errod bars, huwever.
In addition, we remind vhe eeadee that the results prerented in Figs. \[fig:NF\]-\[fij:MFblur\] assume a [*weak tidal fispd ik'act*]{} (i.e. $r_h/r_t \lexssim 0.01$). The impact of a thdxl field upon a population of CFRgs if tightky related to thgir msfs-raspuw relation [see @par11 for a thorkugh divcussion]. Let is assume that all CFRgs age edposed to the same tidal fiele (e.g. rhey are all located in the same rggion of a given galaxy). If the mdan sbrface densutt ov CFRgs is cinstagt, more masslne objewts havr a lower [*voluke*]{} vensuty and are therefore more vulnerable tj tidally-hnbuced mass-losses aftee tas efpulvion. Uf ghe tmdam fielf ia strong ehough, a defucit in high-mass clisevts then charzcterisef the end of violent relaxation (see botuom pznel of fig. 3 in @par11 and dig. 4 in @par10), and the closter mass function becomes [*steeper*]{} than the embedded-clustes masa fuugbion. Un contrast, CFRgs with a given mean [*volume*]{} densyfy akl present the same sensiyigiyi to the exterval tibzl field, thereby pregenting a diwtorsion jf tne cluster mass function shqpe through iqss-dependent tidalpy-induced mcss-losxes.
Thr nature of the CFRg mars-rasius relatiln remaina heavily debated. Baxeg on an analysis of the bigary popuoatipn in yuung star slusters, @mwr12 faymur a weak mass-radlus rglatiot (i.e. close to a constant radius). In contrast, @fal10 infer a cmnsnant-surfaee-denslty relation bafed on the SFE – hence the xmount of atellar feedback – wequired to cnganse a CFRg hrom its wesieual star-fofoing gas. Note yhat theig cnalysis eoes not explore tme imkadt of a mass-depzudtnr gas-expulsion tioe-ssape ow of a tidal xiela [sdr secgion 4.1 in @pav11 fur a detailed discussion of fhese points].
We strpnnly encoutage obsewvers to repott – without any prtconcemved ivea – hpw jhe cluster mass function evolvss with tlme, and to vary the size of the cluster age ranges over which cluster nass functions arg integrated. This alsp requests a caresul assesvment of the errors qffecting the masx function slope and, fherefmre, a good control of the errors affecting individual cluster mass estimates.
Rerurninj eo virial esuilobriuk: | -100Myr. This effect is hardly significant compared error however. In we remind the in \[fig:MF\]-\[fig:MFblur\] assume a tidal field impact*]{} $r_h/r_t \lesssim 0.01$). The impact of tidal field upon a population of CFRgs is tightly related to their mass-radius [see @par11 for a thorough discussion]. Let us assume that all CFRgs are to same field they are all located in the same region of a given galaxy). If the mean surface of CFRgs is constant, more massive objects have lower [*volume*]{} density and therefore more vulnerable to tidally-induced after expulsion. If tidal is enough, a deficit high-mass clusters then characterizes the end of violent relaxation (see bottom panel of fig. 3 in @par11 fig. 4 and the mass becomes than the embedded-cluster In contrast, CFRgs with a given all present the same sensitivity to the external field, thereby a distorsion of the cluster mass shape through mass-dependent tidally-induced mass-losses. The nature of CFRg mass-radius relation remains heavily debated. Based on an analysis of the binary population in clusters, @mar12 favour a mass-radius relation (i.e. to constant In @fal10 infer constant-surface-density relation based on the SFE – hence the amount of feedback – required to cleanse a CFRg from its residual Note their analysis does explore the impact of mass-dependent time-scale or of a [see 4.1 a discussion these points]. We strongly observers to report – without preconceived idea – how with time, and to vary the size of cluster age ranges over which cluster mass are integrated. This also requests a careful assessment of the errors affecting mass function therefore, a good control of the errors affecting cluster mass estimates. Returning virial equilibrium: | -100Myr. This effect is hardly signIficant comPared To tHe eRrOr baRs, hoWever.
In additioN, We reMind the reader that the reSults PrESentED iN Figs. \[Fig:MF\]-\[fiG:mFBLUr\] aSsUmE a [*wEaK TiDal fiEld Impact*]{} (i.E. $r_h/r_t \lesssIm 0.01$). THe Impact of a tidAL fIeld upon a pOpuLation of CFRgS is TightlY rElaTEd to tHeiR mass-Radius RElatioN [see @par11 foR a THorougH DiscussION]. LEt us Assume that all CFRgS ArE Exposed to the saMe tidaL fIElD (E.G. thEy aRe all locatEd In the SAme regiON oF A GIveN Galaxy). If the meAn surface deNSitY of CFRGs Is cONstant, More mAsSIve Objects have A lowEr [*volume*]{} dEnsity ANd are thERefore mOre vulNerAblE to tIDaLlY-inDuCEd mASs-LosSEs aFter gas eXpUlSion. IF the TIDAL fieLd iS strOng enOugh, a deficit iN hiGh-maSS clUsterS then CharAcTerizEs the eNd of vIoLent relaxation (sEe boTtom panel Of fIg. 3 In @pAr11 And fiG. 4 In @par10), aNd tHe cLuster mAss funcTIon BeCOMEs [*Steeper*]{} than the embeDdED-ClUster masS functIOn. in COntrast, CfRGs wIth a GIVen meAn [*voLUmE*]{} density All preSEnT tHe same sEnSitiviTy To tHe eXternAL tidAl fielD, thereby PreveNTing a distorsioN Of the cluster mASs FUNcTIon sHapE through masS-depENdenT tidALlY-inDUced mAss-loSsES.
THE nature of the CFRg masS-rAdius rElatiOn remains heavIly debated. bASEd on an anAlysIS oF The binary populAtion In young staR Clusters, @Mar12 faVour a weaK mass-radiUS Relation (I.e. cLosE to A coNSTaNt radius). In conTRAst, @fAl10 Infer a cOnsTant-surFacE-deNsiTy rElAtion baseD on the SFe – hEnCe ThE amOunt oF Stellar fEeDbaCk – ReqUired TO cleanSe a CFrg frOm ItS ResIdual stAR-fORMing GaS. NOte tHat ThEir anAlysIS doEs not exPlore the iMpaCT of a MaSs-DependeNt gas-expulsioN tIme-scale or Of A tiDal fieLD [See sectiOn 4.1 in @par11 for a detailed discUSsion of TheSe poiNts].
WE strongly EncOurage ObsERvers tO reporT – withOuT anY PReconCEIvEd iDeA – how the cluSTEr mAss fuNcTion Evolves With time, and to vary tHE siZe of the clusteR agE ranGES oVer WHiCH clUsTEr mASS functions are inTegrated. ThIs ALsO requests a CAreFuL assessMent of tHe errORs affecTing the maSs functioN sLope AND, thErefore, a goOd controL of the errORs affECtIng inDivIdual cLuSteR mass EstimaTEs.
REturnIng to vIrIal equIlibrIuM: | -100Myr. This effect is ha rdly signi fican t c omp ar ed t o th e error bars,h owev er.
In addition, we r emind t h e re a de r tha t the r e su l t s p re se nte di nFigs. \[ fig:MF\ ]-\[fig:MF blu r\ ] assume a [ * we ak tidal f iel d impact*]{} (i .e. $r _h /r_ t \les ssi m 0.0 1$). T h e impa ct of a t id a l fiel d upon a p op ulat ion of CFRgs is t i gh t ly related totheirma s s- r a diu s r elation [s ee @par 1 1 for a th o r o ugh discussion].Let us assu m e t hat al lCFR g s areexpos ed tothe same ti dalfield (e. g. the y are al l locate d in t hesam e re g io nofag ive n g ala x y). If theme an surf aced e n s ityofCFRg s isconstant, mor e m assi v e o bject s hav e alo wer [ *volum e*]{} d ensity and arether efore mor e v ul ner ab le to tidall y-i ndu ced mas s-losse s af te r g as expulsion. If the t i d al field i s stro n gen o ugh, a d ef ici t in h igh-m assc lu sters th en cha r ac te rizes t he end o fvio len t rel a xati on (se e bottom pane l of fig. 3 in@ par11 and fig . 4 i n@ par1 0), and the cl uste r mas s fu n ct ion becom es [* st e ep e r*]{} than the embe dd ed-clu stermass function . In contr a s t , CFRgswith ag iven mean [*vo lume* ]{} densit y all pre sentthe same sensitiv i t y to the ex ter nal ti d a lfield, thereb y prev en ting adis torsion of th e c lus te r mass fu nction s ha pe t hr oug h mas s -depende nt ti da lly -indu c ed mas s-los ses.
Th e na ture of th e CFRg m as s-ra diu srelat ionr ema ins hea vily deba ted . Bas ed o n an an alysis of the b inary popu la tio n in y o u ng starclusters, @mar12 favour a weakmas s-rad iusrelation(i. e. clo set o a co nstant radi us ).I n cont r a st , @ fa l10 infera con stant -s urfa ce-dens ity relation based onthe SFE – hen cethea m ou nto fs tel la r fe e d back – required to cleans ea C FRg from i t s r es idual s tar-for mingg as. Not e that th eir analy si s do e s no t explorethe impa ct of a m a ss-de p en dentgas -expul si ontime- scaleo r o f a t idal f ie ld [se e sec ti on 4.1 i n @par11 for a detailed discu ssion of these po int s ].
We stron glyencourageobs erv ers t o r e port– wi t ho uta ny pr econ c eived ide a – ho w th e cluster m a s s fu nctio n e v olveswith time, and to var y the size of t he c l u ste r a g e ra ng es over whichclu st e r mass fu nc tions are i ntegrate d. Thisalso r equest s a car e f ul assess ment of the erro rsaf f ectingth em ass fu ncti on slope and,t here f o re, a good contr ol of t he er r ors affe ct ing ind i vidu al cluster mass estim ates.
Ret urnin g to vi ri al equ ili br ium: | -100Myr. This_effect is_hardly significant compared to_the error_bars,_however.
In addition,_we_remind the reader_that the results_presented in Figs. \[fig:MF\]-\[fig:MFblur\] assume_a [*weak tidal_field_impact*]{} (i.e. $r_h/r_t \lesssim 0.01$). The impact of a tidal field upon a population_of_CFRgs is_tightly_related_to their mass-radius relation [see_@par11 for a thorough discussion]._Let us_assume that all CFRgs are exposed to the_same_tidal field (e.g._they are all located in the same region of_a given galaxy). If the mean_surface density of_CFRgs_is_constant, more massive objects_have a lower [*volume*]{} density and_are therefore more vulnerable to tidally-induced_mass-losses after gas expulsion. If the tidal_field is strong enough, a deficit_in high-mass clusters then characterizes_the end_of violent relaxation (see bottom_panel of fig. 3_in @par11_and fig. 4 in_@par10), and the cluster mass function_becomes [*steeper*]{} than_the embedded-cluster mass function. In contrast,_CFRgs_with a given_mean_[*volume*]{}_density all_present the same_sensitivity_to the_external_tidal field, thereby preventing a distorsion_of_the cluster mass function shape through mass-dependent_tidally-induced mass-losses.
The nature of_the_CFRg mass-radius relation remains_heavily debated. Based on an_analysis of the binary population in_young star_clusters, @mar12_favour a weak mass-radius relation (i.e. close to a constant radius)._In contrast, @fal10 infer a constant-surface-density_relation based on the_SFE –_hence_the amount of_stellar_feedback –_required to cleanse a CFRg from its_residual star-forming_gas. Note that their analysis does_not explore the impact_of_a mass-dependent gas-expulsion time-scale or of_a tidal field [see section 4.1_in @par11 for a detailed_discussion_of_these points].
We strongly encourage observers_to report – without any preconceived_idea – how_the cluster mass function evolves with time,_and_to vary the size of the_cluster_age ranges over which cluster mass_functions_are_integrated. This also requests a_careful assessment of the errors affecting_the mass function slope and, therefore, a good control_of the errors_affecting individual cluster mass estimates.
Returning_to_virial_equilibrium: |
\]) and (\[d24\]), that $$\begin{tabular}{l}
$\mathcal{L} ^{\prime }(t)\leq -\left( \lambda -\delta \right) \left\Vert
u^{\prime }\left( t\right) \right\Vert ^{2}\bigskip $ \\
$\ \ \ \ \ \ \ \ +\left[ \frac{2}{\mu _{\ast }}\left( \widetilde{h}_{0}^{2}(t)+\widetilde{h}_{1}^{2}(t)\right) +2\delta \left\vert \widetilde{h}_{0}(t)+\widetilde{h}_{1}(t)\right\vert -\delta (1-\eta ^{\ast }-\varepsilon
_{1})\right] \left\Vert u\left( t\right) \right\Vert _{1}^{2}\bigskip $ \\
$\ \ \ \ \ \ \ \ -\left[ \frac{1}{4}\mu _{\ast }-\frac{\delta }{\varepsilon
_{1}}\left( \widetilde{\lambda }_{0}^{2}+\widetilde{\lambda }_{1}^{2}\right) \right] \left[ \left\vert u^{\prime }\left( 0,t\right) \right\vert
^{2}+\left\vert u^{\prime }\left( 1,t\right) \right\vert ^{2}\right] $\end{tabular}
\tag{4.26} \label{d26}$$for all $\delta,$ $\varepsilon _{1}>0.$
Let $$0<\varepsilon _{1}<1-\eta ^{\ast }. \tag{4.27} \label{d27}$$
Then, for $\delta $ small enough, with $0<\delta <\lambda $ and if $\widetilde{h}_{0},$ $\widetilde{h}_{1}$ satisfy $$\frac{2}{\mu _{\ast }}\left( \left\Vert \widetilde{h}_{0}\right\Vert
_{L^{\infty }(\mathbb{R}_{+})}^{2}+\left\Vert \widetilde{h}_{1}\right\Vert
_{L^{\infty }(\mathbb{R}_{+})}^{2}\right) +2\delta \left( \left\Vert
\widetilde{h}_{0}\right\Vert _{L^{\infty }(\mathbb{R}_{+})}+\left\Vert
\widetilde{h}_{1}\right\ | \ ]) and (\[d24\ ]), that $ $ \begin{tabular}{l }
$ \mathcal{L } ^{\prime } (t)\leq -\left (\lambda -\delta \right) \left\Vert
u^{\prime } \left (t\right) \right\Vert ^{2}\bigskip $ \\
$ \ \ \ \ \ \ \ \ + \left [ \frac{2}{\mu _ { \ast } } \left (\widetilde{h}_{0}^{2}(t)+\widetilde{h}_{1}^{2}(t)\right) +2\delta \left\vert \widetilde{h}_{0}(t)+\widetilde{h}_{1}(t)\right\vert -\delta (1-\eta ^{\ast } -\varepsilon
_ { 1})\right ] \left\Vert u\left (t\right) \right\Vert _ { 1}^{2}\bigskip $ \\
$ \ \ \ \ \ \ \ \ -\left [ \frac{1}{4}\mu _ { \ast } -\frac{\delta } { \varepsilon
_ { 1}}\left (\widetilde{\lambda } _ { 0}^{2}+\widetilde{\lambda } _ { 1}^{2}\right) \right ] \left [ \left\vert u^{\prime } \left (0,t\right) \right\vert
^{2}+\left\vert u^{\prime } \left (1,t\right) \right\vert ^{2}\right ] $ \end{tabular }
\tag{4.26 } \label{d26}$$for all $ \delta,$ $ \varepsilon _ { 1}>0.$
Let $ $ 0<\varepsilon _ { 1}<1-\eta ^{\ast }. \tag{4.27 } \label{d27}$$
Then, for $ \delta $ small enough, with $ 0<\delta < \lambda $ and if $ \widetilde{h}_{0},$ $ \widetilde{h}_{1}$ satisfy $ $ \frac{2}{\mu _ { \ast } } \left (\left\Vert \widetilde{h}_{0}\right\Vert
_ { L^{\infty } (\mathbb{R}_{+})}^{2}+\left\Vert \widetilde{h}_{1}\right\Vert
_ { L^{\infty } (\mathbb{R}_{+})}^{2}\right) +2\delta \left (\left\Vert
\widetilde{h}_{0}\right\Vert _ { L^{\infty } (\mathbb{R}_{+})}+\left\Vert
\widetilde{h}_{1}\right\ | \]) anf (\[d24\]), that $$\begin{tabular}{l}
$\mauhcal{L} ^{\prime }(t)\leq -\left( \lembda -\dslta \rigft) \left\Vert
u^{\prime }\left( t\righv) \ritht\Veet ^{2}\bigskip $ \\
$\ \ \ \ \ \ \ \ +\lewt[ \frac{2}{\mu _{\ast }}\lefr( \wivetilde{h}_{0}^{2}(t)+\widetilvs{h}_{1}^{2}(t)\righb) +2\delfw \leyt\tert \widetilde{h}_{0}(j)+\widetilde{h}_{1}(t)\sight\vert -\deltd (1-\dtc ^{\ast }-\varepsilon
_{1})\right] \left\Vert u\left( t\right) \rlght\Vert _{1}^{2}\bigskyp $ \\
$\ \ \ \ \ \ \ \ -\left[ \frac{1}{4}\mu _{\ast }-\frac{\delta }{\varepailon
_{1}}\leht( \widetilde{\lamnda }_{0}^{2}+\widetilde{\lambda }_{1}^{2}\right) \gighh] \left[ \left\vert u^{\pgime }\left( 0,t\tjghe) \right\vert
^{2}+\lewt\vert u^{\prpke }\left( 1,t\rjght) \right\vert ^{2}\right] $\end{tabular}
\gag{4.26} \kabel{d26}$$for qlo $\dfnta,$ $\varepsioon _{1}>0.$
Lvt $$0<\varepsilon _{1}<1-\eta ^{\ast }. \tag{4.27} \label{d27}$$
Then, fov $\delva $ wmall enough, with $0<\delva <\lambda $ and if $\wiqetilde{h}_{0},$ $\fibetilde{h}_{1}$ satisfy $$\frac{2}{\nu _{\ast }}\neft( \lefg\Cerg \wjdxtimde{h}_{0}\rihht\Tert
_{L^{\infty }(\jathbb{R}_{+})}^{2}+\left\Cert \widetilde{h}_{1}\righu\Vewn
_{K^{\infty }(\mathbg{R}_{+})}^{2}\righe) +2\qelta \left( \left\Vert
\widetilde{h}_{0}\right\Vert _{N^{\inrty }(\mathbb{R}_{+})}+\left\Vert
\widerilde{h}_{1}\right\ | \]) and (\[d24\]), that $$\begin{tabular}{l} $\mathcal{L} ^{\prime \lambda \right) \left\Vert }\left( t\right) \right\Vert \ \ \ \ +\left[ \frac{2}{\mu _{\ast \widetilde{h}_{0}^{2}(t)+\widetilde{h}_{1}^{2}(t)\right) +2\delta \left\vert \widetilde{h}_{0}(t)+\widetilde{h}_{1}(t)\right\vert -\delta (1-\eta }-\varepsilon _{1})\right] \left\Vert u\left( t\right) \right\Vert _{1}^{2}\bigskip $ \\ $\ \ \ \ \ \ \ -\left[ \frac{1}{4}\mu _{\ast }-\frac{\delta }{\varepsilon _{1}}\left( \widetilde{\lambda }_{0}^{2}+\widetilde{\lambda }_{1}^{2}\right) \right] \left\vert }\left( \right\vert u^{\prime }\left( 1,t\right) \right\vert ^{2}\right] $\end{tabular} \tag{4.26} \label{d26}$$for all $\delta,$ $\varepsilon _{1}>0.$ Let $$0<\varepsilon _{1}<1-\eta ^{\ast \tag{4.27} \label{d27}$$ Then, for $\delta $ small enough, $0<\delta <\lambda $ and $\widetilde{h}_{0},$ $\widetilde{h}_{1}$ satisfy $$\frac{2}{\mu _{\ast \left\Vert _{L^{\infty }(\mathbb{R}_{+})}^{2}+\left\Vert _{L^{\infty +2\delta \left\Vert \widetilde{h}_{0}\right\Vert _{L^{\infty \widetilde{h}_{1}\right\ | \]) and (\[d24\]), that $$\begin{tabular}{l}
$\mathCal{L} ^{\prime }(t)\Leq -\leFt( \lAmbDa -\DeltA \rigHt) \left\Vert
u^{\priME }\lefT( t\right) \right\Vert ^{2}\bigskiP $ \\
$\ \ \ \ \ \ \ \ +\left[ \FrAC{2}{\mu _{\aST }}\lEft( \wiDetilde{H}_{0}^{2}(T)+\wIDEtiLdE{h}_{1}^{2}(T)\riGhT) +2\DeLta \leFt\vErt \wideTilde{h}_{0}(t)+\widEtiLdE{h}_{1}(t)\right\vert -\DElTa (1-\eta ^{\ast }-\vaRepSilon
_{1})\right] \leFt\VErt u\leFt( T\riGHt) \rigHt\VErt _{1}^{2}\biGskip $ \\
$\ \ \ \ \ \ \ \ -\lEFt[ \frac{1}{4}\Mu _{\ast }-\frac{\DeLTa }{\varePSilon
_{1}}\leFT( \WiDetiLde{\lambda }_{0}^{2}+\widetildE{\LaMBda }_{1}^{2}\right) \right] \lEft[ \lefT\vERt U^{\PRimE }\leFt( 0,t\right) \riGhT\vert
^{2}+\LEft\vert U^{\PrIME }\LefT( 1,T\right) \right\veRt ^{2}\right] $\end{tABulAr}
\tag{4.26} \lAbEl{d26}$$FOr all $\dElta,$ $\vArEPsiLon _{1}>0.$
Let $$0<\varepSiloN _{1}<1-\eta ^{\ast }. \taG{4.27} \label{D27}$$
then, for $\DElta $ smaLl enouGh, wIth $0<\DeltA <\LaMbDa $ aNd IF $\wiDEtIldE{H}_{0},$ $\wiDetilde{h}_{1}$ SaTiSfy $$\frAc{2}{\mu _{\AST }}\LEft( \lEft\vert \WidetIlde{h}_{0}\right\VerT
_{L^{\iNfty }(\MAthBb{R}_{+})}^{2}+\leFt\VerT \widEtIlde{h}_{1}\Right\VErt
_{L^{\iNfTy }(\mathbb{R}_{+})}^{2}\right) +2\dElta \Left( \left\VErt
\WiDetIlDe{h}_{0}\riGHt\Vert _{l^{\inFty }(\Mathbb{R}_{+})}+\Left\VerT
\WidEtILDE{h}_{1}\Right\ | \]) and (\[d24\]), that $$ \begin{tab ular} {l}
$\ ma thca l{L} ^{\prime }(t) \ leq-\left( \lambda -\delt a \ri gh t ) \l e ft \Vert
u^{\pr i me } \le ft (t\r ig h t) \rig ht\ Vert ^{ 2}\bigskip $\\
$\ \ \ \ \\ \ \ +\left[ \f rac{2}{\mu _ {\a st }}\ le ft( \wide til de{h} _{0}^{ 2 }(t)+\ widetilde {h } _{1}^{ 2 }(t)\ri g h t) +2\ delta \left\vert\ wi d etilde{h}_{0}( t)+\wi de t il d e {h} _{1 }(t)\right \v ert - \ delta ( 1 -\ e t a ^{ \ ast }-\vareps ilon
_{1})\ r igh t] \le ft \Ve r t u\le ft( t \r i ght ) \right\Ve rt _ {1}^{2}\b igskip $ \\
$\ \ \ \ \ \ \ \ -\ lef t[ \ f ra c{ 1}{ 4} \ mu_ {\ ast }-\ frac{\de lt a}{\va reps i l o n
_{1 }}\ left ( \wi detilde{\lamb da}_{0 } ^{2 }+\wi detil de{\ la mbda}_{1}^ {2}\r ig ht) \right] \le ft[\left\ver t u ^{ \pr im e }\l e ft( 0, t\r igh t) \rig ht\vert ^{2 }+ \ l e ft \vert u^{\prime }\ le f t (1,t\righ t) \ri g ht \v e rt ^{2}\ ri ght ] $\ e n d{tab ular }
\ tag{4.26 } \la b el {d 26}$$fo rall $\ de lta ,$$\var e psil on _{1 }>0.$
L et $$ 0 <\varepsilon _ { 1}<1-\eta ^{\ a st } .\tag {4. 27} \label {d27 } $$
Then , f or$ \delt a $ s ma l le nough, with $0<\del ta <\lam bda $ and if $\wid etilde{h}_ { 0 } ,$ $\wid etil d e{ h }_{1}$ satisfy $$\f rac{2}{\mu _{\ast } }\lef t( \left \Vert \wi d e tilde{h} _{0 }\r igh t\V e r t_{L^{\infty } ( \ math bb {R}_{+} )}^ {2}+\le ft\ Ver t \ wid et ilde{h}_{ 1}\right \V er t_{ L^{ \inft y }(\math bb {R} _{ +}) }^{2} \ right) +2\d elta \ le f t(\left\V e rt \ wide ti ld e{h} _{0 }\ right \Ver t _{ L^{\inf ty }(\mat hbb { R}_{ +} )} +\left\ Vert
\widetil de {h}_{1}\ri gh t\ | \]) and_(\[d24\]), that_$$\begin{tabular}{l}
$\mathcal{L} ^{\prime }(t)\leq -\left(_\lambda -\delta_\right)_\left\Vert
u^{\prime }\left(_t\right)_\right\Vert ^{2}\bigskip $_\\
$\ \ \_\ \ \ \_\ +\left[ \frac{2}{\mu__{\ast_}}\left( \widetilde{h}_{0}^{2}(t)+\widetilde{h}_{1}^{2}(t)\right) +2\delta \left\vert \widetilde{h}_{0}(t)+\widetilde{h}_{1}(t)\right\vert -\delta (1-\eta ^{\ast }-\varepsilon
_{1})\right] \left\Vert u\left( t\right) \right\Vert _{1}^{2}\bigskip_$_\\
$\ \_\_\_\ \ \ \ -\left[_\frac{1}{4}\mu _{\ast }-\frac{\delta }{\varepsilon
_{1}}\left( \widetilde{\lambda_}_{0}^{2}+\widetilde{\lambda }_{1}^{2}\right)_\right] \left[ \left\vert u^{\prime }\left( 0,t\right) \right\vert
^{2}+\left\vert u^{\prime_}\left(_1,t\right) \right\vert ^{2}\right]_$\end{tabular}
\tag{4.26} \label{d26}$$for all $\delta,$ $\varepsilon _{1}>0.$
Let $$0<\varepsilon _{1}<1-\eta_^{\ast }. \tag{4.27} \label{d27}$$
Then,_for $\delta $_small_enough,_with $0<\delta <\lambda $_and if $\widetilde{h}_{0},$ $\widetilde{h}_{1}$ satisfy $$\frac{2}{\mu__{\ast }}\left( \left\Vert \widetilde{h}_{0}\right\Vert
_{L^{\infty }(\mathbb{R}_{+})}^{2}+\left\Vert \widetilde{h}_{1}\right\Vert
_{L^{\infty_}(\mathbb{R}_{+})}^{2}\right) +2\delta \left( \left\Vert
\widetilde{h}_{0}\right\Vert _{L^{\infty }(\mathbb{R}_{+})}+\left\Vert
\widetilde{h}_{1}\right\ |
33,0.0062083)(481.9334859,0.0061949)(482.9610619,0.0061818)(483.9194011,0.0061696)(484.9588041,0.0061565)(485.9289704,0.0061442)(486.9802005,0.006131)(488.012294,0.0061181)(488.9751507,0.0061061)(490.0190713,0.0060932)(490.9937551,0.0060811)(491.9493023,0.0060694)(492.9859133,0.0060567)(493.9532876,0.0060449)(495.0017257,0.0060321)(496.0310272,0.0060197)(496.991092,0.0060081)(498.0322205,0.0059956)(499.0041124,0.005984)(499.0196731,0.0059838)(499.0352338,0.0059836)(499.0663553,0.0059832)(499.1285982,0.0059825)(499.253084,0.005981)(499.5020557,0.005978)(499.5176164,0.0059779)(499.5331771,0.0059777)(499.5642986,0.0059773)(499.6265415,0.0059766)(499.7510273,0.0059751)(499.7665881,0.0059749)(499.7821488,0.0059747)(499.8132703,0.0059743)(499.8755132,0.0059736)(499.8910739,0.0059734)(499.9066346,0.0059732)(499.9377561,0.0059729)(499.9533168,0.0059727)(499.9688775,0.0059725)(499.9844383,0.0059723)(499.999999,0.0059721) ]{}; coordinates [ (0,0.965936) (200,0.965936)]{};\[a=0\] coordinates [ (0,0.379322 | 33,0.0062083)(481.9334859,0.0061949)(482.9610619,0.0061818)(483.9194011,0.0061696)(484.9588041,0.0061565)(485.9289704,0.0061442)(486.9802005,0.006131)(488.012294,0.0061181)(488.9751507,0.0061061)(490.0190713,0.0060932)(490.9937551,0.0060811)(491.9493023,0.0060694)(492.9859133,0.0060567)(493.9532876,0.0060449)(495.0017257,0.0060321)(496.0310272,0.0060197)(496.991092,0.0060081)(498.0322205,0.0059956)(499.0041124,0.005984)(499.0196731,0.0059838)(499.0352338,0.0059836)(499.0663553,0.0059832)(499.1285982,0.0059825)(499.253084,0.005981)(499.5020557,0.005978)(499.5176164,0.0059779)(499.5331771,0.0059777)(499.5642986,0.0059773)(499.6265415,0.0059766)(499.7510273,0.0059751)(499.7665881,0.0059749)(499.7821488,0.0059747)(499.8132703,0.0059743)(499.8755132,0.0059736)(499.8910739,0.0059734)(499.9066346,0.0059732)(499.9377561,0.0059729)(499.9533168,0.0059727)(499.9688775,0.0059725)(499.9844383,0.0059723)(499.999999,0.0059721) ] { }; coordinates [ (0,0.965936) (200,0.965936)]{};\[a=0\ ] coordinates [ (0,0.379322 | 33,0.0062083)(481.9334859,0.0061949)(482.9610619,0.0061818)(483.9194011,0.0061696)(484.9588041,0.0061565)(485.9289704,0.0061442)(486.9802005,0.006131)(488.012294,0.0061181)(488.9751507,0.0061061)(490.0190713,0.0060932)(490.9937551,0.0060811)(491.9493023,0.0060694)(492.9859133,0.0060567)(493.9532876,0.0060449)(495.0017257,0.0060321)(496.0310272,0.0060197)(496.991092,0.0060081)(498.0322205,0.0059956)(499.0041124,0.005984)(499.0196731,0.0059838)(499.0352338,0.0059836)(499.0663553,0.0059832)(499.1285982,0.0059825)(499.253084,0.005981)(499.5020557,0.005978)(499.5176164,0.0059779)(499.5331771,0.0059777)(499.5642986,0.0059773)(499.6265415,0.0059766)(499.7510273,0.0059751)(499.7665881,0.0059749)(499.7821488,0.0059747)(499.8132703,0.0059743)(499.8755132,0.0059736)(499.8910739,0.0059734)(499.9066346,0.0059732)(499.9377561,0.0059729)(499.9533168,0.0059727)(499.9688775,0.0059725)(499.9844383,0.0059723)(499.999999,0.0059721) ]{}; clordinates [ (0,0.965936) (200,0.965936)]{};\[a=0\] coordinatts [ (0,0.379322 | 33,0.0062083)(481.9334859,0.0061949)(482.9610619,0.0061818)(483.9194011,0.0061696)(484.9588041,0.0061565)(485.9289704,0.0061442)(486.9802005,0.006131)(488.012294,0.0061181)(488.9751507,0.0061061)(490.0190713,0.0060932)(490.9937551,0.0060811)(491.9493023,0.0060694)(492.9859133,0.0060567)(493.9532876,0.0060449)(495.0017257,0.0060321)(496.0310272,0.0060197)(496.991092,0.0060081)(498.0322205,0.0059956)(499.0041124,0.005984)(499.0196731,0.0059838)(499.0352338,0.0059836)(499.0663553,0.0059832)(499.1285982,0.0059825)(499.253084,0.005981)(499.5020557,0.005978)(499.5176164,0.0059779)(499.5331771,0.0059777)(499.5642986,0.0059773)(499.6265415,0.0059766)(499.7510273,0.0059751)(499.7665881,0.0059749)(499.7821488,0.0059747)(499.8132703,0.0059743)(499.8755132,0.0059736)(499.8910739,0.0059734)(499.9066346,0.0059732)(499.9377561,0.0059729)(499.9533168,0.0059727)(499.9688775,0.0059725)(499.9844383,0.0059723)(499.999999,0.0059721) ]{}; coordinates [ (0,0.965936) (200,0.965936)]{};\[a=0\] coordinates | 33,0.0062083)(481.9334859,0.0061949)(482.9610619,0.0061818)(483.9194011,0.0061696)(484.9588041,0.0061565)(485.9289704,0.0061442)(486.9802005,0.006131)(488.012294,0.0061181)(488.9751507,0.0061061)(490.0190713,0.0060932)(490.9937551,0.0060811)(491.9493023,0.0060694)(492.9859133,0.0060567)(493.9532876,0.0060449)(495.0017257,0.0060321)(496.0310272,0.0060197)(496.991092,0.0060081)(498.0322205,0.0059956)(499.0041124,0.005984)(499.0196731,0.0059838)(499.0352338,0.0059836)(499.0663553,0.0059832)(499.1285982,0.0059825)(499.253084,0.005981)(499.5020557,0.005978)(499.5176164,0.0059779)(499.5331771,0.0059777)(499.5642986,0.0059773)(499.6265415,0.0059766)(499.7510273,0.0059751)(499.7665881,0.0059749)(499.7821488,0.0059747)(499.8132703,0.0059743)(499.8755132,0.0059736)(499.8910739,0.0059734)(499.9066346,0.0059732)(499.9377561,0.0059729)(499.9533168,0.0059727)(499.9688775,0.0059725)(499.9844383,0.0059723)(499.999999,0.0059721) ]{}; coordinates [ (0,0.965936) (200,0.965936)]{};\[a=0\] coordinates [ (0,0.379322 | 33,0.0062083)(481.9334859, 0.0061949) (482. 961 061 9, 0.00 6181 8)(483.9194011 , 0.00 61696)(484.9588041,0.0 06156 5) ( 485. 9 28 9704, 0.00614 4 2) ( 4 86. 98 02 005 ,0 . 00 6131) (48 8.01229 4,0.006118 1)( 48 8.9751507,0. 0 06 1061)(490. 019 0713,0.00609 32) (490.9 93 755 1 ,0.00 608 11)(4 91.949 3 023,0. 0060694)( 49 2 .98591 3 3,0.006 0 5 67 )(49 3.9532876,0.00604 4 9) ( 495.0017257,0. 006032 1) ( 49 6 . 031 027 2,0.006019 7) (496. 9 91092,0 . 00 6 0 0 81) ( 498.0322205,0 .0059956)(4 9 9.0 041124 ,0 .00 5 984)(4 99.01 96 7 31, 0.0059838)( 499. 0352338,0 .00598 3 6)(499. 0 663553, 0.0059 832 )(4 99.1 2 85 98 2,0 .0 0 598 2 5) (49 9 .25 3084,0.0 05 98 1)(49 9.50 2 0 5 5 7,0. 005 978) (499. 5176164,0.005 977 9)(4 9 9.5 33177 1,0.0 0597 77 )(499 .56429 86,0. 00 59773)(499.6265 415, 0.0059766 )(4 99 .75 10 273,0 . 005975 1)( 499 .766588 1,0.005 9 749 )( 4 9 9 .7 821488,0.0059747)( 49 9 . 81 32703,0. 005974 3 )( 49 9 .8755132 ,0 .00 5973 6 ) (499. 8910 7 39 ,0.00597 34)(49 9 .9 06 6346,0. 00 59732) (4 99. 937 7561, 0 .005 9729)( 499.9533 168,0 . 0059727)(499.9 6 88775,0.00597 2 5) ( 4 99 . 9844 383 ,0.0059723) (499 . 9999 99,0 . 00 597 2 1) ]{ }; co or d in a tes [ (0,0.965936)(2 00,0.9 65936 )]{};\[a=0\]coordinate s [ (0,0.37 9322 | 33,0.0062083)(481.9334859,0.0061949)(482.9610619,0.0061818)(483.9194011,0.0061696)(484.9588041,0.0061565)(485.9289704,0.0061442)(486.9802005,0.006131)(488.012294,0.0061181)(488.9751507,0.0061061)(490.0190713,0.0060932)(490.9937551,0.0060811)(491.9493023,0.0060694)(492.9859133,0.0060567)(493.9532876,0.0060449)(495.0017257,0.0060321)(496.0310272,0.0060197)(496.991092,0.0060081)(498.0322205,0.0059956)(499.0041124,0.005984)(499.0196731,0.0059838)(499.0352338,0.0059836)(499.0663553,0.0059832)(499.1285982,0.0059825)(499.253084,0.005981)(499.5020557,0.005978)(499.5176164,0.0059779)(499.5331771,0.0059777)(499.5642986,0.0059773)(499.6265415,0.0059766)(499.7510273,0.0059751)(499.7665881,0.0059749)(499.7821488,0.0059747)(499.8132703,0.0059743)(499.8755132,0.0059736)(499.8910739,0.0059734)(499.9066346,0.0059732)(499.9377561,0.0059729)(499.9533168,0.0059727)(499.9688775,0.0059725)(499.9844383,0.0059723)(499.999999,0.0059721) ]{};_coordinates [_(0,0.965936) (200,0.965936)]{};\[a=0\] coordinates [_(0,0.379322 |
use in practice due to the finite experimental resolution, however they still serve as useful examples to demonstrate the potential effects of varying bin sizes in the analysis. In the upper left plot we show the co-occurrences for the whole mixed sample with S/B=$5\%$ and four different choices of bin sizes. In each of the other three plots we then show the co-occurrences for different numbers of signal events (again $100$, $500$, and $1000$) and varying bin sizes. As expected, larger bin sizes result in stronger co-occurrences, however the size of this effect is not as large as the effect of removing observables from the analysis completely. For example, in all cases the strength of the co-occurrences at $m_0\simeq m_W$ is almost the same for all choices of the binning. The effect due to different bin sizes is more clearly seen away from these areas of strongest co-occurrence, where larger bin sizes result in stronger co-occurrences across the whole $m_{0}$ range. In particular, this may aid in better modelling of the signal and background distributions away from $m_{0}\simeq m_W$ and $m_0 \simeq m_\phi$. On the other hand, increasing the bin size will also make the signal features less pronounced, potentially reducing the classification power in the same way as a binned likelihood classifier becomes worse and worse approximation to the Neyman-Pearson un-binned likelihood. Therefore there is trade-off here between potential classification power and the ability of VI to extract optimal theme distributions from the data. In particular we find that the constant $\delta m_0 = 10$ GeV bin size provides the best trade-off for the examples satudied here and is also in practice close to the variable binning $\delta m_0 = 0.05 m_0$ mimicking the typical (energy) resolution of modern particle detector calorimeters.
Unsupervised learning with LDA {#sec:ldaresults}
==============================
As our main result we present two applications of the technique outlined in the preceding sections. Using the two benchmark examples discussed in Sec. \[sec:bm1\] and Sec. \[sec:bm2\] we construct various mixed event samples, i.e. mixtures of background and signal events. For the boosted $t\bar{t}$ example we construct mixed samples with $9\times 10^4$ events, and with S/B ratios: $ | use in practice due to the finite experimental resolution, however they still service as utilitarian examples to demonstrate the likely effects of vary bin sizes in the analysis. In the upper leftover plot we express the co - happening for the whole mixed sample distribution with S / B=$5\%$ and four different choices of bin sizes. In each of the early three plots we then show the conscientious objector - occurrences for different number of signal events (again $ 100 $, $ 500 $, and $ 1000 $) and varying bin sizes. As expect, larger bin sizes result in stronger carbon monoxide - occurrences, however the size of this effect is not as large as the effect of murder observables from the analysis completely. For example, in all cases the strength of the co - occurrences at $ m_0\simeq m_W$ is almost the same for all choices of the binning. The effect due to different bin sizes is more clearly seen away from these areas of strong co - occurrence, where big bin size result in stronger co - happening across the whole $ m_{0}$ range. In particular, this may aid in good modelling of the signal and background distributions away from $ m_{0}\simeq m_W$ and $ m_0 \simeq m_\phi$. On the other hand, increase the bin size will also make the signal features less pronounce, potentially reducing the categorization ability in the same way as a binned likelihood classifier becomes worse and worse approximation to the Neyman - Pearson un - binned likelihood. consequently there is craft - off here between potential classification might and the ability of VI to extract optimum theme distributions from the data. In particular we find that the changeless $ \delta m_0 = 10 $ GeV bin size leave the best barter - off for the examples satudied here and is also in practice close to the variable bin $ \delta m_0 = 0.05 m_0 $ mimicking the typical (energy) resoluteness of modern particle detector calorimeters.
Unsupervised learning with LDA { # sec: ldaresults }
= = = = = = = = = = = = = = = = = = = = = = = = = = = = = =
As our main consequence we present two applications of the proficiency outlined in the preceding sections. Using the two benchmark examples discussed in Sec. \[sec: bm1\ ] and Sec. \[sec: bm2\ ] we construct versatile mixed event samples, i.e. mixtures of background and signal events. For the hike $ t\bar{t}$ case we manufacture mixed samples with $ 9\times 10 ^ 4 $ events, and with S / B ratios: $ | usf in practice due to the finite experimgnral revolutikn, howevdr they still serve as usefup wxampoes to demonstrate the potentiap effectw of carying bii sizes lu the wnalvsms. In the upper left plot we show the cm-ozcbrrences for the whole mixed sample rith S/B=$5\%$ ajd four differgnt cnjicea of bin sizes. In each of the othed three plots we them show the co-occurrences flr dlfferent numbers ov signal evghts (qgain $100$, $500$, and $1000$) and varying bin sizes. As expected, larger bin sizes rdsult in stronggx co-lwcurrences, ioweveg the size of this efxect is not as large ss vhe wffect of removing obvervables from the analysis ckmpletely. For exanpoe, in all casdw tfe atcenfth of thx co-occurrehces at $m_0\sineq m_W$ is almost tht swne for all chkices jf the binning. The effect due to differenu bin sizes is more clearly ween away from these wreas of ftrongest co-occurrence, where larger bin sizes resglt ii rtrinner zi-ofcurrences across the whole $m_{0}$ range. In partictmat, nhis may aid in bctter modelling of tje fignal and bazkgrouus sistributions away from $m_{0}\fimeq m_W$ and $m_0 \simrq m_\phi$. On the other hand, ibcreasing thv bib size will also mcke the signcl feajures kess pronounced, potenticlly rsducing the classifidxtion power in tfe xake way as a binned likelihjod classmfier becomer wotse and worse appgoximation to the Neyman-Oearsln un-binned pikelihood. Therefore there is tcede-off here bgtwaen potenticl claxsification pjwer and the ayility oy VI tu extract kptimal theme distwibutions frok the data. In particujar qe fund thag the constant $\celta m_0 = 10$ GeV bin suze provides the bcst ttase-off for the ercmklws satudied hete xnd id elso yt practice cnose to yhe vxriable binkine $\dekta m_0 = 0.05 m_0$ mimicking dhe fypical (energy) resplmtion of nodern pwrticle detecyor calorimeters.
Undupertised uearnong with LDA {#sec:ldaresults}
==============================
As our jain resupt re present tro aiplisations of the technique outlined in the precedinj sections. Using the twi benchmark examplex discussed mn Sec. \[sec:bm1\] ang Sec. \[sec:bm2\] we consteuct various mixec event samples, i.e. mistures of bwckground and signal events. For the boosted $t\bar{t}$ example we construct mized samples with $9\tjmes 10^4$ evettf, and wyth D/U ratios: $ | use in practice due to the finite however still serve useful examples to varying sizes in the In the upper plot we show the co-occurrences for whole mixed sample with S/B=$5\%$ and four different choices of bin sizes. In of the other three plots we then show the co-occurrences for different numbers signal (again $500$, $1000$) and varying bin sizes. As expected, larger bin sizes result in stronger co-occurrences, however the of this effect is not as large as effect of removing observables the analysis completely. For example, all the strength the at m_W$ is almost same for all choices of the binning. The effect due to different bin sizes is more clearly away from of strongest where bin result in stronger the whole $m_{0}$ range. In particular, in better modelling of the signal and background away from m_W$ and $m_0 \simeq m_\phi$. On other hand, increasing the bin size will also the signal features less pronounced, potentially reducing the classification power in the same way as likelihood classifier becomes worse worse approximation to Neyman-Pearson likelihood. there trade-off here potential classification power and the ability of VI to extract optimal distributions from the data. In particular we find that the m_0 10$ GeV bin provides the best trade-off the satudied here and is practice to $\delta = m_0$ mimicking the typical resolution of modern particle detector Unsupervised learning with LDA result we present two applications of the technique in the preceding sections. Using the two examples discussed in Sec. \[sec:bm1\] and Sec. \[sec:bm2\] we construct various mixed samples, i.e. background and signal events. For the boosted $t\bar{t}$ we construct mixed samples $9\times 10^4$ events, and with S/B ratios: $ | use in practice due to the finiTe experimeNtal rEsoLutIoN, howEver They still serve AS useFul examples to demonstraTe the PoTEntiAL eFfectS of varyINg BIN siZeS iN thE aNAlYsis. IN thE upper lEft plot we sHow ThE co-occurrencES fOr the whole MixEd sample with s/B=$5\%$ aNd four DiFfeREnt chOicEs of bIn sizeS. in each Of the otheR tHRee ploTS we then SHOw The cO-occurrences for diFFeREnt numbers of siGnal evEnTS (aGAIn $100$, $500$, aNd $1000$) aNd varying bIn Sizes. aS expectED, lARGEr bIN sizes result iN stronger co-OCcuRrenceS, hOweVEr the sIze of ThIS efFect is not as LargE as the effEct of rEMoving oBServablEs from The AnaLysiS CoMpLetElY. for EXaMplE, In aLl cases tHe StRengtH of tHE CO-OccuRreNces At $m_0\siMeq m_W$ is almost The Same FOr aLl choIces oF the BiNning. the effEct duE tO different bin siZes iS more cleaRly SeEn aWaY from THese arEas Of sTrongesT co-occuRRenCe, WHERe Larger bin sizes resuLt IN StRonger co-OccurrENcEs ACross the WhOle $M_{0}$ ranGE. in parTicuLAr, This may aId in beTTeR mOdellinG oF the siGnAl aNd bAckgrOUnd dIstribUtions awAy froM $M_{0}\simeq m_W$ and $m_0 \siMEq m_\phi$. On the otHEr HANd, INcreAsiNg the bin sizE wilL Also Make THe SigNAl feaTures LeSS pROnounced, potentially ReDucing The clAssification pOwer in the sAME Way as a biNned LIkELihood classifiEr becOmes worse aND worse apProxiMation to The Neyman-pEArson un-bInnEd lIkeLihOOD. THerefore there IS TradE-oFf here bEtwEen poteNtiAl cLasSifIcAtion poweR and the aBiLiTy Of vI tO extrACt optimaL tHemE dIstRibutIOns froM the dAta. IN pArTIcuLar we fiND tHAT the CoNsTant $\DelTa M_0 = 10$ GeV bIn siZE prOvides tHe best traDe-oFF for ThE eXamples Satudied here aNd Is also in prAcTicE close TO The variaBle binning $\delta m_0 = 0.05 m_0$ mimickINg the tyPicAl (eneRgy) rEsolution Of mOdern pArtICle detEctor cAloriMeTerS.
uNsupeRVIsEd lEaRning with Lda {#Sec:LdareSuLts}
==============================
AS our maiN result we present twO AppLications of thE teChniQUE oUtlINeD In tHe PRecEDIng sections. UsinG the two benChMArK examples dIScuSsEd in Sec. \[Sec:bm1\] anD Sec. \[sEC:bm2\] we coNstruct vaRious mixeD eVent SAMplEs, i.e. mixturEs of backGround and SIgnal EVeNts. FoR thE boostEd $T\baR{t}$ exaMple we COnsTruct Mixed sAmPles wiTh $9\timEs 10^4$ Events, anD with S/B ratios: $ | use in practice due to th e finite e xperi men tal r esol utio n, however the y sti ll serve as useful exa mples t o dem o ns trate the po t en t i alef fe cts o f v aryin g b in size s in the a nal ys is. In the u p pe r left plo t w e show the c o-o ccurre nc esf or th e w holemixeds amplewith S/B= $5 \ %$ and four di f f er entchoices of bin si z es . In each of th e othe rt hr e e pl ots we then s ho w the co-occu r re n c e s f o r different n umbers of s i gna l even ts (a g ain $1 00$,$5 0 0$, and $1000$ ) an d varying bin s i zes. As expecte d, lar ger bi n si z es r esu lt ins tr ong e r c o-occurr en ce s, ho weve r t h e si zeof t his e ffect is notaslarg e as theeffec t of r emovi ng obs ervab le s from the anal ysis complete ly. F orex ample , in al l c ase s the s trength ofth e c o- occurrences at $m_ 0\ s i me q m_W$ i s almo s tth e same fo rall cho i c es of the bi nning. T he eff e ct d ue to d if ferent b insiz es is more clear ly seenawayf rom these area s of strongest co - o cc u rren ce, where larg er b i n si zesr es ult in st ronge rc o- o ccurrences across t he whole $m_{ 0}$ range. In particula r , this may aid in better modelli ng of the signa l and bac kgrou nd distr ibutionsa w ay from$m_ {0} \si meq m _W $ and $m_0 \s i m eq m _\ phi$. O n t he othe r h and , i ncr ea sing thebin size w il lal somaket he signa lfea tu res less pronou nced, pot en ti a lly reduci n gt h e cl as si fica tio npower int hesame wa y as a bi nne d lik el ih ood cla ssifier becom es worse and w ors e appr o x imationto the Neyman-Pearson u n -binned li kelih ood. Therefor e t here i s t r ade-of f here betw ee n p o t entia l cl ass if ication po w e r a nd th eabil ity ofVI to extract opti m altheme distrib uti onsf r om th e d a ta. I n pa r t icular we findthat the c on s ta nt $\delta m_0 = 10$ Ge V bin s ize p r ovidesthe besttrade-off f or t h e ex amples sat udied he re and is alsoi npract ice close t o t he va riable bin ning$\delt am_0 =0.05m_ 0$ mimic king the typical (energ y) res oluti onof modern pa r tic le detect or c alorimeter s.
Un super vis e d lea rnin g w ith LDA { #sec : ldaresult s }=== = = == =========== = = = === ===== =
A s ourmain result we presen t two applicati onso f th e t e chni qu e outlined inthe p r e ceding s ec tions. Usin g the tw ob enchm ark ex amples discus s e di n Sec. \[s ec: bm1\] and Se c. \[sec:b m2 \] we con stru ct vario us mix e d ev e n t samples, i.e.mixtu r e s ofb ack groun dand sig n al e vents. For the booste d $t\b ar{t }$ ex ample w econstr uct m ixed sampl e s with $9 \time s 10^4$ e vent s,and wi th S / B rati os:$ | use_in practice_due to the finite_experimental resolution,_however_they still_serve_as useful examples_to demonstrate the_potential effects of varying_bin sizes in_the_analysis. In the upper left plot we show the co-occurrences for the whole mixed_sample_with S/B=$5\%$_and_four_different choices of bin sizes._In each of the other_three plots_we then show the co-occurrences for different numbers_of_signal events (again_$100$, $500$, and $1000$) and varying bin sizes. As_expected, larger bin sizes result in_stronger co-occurrences, however_the_size_of this effect is_not as large as the effect_of removing observables from the analysis_completely. For example, in all cases the_strength of the co-occurrences at $m_0\simeq_m_W$ is almost the same_for all_choices of the binning. The_effect due to_different bin_sizes is more_clearly seen away from these areas_of strongest co-occurrence,_where larger bin sizes result in_stronger_co-occurrences across the_whole_$m_{0}$_range. In_particular, this may_aid_in better_modelling_of the signal and background distributions_away_from $m_{0}\simeq m_W$ and $m_0 \simeq m_\phi$._On the other hand,_increasing_the bin size will_also make the signal features_less pronounced, potentially reducing the classification_power in_the same_way as a binned likelihood classifier becomes worse and worse approximation_to the Neyman-Pearson un-binned likelihood. Therefore_there is trade-off here_between potential_classification_power and the_ability_of VI_to extract optimal theme distributions from the_data. In_particular we find that the constant_$\delta m_0 = 10$_GeV_bin size provides the best trade-off_for the examples satudied here and_is also in practice close_to_the_variable binning $\delta m_0 =_0.05 m_0$ mimicking the typical (energy)_resolution of modern_particle detector calorimeters.
Unsupervised learning with LDA {#sec:ldaresults}
==============================
As_our_main result we present two applications_of_the technique outlined in the preceding_sections._Using_the two benchmark examples discussed_in Sec. \[sec:bm1\] and Sec. \[sec:bm2\]_we construct various mixed event samples, i.e. mixtures of_background and signal_events. For the boosted $t\bar{t}$_example_we_construct mixed samples with $9\times 10^4$ events, and with S/B_ratios: $ |
radius $R=30 {\;\mathrm{fm}}$; for smaller screening radii the convergence in orbital angular momentum is faster. The hadronic interaction is taken into account in two-baryon partial waves with total angular momentum $I \le 5$. Both three-baryon total isospin $\mathcal{T} = \frac12$ and $\mathcal{T} = \frac32$ states are included. The maximal three-baryon total angular momentum $\mathcal{J}$ considered is $\frac{61}{2}$.
![\[fig:R13\] Convergence of the $pd$ breakup observables with screening radius $R$. The differential cross section and the proton analyzing power $A_y(N)$ for $pd$ breakup at 13 MeV proton lab energy are shown as functions of the arclength $S$ along the kinematical curve. Results for CD Bonn potential obtained with screening radius $R= 10$ fm (dotted curves), 20 fm (dash-dotted curves), and 30 fm (solid curves) are compared. Results without Coulomb (dashed curves) are given as reference for the size of the Coulomb effect.](R13.eps)
![\[fig:R130d\] Convergence of the $pd$ breakup observables with screening radius $R$. The differential cross section and the deuteron analyzing power $A_{xx}$ for $pd$ breakup at 130 MeV are shown. Curves as in [Fig.]{} \[fig:R13\].](R130d.eps)
![\[fig:R13ppfsi\] Convergence of the $pd$ breakup observables with screening radius $R$. The differential cross section for $pd$ breakup at 13 MeV proton lab energy in the $pp$-FSI configuration is shown as function of the relative $pp$ energy $E_{pp}$. Results obtained with screening radius $R= 10$ fm (dotted curves), 20 fm (dashed-double-dotted curves), 30 fm (dashed-dotted curves), 40 fm (double-dashed-dotted curves), and 60 fm (solid curves) are compared. Results without Coulomb (dashed curves) are given as reference for the size of the Coulomb effect.](R13eppfsi.eps)
Figures \[fig:R13\] - \[fig:R13ppfsi\] study the convergence of our method with increasing screening radius $R$ according to [Eq.]{} | radius $ R=30 { \;\mathrm{fm}}$; for smaller screening radii the convergence in orbital angular momentum is fast. The hadronic interaction is accept into account in two - baryon partial waves with full angular momentum $ I \le 5$. Both three - baryon total isospin $ \mathcal{T } = \frac12 $ and $ \mathcal{T } = \frac32 $ states are admit. The maximal three - baryon full angular momentum $ \mathcal{J}$ considered is $ \frac{61}{2}$.
! [ \[fig: R13\ ] Convergence of the $ pd$ dissolution observables with screening radius $ R$. The differential hybridization section and the proton analyzing power $ A_y(N)$ for $ pd$ dissolution at 13 MeV proton lab energy are shown as functions of the arclength $ S$ along the kinematical curve. Results for CD Bonn electric potential obtained with screening radius $ R= 10 $ fermium (dotted curves), 20 frequency modulation (dash - dot curves), and 30 fm (solid curves) are compared. result without Coulomb (dashed curves) are given as reference for the size of the Coulomb effect.](R13.eps)
! [ \[fig: R130d\ ] Convergence of the $ pd$ breakup observables with screening radius $ R$. The differential cross section and the deuteron analyzing power $ A_{xx}$ for $ pd$ breakup at 130 MeV are shown. Curves as in [ Fig. ] { } \[fig: R13\].](R130d.eps)
! [ \[fig: R13ppfsi\ ] Convergence of the $ pd$ breakup observables with screening radius $ R$. The differential crisscross section for $ pd$ breakup at 13 MeV proton lab department of energy in the $ pp$-FSI shape is shown as function of the relative $ pp$ energy $ E_{pp}$. result obtained with screening radius $ R= 10 $ fm (dashed curves), 20 fm (dashed - double - dotted curves), 30 fm (dashed - dotted curves), 40 fm (double - dashed - dotted curve), and 60 frequency modulation (solid curves) are compared. Results without Coulomb (dashed bend) are given as reference for the size of the Coulomb effect.](R13eppfsi.eps)
Figures \[fig: R13\ ] - \[fig: R13ppfsi\ ] study the convergence of our method with increasing screening spoke $ R$ according to [ Eq. ] { } | rafius $R=30 {\;\mathrm{fm}}$; for smaluer screening rceii thx convedgence iv orbital angular momentum id dastee. The hadronic interacgion is twken inti acrount in two-baryon partial wavsd wich total angular momentum $H \le 5$. Both thrae-caxyon total isospin $\mathcal{T} = \frac12$ anq $\mathcsl{H} = \frac32$ states are ynclhded. The maximal three-baryon total angulag momentum $\mathcak{J}$ considered is $\frac{61}{2}$.
![\[fig:R13\] Fonvfrgence of the $pd$ hreakup obsgdvafoes with scrdening radpbs $R$. The dirferential cross section and thd procon analyzibg poagr $A_y(N)$ for $pv$ breahup at 13 MeV pvpton ldb enerby are shown ax fnnctuons of the arclength $S$ along the kinemwtical cusvz. Results for CD Bonn pitenthal mbtakbed wiuh xcdeeninh redius $R= 10$ fm (sotted curvws), 20 fm (dash-dotted cutvqw), and 30 fm (solis curvqs) are compared. Results without Coulomb (ddshsd curves) are given as eeference for the sizg of the Cjulomb effect.](R13.eps)
![\[fig:R130d\] Convergence of the $pd$ breanup ousdrvcnjdw aith screening radius $R$. The differential crosf stctpon and the deutevon analyzing powet $W_{xc}$ for $pd$ breakop at 130 MzB zre shown. Curves ad in [Fid.]{} \[fig:R13\].](E130d.eps)
![\[fig:R13kpfsi\] Convergence of the $pd$ breajup observabjws with screening xadius $R$. The difgerenyial cross section for $'d$ brezkup at 13 MeV proton lzc energy in the $op$-FXI configuration is shown as function of che relagive $pp$ enqrgy $E_{pp}$. Rfsults obtained with scrfeniny radhus $R= 10$ fm (dltted curves), 20 fm (dashed-double-dotvxd curves), 30 fm (cavheg-dotted eurves), 40 fm (double-dasred-dotted curvgs), and 60 fm (solia curves) age comparxd. Results wythout Coulomt (dashed curvxs) are giden qs rwferencd for the size pf the Colljmb wffect.](R13eppfsi.eps)
Finures \[wjg:R13\] - \[fig:R13ppfsi\] wtuey the convergemce of olr keevod with incseaskng xcreeving radius $F$ acvording to [Eq.]{} | radius $R=30 {\;\mathrm{fm}}$; for smaller screening radii in angular momentum faster. The hadronic in partial waves with angular momentum $I 5$. Both three-baryon total isospin $\mathcal{T} \frac12$ and $\mathcal{T} = \frac32$ states are included. The maximal three-baryon total angular $\mathcal{J}$ considered is $\frac{61}{2}$. ![\[fig:R13\] Convergence of the $pd$ breakup observables with screening $R$. differential section the proton analyzing power $A_y(N)$ for $pd$ breakup at 13 MeV proton lab energy are shown functions of the arclength $S$ along the kinematical Results for CD Bonn obtained with screening radius $R= fm curves), 20 (dash-dotted and fm (solid curves) compared. Results without Coulomb (dashed curves) are given as reference for the size of the Coulomb effect.](R13.eps) Convergence of breakup observables screening $R$. differential cross section deuteron analyzing power $A_{xx}$ for $pd$ MeV are shown. Curves as in [Fig.]{} \[fig:R13\].](R130d.eps) Convergence of $pd$ breakup observables with screening radius The differential cross section for $pd$ breakup at MeV proton lab energy in the $pp$-FSI configuration is shown as function of the relative $E_{pp}$. Results obtained with radius $R= 10$ (dotted 20 (dashed-double-dotted 30 fm curves), 40 fm (double-dashed-dotted curves), and 60 fm (solid curves) are Results without Coulomb (dashed curves) are given as reference for of Coulomb effect.](R13eppfsi.eps) Figures - \[fig:R13ppfsi\] study the of method with increasing screening according [Eq.]{} | radius $R=30 {\;\mathrm{fm}}$; for smaller Screening rAdii tHe cOnvErGencE in oRbital angular mOMentUm is faster. The hadronic iNteraCtIOn is TAkEn intO accounT In TWO-baRyOn ParTiAL wAves wIth Total anGular momenTum $i \lE 5$. Both three-baRYoN total isosPin $\Mathcal{T} = \frac12$ And $\MathcaL{T} = \FraC32$ StateS arE inclUded. ThE MaximaL three-barYoN Total aNGular moMENtUm $\maThcal{J}$ considered iS $\FrAC{61}{2}$.
![\[fig:R13\] ConvergenCe of thE $pD$ BrEAKup ObsErvables wiTh ScreeNIng radiUS $R$. tHE DifFErential cross Section and tHE prOton anAlYziNG power $a_y(N)$ foR $pD$ BreAkup at 13 MeV prOton Lab energy Are shoWN as funcTIons of tHe arclEngTh $S$ AlonG ThE kIneMaTIcaL CuRve. rEsuLts for CD boNn PotenTial OBTAIned WitH scrEeninG radius $R= 10$ fm (dotTed CurvES), 20 fm (Dash-dOtted CurvEs), And 30 fm (Solid cUrves) ArE compared. ResultS witHout CouloMb (dAsHed CuRves) aRE given As rEfeRence foR the sizE Of tHe cOULoMb effect.](R13.eps)
![\[fig:R130d\] COnVERgEnce of thE $pd$ breAKuP oBServableS wIth ScreENIng raDius $r$. thE differeNtial cROsS sEction aNd The deuTeRon AnaLyzinG PoweR $A_{xx}$ foR $pd$ breakUp at 130 MEv are shown. CurveS As in [Fig.]{} \[fig:R13\].](R130d.EPs)
![\[FIG:R13PPfsi\] conVergence of tHe $pd$ BReakUp obSErVabLEs witH screEnINg RAdius $R$. The differentiAl Cross sEctioN for $pd$ breakup At 13 MeV protoN LAB energy iN the $PP$-Fsi configuration Is shoWn as functiON of the reLativE $pp$ energY $E_{pp}$. ResulTS Obtained WitH scReeNinG RAdIus $R= 10$ fm (dotted cURVes), 20 fM (dAshed-doUblE-dotted CurVes), 30 Fm (dAshEd-Dotted curVes), 40 fm (douBlE-dAsHeD-doTted cURves), and 60 fM (sOliD cUrvEs) are COmpareD. ResuLts wItHoUT CoUlomb (daSHeD CUrveS) aRe GiveN as ReFerenCe foR The Size of tHe Coulomb EffECt.](R13ePpFsI.eps)
FigUres \[fig:R13\] - \[fig:R13pPfSi\] study the CoNveRgence OF Our methoD with increasing screeninG Radius $R$ AccOrdinG to [EQ.]{} | radius $R=30 {\;\mathrm{f m}}$; forsmall erscr ee ning rad ii the converg e ncein orbital angular mom entum i s fas t er . The hadron i ci n ter ac ti onis ta ken i nto accoun t in two-b ary on partial wav e swith total an gular moment um$I \le 5 $.B oth t hre e-bar yon to t al iso spin $\ma th c al{T}= \frac1 2 $ a nd $ \mathcal{T} = \fr a c3 2 $ states are i nclude d. Th e max ima l three-ba ry on to t al angu l ar m o men t um $\mathcal{ J}$ conside r edis $\f ra c{6 1 }{2}$.
![\ [f i g:R 13\] Conver genc e of the$pd$ b r eakup o b servabl es wit h s cre enin g r ad ius $ R $.T he di f fer ential c ro ss sect iona n d thepro tonanaly zing power $A _y( N)$f or$pd$break up a t13 Me V prot on la benergy are show n as function s o fthe a rclen g th $S$ al ong the ki nematic a l c ur v e . R esults for CD Bonn p o t en tial obt ainedw it hs creening r adi us $ R = 10$fm ( d ot ted curv es), 2 0 f m(dash-d ot ted cu rv es) , a nd 30 fm ( solidcurves)are c o mpared. Result s without Coul o mb ( da s hedcur ves) are gi vena s re fere n ce fo r thesizeof th e Coulomb effect.](R 13 .eps)
![\[ fig:R130d\] C onvergence o f the $pd $ br e ak u p observableswithscreeningr adius $R $. Th e differ ential cr o s s sectio n a ndthe de u t er on analyzingp o wer$A _{xx}$for $pd$ b rea kup at 13 0MeV are s hown. Cu rv es a sin[Fig. ] {} \[fig :R 13\ ]. ](R 130d. e ps)
! [\[fi g:R1 3p pf s i\] Conver g en c e ofth e$pd$ br ea kup o bser v abl es with screenin g r a dius $ R$ . The d ifferential c ro ss section f or$pd$ b r e akup at13 MeV proton lab energ y in the $p p$-FS I co nfigurati onis sho wna s func tion o f the r ela t i ve $p p $ e ner gy $E_{pp}$. R esu lts o bt aine d withscreening radius $ R = 1 0$ fm (dotted cu rves ) , 2 0 f m ( d ash ed - dou b l e-dotted curves ), 30 fm ( da s he d-dotted c u rve s) , 40 fm (doubl e-das h ed-dott ed curves ), and 60 f m (s o l idcurves) ar e compar ed. Resul t s wit h ou t Cou lom b (das he d c urves ) areg ive n asrefere nc e forthe s iz e of the Coulomb effect.](R13ep pfsi.e ps)
Fig ures \[fi g:R 1 3\] - \[fig: R13p pfsi\] stu dythe conv erg e nce o f ou r m eth o d wit h in c reasing s c re eni n g r adius $R$ a c c o rdi ng to [E q .]{} | radius_$R=30 {\;\mathrm{fm}}$;_for smaller screening radii_the convergence_in_orbital angular_momentum_is faster. The_hadronic interaction is_taken into account in_two-baryon partial waves_with_total angular momentum $I \le 5$. Both three-baryon total isospin $\mathcal{T} = \frac12$ and_$\mathcal{T}_= \frac32$_states_are_included. The maximal three-baryon total_angular momentum $\mathcal{J}$ considered is_$\frac{61}{2}$.
![\[fig:R13\] Convergence_of the $pd$ breakup observables with screening radius_$R$._The differential cross_section and the proton analyzing power $A_y(N)$ for $pd$_breakup at 13 MeV proton lab energy_are shown as_functions_of_the arclength $S$ along_the kinematical curve. Results for CD_Bonn potential obtained with screening radius_$R= 10$ fm (dotted curves), 20 fm (dash-dotted curves),_and 30 fm (solid curves) are compared._Results without Coulomb (dashed curves)_are given_as reference for the size_of the Coulomb_effect.](R13.eps)
![\[fig:R130d\] Convergence_of the $pd$_breakup observables with screening radius $R$._The differential cross_section and the deuteron analyzing power_$A_{xx}$_for $pd$ breakup_at_130 MeV_are shown._Curves as in_[Fig.]{} \[fig:R13\].](R130d.eps)
![\[fig:R13ppfsi\]_Convergence of_the_$pd$ breakup observables with screening radius_$R$._The differential cross section for $pd$ breakup_at 13 MeV proton lab_energy_in the $pp$-FSI configuration_is shown as function of_the relative $pp$ energy $E_{pp}$. Results_obtained with_screening radius_$R= 10$ fm (dotted curves), 20 fm (dashed-double-dotted curves), 30 fm (dashed-dotted curves), 40 fm_(double-dashed-dotted curves), and 60 fm (solid curves)_are compared. Results without_Coulomb (dashed_curves)_are given as_reference_for the_size of the Coulomb effect.](R13eppfsi.eps)
Figures \[fig:R13\] - \[fig:R13ppfsi\]_study the_convergence of our method with increasing_screening radius $R$ according_to_[Eq.]{} |
the velocity shift caused by the change in the zero-point energy density of the quantized electromagnetic field induced by the presence of Casimirlike plates. Recall that an external electromagnetic field such as that of a propagating light couples to the quantized radiation field through fermionic loops. The Scharnhorst effect is not the only example where non-trivial vacua affects the speed of light. In fact this subject has attracted the attention of many physicists in the last years [@Adler; @Drummond; @DanielsShore; @Shore; @Latorre; @Dittrich].
It is clear from what was stated above that an analysis of the QED vacuum inside cavities is crucial for an understanding of its observable properties. Here we shall consider the QED vacuum confined by an unsual pair of mirrors. Specifically, we shall place an infinite perfectly conducting ($\epsilon\to\infty$) surface parallel to a second infinite perfectly permeable ($\mu\to\infty$) surface held at fixed distance $L$ from the first. This setup was first considered by Boyer in order to compute the corresponding Casimir effect in the framework of random electrodynamics [@BOYER] and leads to a repulsive force. This result is somewhat intriguing, since it seems to contradict the explanation given for the usual attractive Casimir effect which suggests that there is a greater number of modes outside the plates than inside [@Milonnibook]. In fact, this is not true: there is only a rearrangement of modes, for a nice explanation of this problem see [@Hushwater]. For the generalized $\zeta$-function approach applied to the repulsive Casimir effect for parallel plates geometry see [@Andre; @SANTOS].
This paper is organized as follows: in section [**2**]{} we determine the photon field ${\mbox{\boldmath $A$}}({\mbox{\boldmath $r$}},t)$ in the region between Boyer’s plates making use of the Coulomb gauge. Next we also evaluate the field operator correlators $<{\hat E}_i{\hat E}_j>_0$ and $<{\hat B}_i{\hat B}_j>_0$ with the aid of a simple but efficient regularization prescription. In section [**3**]{} we apply our results to reobtain the repulsive Casimir pressure of this setup. In section [**4**]{} we discuss the Scharnhorst effect but for this different situation. In particular, we show that, contrary to the case with of the usual pair of | the velocity shift caused by the change in the zero - item department of energy density of the quantized electromagnetic sphere induce by the presence of Casimirlike plates. remember that an external electromagnetic field such as that of a propagating unaccented couples to the quantized radiotherapy field through fermionic loops. The Scharnhorst impression is not the lone example where non - trivial vacua affects the speed of luminosity. In fact this subject has attracted the attention of many physicist in the last years [ @Adler; @Drummond; @DanielsShore; @Shore; @Latorre; @Dittrich ].
It is clean from what was stated above that an psychoanalysis of the QED vacuum inside cavity is crucial for an understanding of its observable property. Here we shall consider the QED vacuum confined by an unsual pair of mirrors. Specifically, we shall invest an infinite perfectly conducting ($ \epsilon\to\infty$) surface parallel to a second infinite perfectly permeable ($ \mu\to\infty$) surface held at fixed distance $ L$ from the first. This apparatus was first considered by Boyer in decree to calculate the corresponding Casimir effect in the framework of random electrodynamics [ @BOYER ] and leads to a repulsive force. This result is somewhat challenging, since it seems to contradict the explanation given for the usual attractive Casimir effect which suggests that there is a greater number of modes outside the plates than inside [ @Milonnibook ]. In fact, this is not true: there be merely a rearrangement of modes, for a nice explanation of this problem see [ @Hushwater ]. For the generalized $ \zeta$-function approach practice to the repulsive Casimir effect for parallel plates geometry examine [ @Andre; @SANTOS ].
This paper is organized as follows: in section [ * * 2 * * ] { } we decide the photon field $ { \mbox{\boldmath $ A$ } } ({ \mbox{\boldmath $ r$}},t)$ in the region between Boyer ’s plate making function of the Coulomb gauge. Next we besides evaluate the field hustler correlators $ < { \hat E}_i{\hat E}_j>_0 $ and $ < { \hat B}_i{\hat B}_j>_0 $ with the aid of a simple but efficient regularization prescription. In part [ * * 3 * * ] { } we apply our solution to reobtain the repulsive Casimir pressure of this setup. In section [ * * 4 * * ] { } we discuss the Scharnhorst consequence but for this different situation. In finical, we show that, contrary to the character with of the usual pair of | thf velocity shift caused ny the change in the zeco-point energy aensity of the quantized elertronagneupc field induced by tfe presenbe of Casumirouke plates. Recall bkat ah extzrial electromagngtic field sgch as that of a pxopagating light couples to the quaneized rsdlation field trroubr fedmionic loops. The Scharnhorst effedt is nmt the only ecample where non-trivial vafua wffects the speed lf light. In face this subjecg has attracted the atjention of many physicists in thd lasc years [@Adlgx; @Drkkmond; @DanieosShoge; @Shore; @Latovge; @Dittsich].
It os clear from ehav waw stated above that ai analysis of the QEQ vacuum hnaide cavities is xrycial for an jbdefstznvinf of ihs kbservable properties. Here we shall consodqg the QED vachum cogfyned by an unsual pair of mirrors. Specixiczlly, we shall place an unfinite perfectly cojducting ($\qpsilon\to\infty$) surface parallel to a second infinhte pxrweculn owrleable ($\mu\to\infty$) surface held at fixed distanss $K$ nrom the first. Tmis setup was firsy fomfidered by Boier in ordsr to compute the forrespjndint Casimir effrct in the framework of raneom electrodjnamucs [@BOYER] and leadd to a repuusivg forcr. This result is somewhct intdiguing, sinfe it seejr to contradict ghe efplanation given for the ufual attrectivz Casimif efgect wrich suggedts tmdt there is a greaher nomber mf modes oktside the plates than inside [@Mmkonnibook]. In fdct, this is not brue: there is ogly a rearranggment of iodes, for a nicv explanavion of this problem see [@Voshwater]. For vhe generwlizwd $\zwta$-funcgkon approach alplied to the repulwive Casimir effecb for larallel plates geimetry see [@Andrr; @SXNTJS].
Nhix [dper is orgatizea ar folluws: in sectlon [**2**]{} we determine the photot fisld ${\mbox{\boldmath $A$}}({\kbjx{\boldmajh $r$}},t)$ in ehe region beyween Boyer’s plated makmng usx of tne Soulomb gauge. Next we also evamuate the ficld operator sorrclatjrs $<{\hat E}_i{\kat E}_j>_0$ and $<{\hat B}_i{\hat B}_j>_0$ with the aid of e simple but efficient eegularization preserlption. In sertion [**3**]{} we apply our results to reobrain the repulsivt Casimir pressure of tgis sedup. Ij section [**4**]{} we discuss the Scharnhorst effect but for this different situarion. Ii [articular, se snow tvac, cpntrarr to vhe case with of nhe usual pair of | the velocity shift caused by the change zero-point density of quantized electromagnetic field Casimirlike Recall that an electromagnetic field such that of a propagating light couples the quantized radiation field through fermionic loops. The Scharnhorst effect is not the example where non-trivial vacua affects the speed of light. In fact this subject attracted attention many in the last years [@Adler; @Drummond; @DanielsShore; @Shore; @Latorre; @Dittrich]. It is clear from what was above that an analysis of the QED vacuum cavities is crucial for understanding of its observable properties. we consider the vacuum by unsual pair of Specifically, we shall place an infinite perfectly conducting ($\epsilon\to\infty$) surface parallel to a second infinite perfectly permeable surface held distance $L$ the This was first considered in order to compute the corresponding the framework of random electrodynamics [@BOYER] and leads a repulsive This result is somewhat intriguing, since seems to contradict the explanation given for the attractive Casimir effect which suggests that there is a greater number of modes outside the inside [@Milonnibook]. In fact, is not true: is a of for a explanation of this problem see [@Hushwater]. For the generalized $\zeta$-function approach to the repulsive Casimir effect for parallel plates geometry see This is organized as in section [**2**]{} we the field ${\mbox{\boldmath $A$}}({\mbox{\boldmath $r$}},t)$ region Boyer’s of Coulomb Next we also evaluate field operator correlators $<{\hat E}_i{\hat and $<{\hat B}_i{\hat B}_j>_0$ simple but efficient regularization prescription. In section [**3**]{} apply our results to reobtain the repulsive pressure of this setup. In section [**4**]{} we discuss the Scharnhorst effect for this In particular, we show that, contrary to the with of the usual of | the velocity shift caused by tHe change in The zeRo-pOinT eNergY denSity of the quantIZed eLectromagnetic field indUced bY tHE preSEnCe of CAsimirlIKe PLAteS. REcAll ThAT aN exteRnaL electrOmagnetic fIelD sUch as that of a PRoPagating liGht Couples to the QuaNtized RaDiaTIon fiEld ThrouGh fermIOnic loOps. The SchArNHorst eFFect is nOT ThE onlY example where non-tRIvIAl vacua affects The speEd OF lIGHt. IN faCt this subjEcT has aTTracted THe ATTEntIOn of many physiCists in the lASt yEars [@AdLeR; @DrUMmond; @DAnielSSHOre; @shore; @LatorrE; @DitTrich].
It is Clear fROm what wAS stated Above tHat An aNalySIs Of The qEd VacUUm InsIDe cAvities iS cRuCial fOr an UNDERstaNdiNg of Its obServable propeRtiEs. HeRE we Shall ConsiDer tHe qED vaCuum coNfineD bY an unsual pair of MirrOrs. SpecifIcaLlY, we ShAll plACe an inFinIte PerfectLy conduCTinG ($\ePSILoN\to\infty$) surface parAlLEL tO a second InfiniTE pErFEctly perMeAblE ($\mu\tO\INfty$) sUrfaCE hEld at fixEd distANcE $L$ From the FiRst. ThiS sEtuP waS firsT ConsIdered By Boyer iN ordeR To compute the coRResponding CasIMiR EFfECt in The Framework of RandOM eleCtroDYnAmiCS [@BOYEr] and lEaDS tO A repulsive force. This ReSult is SomewHat intriguing, Since it seeMS TO contradIct tHE eXPlanation given For thE usual attrACtive CasImir eFfect whiCh suggestS THat there Is a GreAteR nuMBEr Of modes outsidE THe plAtEs than iNsiDe [@MilonNibOok]. in fAct, ThIs is not trUe: there iS oNlY a ReArrAngemENt of modeS, fOr a NiCe eXplanATion of This pRoblEm SeE [@husHwater]. FOR tHE GeneRaLiZed $\zEta$-FuNctioN appROacH applieD to the repUlsIVe CaSiMiR effect For parallel plAtEs geometry SeE [@AnDre; @SANtos].
This papEr is organized as follows: iN Section [**2**]{} We dEtermIne tHe photon fIelD ${\mbox{\bOldMAth $A$}}({\mbOx{\boldMath $r$}},T)$ iN thE REgion BETwEen boYer’s plates MAKinG use oF tHe CoUlomb gaUge. Next we also evaluATe tHe field operatOr cOrreLAToRs $<{\hAT E}_I{\Hat e}_j>_0$ ANd $<{\hAT b}_i{\hat B}_j>_0$ with the aId of a simplE bUT eFficient reGUlaRiZation pRescripTion. IN Section [**3**]{} We apply ouR results tO rEobtAIN thE repulsive casimir pRessure of THis seTUp. in secTioN [**4**]{} we disCuSs tHe SchArnhorST efFect bUt for tHiS diffeRent sItUation. In Particular, we show that, conTrary tO the cAse With of the UsuAL paIr of | the velocity shift caused by the ch angeinthe z ero- poin t energy densi t y of the quantized electro magne ti c fie l dinduc ed by t h ep r ese nc eofCa s im irlik e p lates.Recall tha t a nexternal ele c tr omagneticfie ld such as t hat of apr opa g ating li ght c ouples to the quantize dr adiati o n field t hr ough fermionic loops. Th e Scharnhorst e ffectis no t the on ly example w heren on-triv i al v a cua affects the s peed of lig h t.In fac tthi s subje ct ha sa ttr acted the a tten tion of m any ph y sicists in thelast y ear s [ @Adl e r; @ Dru mm o nd; @D ani e lsS hore; @S ho re ; @La torr e ; @ Ditt ric h].
It i s clear fromwha t wa s st atedabove tha tan an alysis of t he QED vacuum ins idecavitiesiscr uci al fora n unde rst and ing ofits obs e rva bl e p ro perties. Here we s ha l l c onsiderthe QE D v ac u um confi ne d b y an u nsual pai r o f mirror s. Spe c if ic ally, w eshallpl ace an infi n iteperfec tly cond uctin g ($\epsilon\to \ infty$) surfa c ep a ra l leltoa second in fini t e pe rfec t ly pe r meabl e ($\ mu \ to \ infty$) surface hel dat fix ed di stance $L$ fr om the fir s t . This se tupw as first consider ed by Boyer ino rder tocompu te the c orrespond i n g Casimi r e ffe ctint h eframework ofr a ndom e lectrod yna mics [@ BOY ER] an d l ea ds to a r epulsive f or ce .Thi s res u lt is so me wha tint rigui n g, sin ce it see ms t o co ntradic t t h e exp la na tion gi ve n for the usu al attr active Ca sim i r ef fe ct whichsuggests that t here is agr eat er num b e r of mod es outside the plates t h an insi de[@Mil onni book]. In fa ct, th isi s nottrue:there i s o n l y a r e a rr ang em ent of mod e s , f or ani ce e xplanat ion of this proble m se e [@Hushwater ].Fort h egen e ra l ize d$ \ze t a $-function appr oach appli ed to the repul s ive C asimireffectfor p a rallelplates ge ometry se e[@An d r e;@SANTOS].
This pa per is or g anize d a s fol low s: inse cti on [* *2**]{ } we dete rmineth e phot on fi el d ${\mbo x{\boldmath $A$}}({\mbo x{\bol dmath $r $}},t)$ i n t h e r egion bet ween Boyer’s p lat esmakin g u s e oftheC ou lom b gaug e. N e xt we als o e val u a te the fieldo p e rat or co rre l ators$<{\ hat E}_i{\hat E}_ j >_0$ and $<{\h at B } _ i{\ hat B}_j >_ 0$ with the ai d o fa simple b ut efficientregulari za t ion p rescri ption. In sec t i on [**3** ]{}weapply our re su l ts to r eo bt a in the rep ul sive C asimir pres s u re of this setup . Ins e ction [** 4**]{ }we disc u ss t he Scharnh orst effect but f or t his d ifferen tsituat ion .In particu l ar, we sh ow th at, con tr arytothe ca se w i t h oftheus ual pair of | the_velocity shift_caused by the change_in the_zero-point_energy density_of_the quantized electromagnetic_field induced by_the presence of Casimirlike_plates. Recall that_an_external electromagnetic field such as that of a propagating light couples to the quantized_radiation_field through_fermionic_loops._The Scharnhorst effect is not_the only example where non-trivial_vacua affects_the speed of light. In fact this subject_has_attracted the attention_of many physicists in the last years [@Adler; @Drummond;_@DanielsShore; @Shore; @Latorre; @Dittrich].
It is clear_from what was_stated_above_that an analysis of_the QED vacuum inside cavities is_crucial for an understanding of its_observable properties. Here we shall consider the_QED vacuum confined by an unsual_pair of mirrors. Specifically, we_shall place_an infinite perfectly conducting ($\epsilon\to\infty$)_surface parallel to_a second_infinite perfectly permeable_($\mu\to\infty$) surface held at fixed distance_$L$ from the_first. This setup was first considered_by_Boyer in order_to_compute_the corresponding_Casimir effect in_the_framework of_random_electrodynamics [@BOYER] and leads to a_repulsive_force. This result is somewhat intriguing, since_it seems to contradict_the_explanation given for the_usual attractive Casimir effect which_suggests that there is a greater_number of_modes outside_the plates than inside [@Milonnibook]. In fact, this is not true:_there is only a rearrangement of_modes, for a nice_explanation of_this_problem see [@Hushwater]._For_the generalized_$\zeta$-function approach applied to the repulsive Casimir_effect for_parallel plates geometry see [@Andre; @SANTOS].
This_paper is organized as_follows:_in section [**2**]{} we determine the_photon field ${\mbox{\boldmath $A$}}({\mbox{\boldmath $r$}},t)$ in_the region between Boyer’s plates_making_use_of the Coulomb gauge. Next_we also evaluate the field operator_correlators $<{\hat E}_i{\hat_E}_j>_0$ and $<{\hat B}_i{\hat B}_j>_0$ with the_aid_of a simple but efficient regularization_prescription._In section [**3**]{} we apply our_results_to_reobtain the repulsive Casimir pressure_of this setup. In section [**4**]{}_we discuss the Scharnhorst effect but for this different_situation. In particular,_we show that, contrary to_the_case_with of the usual pair of |
’s outcome is then $\langle a \rangle=\cos\alpha\cos\theta_x$.
Concerning Charlie, his outcome $c$ is totally determined by $\vec \lambda$ and $\vec z$. For simplicity we assume that $0\leq\theta_z\leq\pi/2$, but the other situation can be treated similarly. In this case, for $\alpha\leq\frac\pi2-\theta_z$, Charles always has $c=+1$, and for $\alpha\geq\frac\pi2+\theta_z$, he always has $c=-1$. Now for $\alpha\in[\frac\pi2-\theta_z,\frac\pi2+\theta_z]$, we can write $\mathcal{I}_+=[-\Phi_c+\varphi_z,\Phi_c+\varphi_z]$ the interval of $\beta$ for which the product $\vec\lambda\cdot\vec z \geq 0$, and $\mathcal{I}_-=[\varphi_z-\pi,\varphi_z+\pi]\backslash \mathcal{I}_+$ its complement, with $\Phi_c=\arccos(-\cot\alpha\cot\theta_z)$. We thus have the three following cases:
1. $\alpha\leq\frac{\pi}2-\theta_z$. In this case $c=+1$, which gives $\langle ac\rangle=\cos\alpha\cos\theta_x$.
2. $\alpha\geq\frac{\pi}2+\theta_z$. In this case $c=-1$, which gives $\langle ac\rangle=-\cos\alpha\cos\theta_x$.
3. $\frac{\pi}2-\theta_z\leq\alpha\leq\frac{\pi}2+\theta_z$. In this case one has $\langle ac\rangle=\cos\alpha\cos\theta_x(\chi_\mathcal{I_+}(\beta)-\chi_\mathcal{I_-}(\beta))$ where $$\chi_\mathcal{I}(\beta)=\begin{cases}1&\text{if }\beta\in\mathcal{I}\\0&\text{if }\beta\notin\mathcal{I}\end{cases}$$ is the indicator function.
In total, after integration over $\alpha$ and $\beta$, this gives $\overline{\langle ac\rangle}=\frac12\cos\theta_x\cos\ | ’s outcome is then $ \langle a \rangle=\cos\alpha\cos\theta_x$.
Concerning Charlie, his outcome $ c$ is totally determine by $ \vec \lambda$ and $ \vec z$. For ease we assume that $ 0\leq\theta_z\leq\pi/2 $, but the other situation can be process similarly. In this case, for $ \alpha\leq\frac\pi2-\theta_z$, Charles always have $ c=+1 $, and for $ \alpha\geq\frac\pi2+\theta_z$, he always has $ c=-1$. Now for $ \alpha\in[\frac\pi2-\theta_z,\frac\pi2+\theta_z]$, we can write $ \mathcal{I}_+=[-\Phi_c+\varphi_z,\Phi_c+\varphi_z]$ the interval of $ \beta$ for which the intersection $ \vec\lambda\cdot\vec z \geq 0 $, and $ \mathcal{I}_-=[\varphi_z-\pi,\varphi_z+\pi]\backslash \mathcal{I}_+$ its complement, with $ \Phi_c=\arccos(-\cot\alpha\cot\theta_z)$. We thus have the three following case:
1. $ \alpha\leq\frac{\pi}2-\theta_z$. In this case $ c=+1 $, which gives $ \langle ac\rangle=\cos\alpha\cos\theta_x$.
2. $ \alpha\geq\frac{\pi}2+\theta_z$. In this case $ c=-1 $, which gives $ \langle ac\rangle=-\cos\alpha\cos\theta_x$.
3. $ \frac{\pi}2-\theta_z\leq\alpha\leq\frac{\pi}2+\theta_z$. In this case one has $ \langle ac\rangle=\cos\alpha\cos\theta_x(\chi_\mathcal{I_+}(\beta)-\chi_\mathcal{I_-}(\beta))$ where $ $ \chi_\mathcal{I}(\beta)=\begin{cases}1&\text{if } \beta\in\mathcal{I}\\0&\text{if } \beta\notin\mathcal{I}\end{cases}$$ is the index function.
In total, after integration over $ \alpha$ and $ \beta$, this yield $ \overline{\langle ac\rangle}=\frac12\cos\theta_x\cos\ | ’s oktcome is then $\langle a \vangle=\cos\alpha\cos\theta_x$.
Roncernjng Charuie, his outcome $c$ is totally dwtermuned by $\vec \lambda$ and $\vec z$. Fog simplicuty xe assume that $0\lxs\theta_z\leq\pi/2$, gmt thz ither situatiok can be traated similarlf. Kn this case, for $\alpha\leq\frac\pi2-\theta_z$, Charlex wlways has $c=+1$, agd fpw $\allha\geq\frac\pi2+\theta_z$, he always has $c=-1$. Now fog $\alpha\in[\frac\pi2-\thrta_z,\frac\pi2+\theta_z]$, we can wrlte $\lathcal{I}_+=[-\Phi_c+\varphi_x,\Phi_c+\varphi_e]$ thq interval of $\beta$ for which the prkduct $\vec\lambda\cdot\vec z \geq 0$, avd $\machcal{I}_-=[\varphu_z-\pi,\vwtphi_z+\pi]\backsoash \iathcal{I}_+$ its complemant, witn $\Phi_c=\arccos(-\cob\alphe\cot\rheta_z)$. We thus have tie three following cwses:
1. $\alpva\meq\frac{\pi}2-\theta_z$. Ib rhis wase $c=+1$, wfuch gibex $\mangle ac\cangle=\cos\allha\cos\theta_z$.
2. $\alpha\geq\frac{\pi}2+\thtta_s$. In this case $c=-1$, whish gives $\langle ac\rangle=-\cos\alpha\cos\theta_x$.
3. $\fdac{\pi}2-\theta_z\leq\alpha\leq\feac{\pi}2+\theta_z$. In this cwse one hws $\langle ac\rangle=\cos\alpha\cos\theta_x(\chi_\mathcal{I_+}(\betd)-\chi_\metfcao{I_-}(\neta))$ qhfre $$\chi_\mathcal{I}(\beta)=\begin{cases}1&\text{if }\beta\in\matrdak{I}\\0&\next{if }\beta\notin\mcthcal{I}\end{cases}$$ os tng indicator fuvction.
Nh fotal, after integrwtion oder $\aopha$ and $\feta$, this gives $\overline{\langle qc\rangle}=\frac12\bos\tyeta_x\cos\ | ’s outcome is then $\langle a \rangle=\cos\alpha\cos\theta_x$. his $c$ is determined by $\vec simplicity assume that $0\leq\theta_z\leq\pi/2$, the other situation be treated similarly. In this case, $\alpha\leq\frac\pi2-\theta_z$, Charles always has $c=+1$, and for $\alpha\geq\frac\pi2+\theta_z$, he always has $c=-1$. Now $\alpha\in[\frac\pi2-\theta_z,\frac\pi2+\theta_z]$, we can write $\mathcal{I}_+=[-\Phi_c+\varphi_z,\Phi_c+\varphi_z]$ the interval of $\beta$ for which the product z 0$, $\mathcal{I}_-=[\varphi_z-\pi,\varphi_z+\pi]\backslash its complement, with $\Phi_c=\arccos(-\cot\alpha\cot\theta_z)$. We thus have the three following cases: 1. $\alpha\leq\frac{\pi}2-\theta_z$. In this case which gives $\langle ac\rangle=\cos\alpha\cos\theta_x$. 2. $\alpha\geq\frac{\pi}2+\theta_z$. In this $c=-1$, which gives $\langle 3. $\frac{\pi}2-\theta_z\leq\alpha\leq\frac{\pi}2+\theta_z$. In this case has ac\rangle=\cos\alpha\cos\theta_x(\chi_\mathcal{I_+}(\beta)-\chi_\mathcal{I_-}(\beta))$ where }\beta\in\mathcal{I}\\0&\text{if is indicator function. In after integration over $\alpha$ and $\beta$, this gives $\overline{\langle ac\rangle}=\frac12\cos\theta_x\cos\ | ’s outcome is then $\langle a \rangLe=\cos\alpha\Cos\thEta_X$.
CoNcErniNg ChArlie, his outcomE $C$ is tOtally determined by $\vec \lAmbda$ AnD $\Vec z$. fOr SimplIcity we ASsUME thAt $0\LeQ\thEtA_Z\lEq\pi/2$, bUt tHe other Situation cAn bE tReated similaRLy. in this case, For $\Alpha\leq\frac\Pi2-\tHeta_z$, CHaRleS AlwayS haS $c=+1$, and For $\alpHA\geq\frAc\pi2+\theta_Z$, hE Always HAs $c=-1$. Now fOR $\AlPha\iN[\frac\pi2-\theta_z,\frac\PI2+\tHEta_z]$, we can write $\MathcaL{I}_+=[-\pHi_C+\VArpHi_z,\phi_c+\varphi_Z]$ tHe intERval of $\bETa$ FOR WhiCH the product $\veC\lambda\cdot\VEc z \Geq 0$, and $\MaThcAL{I}_-=[\varpHi_z-\pi,\VaRPhi_Z+\pi]\backslasH \matHcal{I}_+$ its cOmplemENt, with $\PHI_c=\arccoS(-\cot\alPha\Cot\ThetA_Z)$. WE tHus HaVE thE ThRee FOllOwing casEs:
1. $\AlPha\leQ\fraC{\PI}2-\THeta_Z$. In This Case $c=+1$, Which gives $\lanGle Ac\raNGle=\Cos\alPha\coS\theTa_X$.
2. $\alphA\geq\frAc{\pi}2+\tHeTa_z$. In this case $c=-1$, wHich Gives $\langLe aC\rAngLe=-\Cos\alPHa\cos\tHetA_x$.
3. $\fRac{\pi}2-\thEta_z\leq\ALphA\lEQ\FRaC{\pi}2+\theta_z$. In this casE oNE HaS $\langle aC\ranglE=\CoS\aLPha\cos\thEtA_x(\cHi_\maTHCal{I_+}(\bEta)-\cHI_\mAthcal{I_-}(\bEta))$ wheRE $$\cHi_\Mathcal{i}(\bEta)=\begIn{CasEs}1&\tExt{if }\BEta\iN\mathcAl{I}\\0&\text{iF }\beta\NOtin\mathcal{I}\enD{Cases}$$ is the indICaTOR fUNctiOn.
IN total, after InteGRatiOn ovER $\aLphA$ And $\beTa$, thiS gIVeS $\Overline{\langle ac\ranGlE}=\frac12\cOs\theTa_x\cos\ | ’s outcome is then $\langl e a \rangl e=\co s\a lph a\ cos\ thet a_x$.
Concern i ng C harlie, his outcome $c $ isto t ally de termi ned by$ \v e c \l am bd a$an d $ \vecz$. For si mplicity w e a ss ume that $0\ l eq \theta_z\l eq\ pi/2$, but t heothersi tua t ion c anbe tr eateds imilar ly. In th is case,f or $\al p h a\ leq\ frac\pi2-\theta_z $ ,C harles alwayshas $c =+ 1 $, a ndfor $\alpha\g eq \frac \ pi2+\th e ta _ z $ , h e always has $ c=-1$. Nowf or$\alph a\ in[ \ frac\p i2-\t he t a_z ,\frac\pi2+ \the ta_z]$, w e canw rite $\ m athcal{ I}_+=[ -\P hi_ c+\v a rp hi _z, \P h i_c + \v arp h i_z ]$ the i nt er val o f $\ b e t a $ fo r w hich theproduct $\vec \la mbda \ cdo t\vec z \g eq 0 $, and$\math cal{I }_ -=[\varphi_z-\p i,\v arphi_z+\ pi] \b ack sl ash \ m athcal {I} _+$ its co mplemen t , w it h $ \P hi_c=\arccos(-\cot \a l p ha \cot\the ta_z)$ . W et hus have t hethre e follo wing ca ses:
1. $\al p ha \l eq\frac {\ pi}2-\ th eta _z$ . Int hiscase $ c=+1$, w hichg ives $\langlea c\rangle=\cos \ al p h a\ c os\t het a_x$.
2. $\al p ha\g eq\f r ac {\p i }2+\t heta_ z$ . I n this case $c=-1$,wh ich gi ves $ \langle ac\ra ngle=-\cos \ a l pha\cos\ thet a _x $ .
3. $\frac{ \pi}2 -\theta_z\ l eq\alpha \leq\ frac{\pi }2+\theta _ z $. In th iscas e o neh a s$\langle ac\r a n gle= \c os\alph a\c os\thet a_x (\c hi_ \ma th cal{I_+}( \beta)-\ ch i_ \m at hca l{I_- } (\beta)) $whe re $$ \chi_ \ mathca l{I}( \bet a) =\ b egi n{cases } 1& \ t ext{ if } \bet a\i n\ mathc al{I } \\0 &\text{ if }\beta \no t in\m at hc al{I}\e nd{cases}$$ i sthe indica to r f unctio n .
In tot al, after integration o v er $\al pha $ and $\b eta$, thi s g ives $ \ov e rline{ \langl e ac\ ra ngl e } =\fra c 1 2\ cos \t heta_x\cos \ | ’s outcome_is then_$\langle a \rangle=\cos\alpha\cos\theta_x$.
Concerning Charlie,_his outcome_$c$_is totally_determined_by $\vec \lambda$_and $\vec z$._For simplicity we assume_that $0\leq\theta_z\leq\pi/2$, but_the_other situation can be treated similarly. In this case, for $\alpha\leq\frac\pi2-\theta_z$, Charles always has_$c=+1$,_and for_$\alpha\geq\frac\pi2+\theta_z$,_he_always has $c=-1$. Now for_$\alpha\in[\frac\pi2-\theta_z,\frac\pi2+\theta_z]$, we can write $\mathcal{I}_+=[-\Phi_c+\varphi_z,\Phi_c+\varphi_z]$_the interval_of $\beta$ for which the product $\vec\lambda\cdot\vec z_\geq_0$, and $\mathcal{I}_-=[\varphi_z-\pi,\varphi_z+\pi]\backslash_\mathcal{I}_+$ its complement, with $\Phi_c=\arccos(-\cot\alpha\cot\theta_z)$. We thus have the_three following cases:
1. $\alpha\leq\frac{\pi}2-\theta_z$. In_this case $c=+1$,_which_gives_$\langle ac\rangle=\cos\alpha\cos\theta_x$.
2. $\alpha\geq\frac{\pi}2+\theta_z$._In this case $c=-1$, which gives_$\langle ac\rangle=-\cos\alpha\cos\theta_x$.
3. $\frac{\pi}2-\theta_z\leq\alpha\leq\frac{\pi}2+\theta_z$. In this_case one has $\langle ac\rangle=\cos\alpha\cos\theta_x(\chi_\mathcal{I_+}(\beta)-\chi_\mathcal{I_-}(\beta))$ where $$\chi_\mathcal{I}(\beta)=\begin{cases}1&\text{if_}\beta\in\mathcal{I}\\0&\text{if }\beta\notin\mathcal{I}\end{cases}$$ is the indicator function.
In_total, after integration over $\alpha$_and $\beta$,_this gives $\overline{\langle ac\rangle}=\frac12\cos\theta_x\cos\ |
the species $\ct{k}$ of $k$-coding trees first, then use dissymmetry to apply the results to unrooted enumeration.
\[thm:funcdecompct\] The species $\ctx{k}$ of $X$-rooted $k$-coding trees, $\cty{k}$ of $Y$-rooted $k$-coding trees, and $\ctxy{k}$ of edge-rooted $k$-coding trees satisfy the functional equations
\[eq:ctfunc\] $$\begin{aligned}
\ctx{k} &= X \cdot \specname{C}_{k+1} \pbrac[\big]{\cty{k}} \label{eq:ctxfunc} \\
\cty{k} &= Y \cdot \specname{E} \pbrac*{X \cdot \specname{L}_{k} \pbrac[\big]{\cty{k}}} \label{eq:ctyfunc} \\
\ctxy{k} &= \cty{k} \cdot X \cdot \specname{L}_{k} \pbrac[\big]{\cty{k}} = X \cdot \specname{L}_{k+1} \pbrac[\big]{\cty{k}} \label{eq:ctxyfunc}
\end{aligned}$$
as isomorphisms of species.
By construction, a $\ctx{k}$-structure consists of a single $X$-label and a cyclically-ordered $\pbrac{k+1}$-set of $\cty{k}$-structures. This gives \[eq:ctxfunc\]. See \[fig:ctxconst\] for an example of this construction.
(root) at (0,0) [$X$]{}; ;
(180/:1) node [$\specname{C}_{\numfronts}$]{};
iin [0,..., ]{}
(childi) at (:3) [$Y$]{}; (root) – (childi); (childi) ++(+90:1) node [$\cty{\kval}$]{};
(childi) ++(:1cm) ++(180+-:2cm) arc (180+-:180++:2cm);
(childi) – ++(:2) node \[rotate=,fill=white\] [$\cdots$]{}; (childi) – ++(+:2); | the species $ \ct{k}$ of $ k$-coding trees first, then use asymmetry to lend oneself the results to unrooted enumeration.
\[thm: funcdecompct\ ] The coinage $ \ctx{k}$ of $ X$-rooted $ k$-coding trees, $ \cty{k}$ of $ Y$-rooted $ k$-coding tree, and $ \ctxy{k}$ of edge - settle $ k$-coding trees satisfy the running equation
\[eq: ctfunc\ ] $ $ \begin{aligned }
\ctx{k } & = X \cdot \specname{C}_{k+1 } \pbrac[\big]{\cty{k } } \label{eq: ctxfunc } \\
\cty{k } & = Y \cdot \specname{E } \pbrac*{X \cdot \specname{L}_{k } \pbrac[\big]{\cty{k } } } \label{eq: ctyfunc } \\
\ctxy{k } & = \cty{k } \cdot X \cdot \specname{L}_{k } \pbrac[\big]{\cty{k } } = X \cdot \specname{L}_{k+1 } \pbrac[\big]{\cty{k } } \label{eq: ctxyfunc }
\end{aligned}$$
as isomorphisms of coinage.
By construction, a $ \ctx{k}$-structure consists of a single $ X$-label and a cyclically - ordered $ \pbrac{k+1}$-set of $ \cty{k}$-structures. This gives \[eq: ctxfunc\ ]. determine \[fig: ctxconst\ ] for an example of this construction.
(root) at (0,0) [ $ X$ ] { }; ;
(180/:1) node [ $ \specname{C}_{\numfronts}$ ] { };
iin [ 0, ..., ] { }
(childi) at (:3) [ $ Y$ ] { }; (root) – (childi); (childi) + + (+90:1) node [ $ \cty{\kval}$ ] { };
(childi) + + (:1 cm) + + (180 + -:2 cm) discharge (180 + -:180++:2 cm);
(childi) – + + (: 2) node \[rotate=,fill = white\ ] [ $ \cdots$ ] { }; (childi) – + + (+: 2); | thf species $\ct{k}$ of $k$-coding trees first, thgn use dmssymmefry to aoply the results to unrooted ebumerqtion.
\[thm:funcdecompct\] Tfe specied $\ctx{k}$ od $X$-riited $k$-codiif trees, $\cty{k}$ kn $Y$-romved $k$-coding tregs, and $\ctxy{k}$ of edge-rooted $k$-zobing trees satisfy the functional eqtations
\[rq:ftfunc\] $$\begin{alygnec}
\ctx{k} &= X \cdot \specname{C}_{k+1} \pbrac[\big]{\dty{k}} \lauel{eq:ctxfunc} \\
\cty{k} &= Y \cdot \specname{E} \pbrwc*{X \cdot \specname{L}_{n} \pbrac[\big]{\cjg{k}}} \jqbel{eq:ctyfunz} \\
\ctqv{k} &= \cty{k} \cdkt X \cdot \specname{L}_{k} \pbrac[\big]{\cth{k}} = X \cdot \specbane{L}_{n+1} \pbrac[\big]{\ctb{k}} \labvl{eq:ctxyfunc}
\end{alhgned}$$
as isomorphisms pf vpexies.
By construction, a $\ctx{k}$-structure confists of d aingle $X$-label and a cyclhcalny-orawrea $\pgrec{k+1}$-aet of $\ctb{k}$-structurea. This givew \[eq:ctxfunc\]. See \[fig:vtvbpnst\] for an sxamplq jf this construction.
(root) at (0,0) [$X$]{}; ;
(180/:1) node [$\sptcnams{C}_{\numfronts}$]{};
iin [0,..., ]{}
(childi) qt (:3) [$Y$]{}; (root) – (childi); (chlldi) ++(+90:1) nodq [$\cty{\kval}$]{};
(childi) ++(:1cm) ++(180+-:2cm) arc (180+-:180++:2cm);
(childi) – ++(:2) node \[rotate=,fhll=whmtd\] [$\cbits$]{}; (cfulfi) – ++(+:2); | the species $\ct{k}$ of $k$-coding trees first, dissymmetry apply the to unrooted enumeration. $X$-rooted trees, $\cty{k}$ of $k$-coding trees, and of edge-rooted $k$-coding trees satisfy the equations \[eq:ctfunc\] $$\begin{aligned} \ctx{k} &= X \cdot \specname{C}_{k+1} \pbrac[\big]{\cty{k}} \label{eq:ctxfunc} \\ \cty{k} &= \cdot \specname{E} \pbrac*{X \cdot \specname{L}_{k} \pbrac[\big]{\cty{k}}} \label{eq:ctyfunc} \\ \ctxy{k} &= \cty{k} \cdot X \specname{L}_{k} = \cdot \pbrac[\big]{\cty{k}} \label{eq:ctxyfunc} \end{aligned}$$ as isomorphisms of species. By construction, a $\ctx{k}$-structure consists of a single $X$-label a cyclically-ordered $\pbrac{k+1}$-set of $\cty{k}$-structures. This gives \[eq:ctxfunc\]. \[fig:ctxconst\] for an example this construction. (root) at (0,0) ; node [$\specname{C}_{\numfronts}$]{}; [0,..., (childi) (:3) [$Y$]{}; (root) (childi); (childi) ++(+90:1) node [$\cty{\kval}$]{}; (childi) ++(:1cm) ++(180+-:2cm) arc (180+-:180++:2cm); (childi) – ++(:2) node \[rotate=,fill=white\] [$\cdots$]{}; (childi) ++(+:2); | the species $\ct{k}$ of $k$-coding treEs first, theN use dIssYmmEtRy to ApplY the results to uNRootEd enumeration.
\[thm:funcdeCompcT\] THE speCIeS $\ctx{k}$ Of $X$-rootED $k$-CODinG tReEs, $\cTy{K}$ Of $y$-rootEd $k$-Coding tRees, and $\ctxY{k}$ oF eDge-rooted $k$-coDInG trees satiSfy The functionaL eqUationS
\[eQ:ctFUnc\] $$\beGin{AlignEd}
\ctx{k} &= x \Cdot \spEcname{C}_{k+1} \pBrAC[\big]{\ctY{K}} \label{eQ:CTxFunc} \\
\Cty{k} &= Y \cdot \specname{e} \PbRAc*{X \cdot \specnamE{L}_{k} \pbrAc[\BIg]{\CTY{k}}} \lAbeL{eq:ctyfunc} \\
\CtXy{k} &= \ctY{K} \cdot X \cDOt \SPECnaME{L}_{k} \pbrac[\big]{\ctY{k}} = X \cdot \specNAme{l}_{k+1} \pbraC[\bIg]{\cTY{k}} \labeL{eq:ctXyFUnc}
\End{aligned}$$
aS isoMorphisms Of specIEs.
By conSTructioN, a $\ctx{k}$-StrUctUre cONsIsTs oF a SIngLE $X$-LabEL anD a cyclicAlLy-OrderEd $\pbRAC{K+1}$-Set oF $\ctY{k}$-stRuctuRes. This gives \[eQ:ctXfunC\]. see \[Fig:ctXconsT\] for An ExampLe of thIs conStRuction.
(root) at (0,0) [$X$]{}; ;
(180/:1) nOde [$\sPecname{C}_{\nUmfRoNts}$]{};
IiN [0,..., ]{}
(chilDI) at (:3) [$Y$]{}; (roOt) – (cHilDi); (childI) ++(+90:1) node [$\ctY{\KvaL}$]{};
(cHILDi) ++(:1Cm) ++(180+-:2cm) arc (180+-:180++:2cm);
(childi) – ++(:2) nodE \[rOTAtE=,fill=whiTe\] [$\cdotS$]{}; (ChIlDI) – ++(+:2); | the species $\ct{k}$ of $ k$-codingtrees fi rst ,then use dissymmetry t o app ly the results to unro oteden u mera t io n.
\ [thm:fu n cd e c omp ct \] Th es pe cies$\c tx{k}$of $X$-roo ted $ k$-coding tr e es , $\cty{k} $ o f $Y$-rooted $k $-codi ng tr e es, a nd$\ctx y{k}$o f edge -rooted $ k$ - coding trees s a t is fy t he functional equ a ti o ns
\[eq:ctfun c\] $$ \b e gi n { ali gne d}
\ ct x{k}& = X \cd o t\ s p ecn a me{C}_{k+1} \ pbrac[\big] { \ct y{k}}\l abe l {eq:ct xfunc }\ \
\cty{k } &= Y \cdot\specn a me{E} \ p brac*{X \cdot \s pec name { L} _{ k}\p b rac [ \b ig] { \ct y{k}}} \ la be l{eq: ctyf u n c } \\ \ ctxy{ k} &= \cty{k} \c dotX \c dot \ specn ame{ L} _{k}\pbrac [\big ]{ \cty{k}} = X \c dot\specname {L} _{ k+1 }\pbra c [\big] {\c ty{ k}} \la bel{eq: c txy fu n c }
\end{aligned}$$
a s i somorphi sms of sp ec i es.
Byco nst ruct i o n, a$\ct x {k }$-struc ture c o ns is ts of a s ingle$X $-l abe l and a cy clical ly-order ed $\ p brac{k+1}$-set of $\cty{k}$- s tr u c tu r es.Thi s gives \[e q:ct x func \].S ee \[ f ig:ct xcons t\ ] f o r an example of thi sconstr uctio n.
(root) at (0,0) [$X $ ] { }; ;
(1 80/: 1 )n ode [$\specnam e{C}_ {\numfront s }$]{};
iin [ 0,..., ] {}
(chil d i ) at (:3 ) [ $Y$ ]{} ; ( r o ot ) – (childi); ( chil di ) ++(+9 0:1 ) node[$\ cty {\k val }$ ]{};
(ch ildi) ++ (: 1c m) + +(1 80+-: 2 cm) arc(1 80+ -: 180 ++:2c m );
(c hildi ) –++ (: 2 ) n ode \[r o ta t e =,fi ll =w hite \][$ \cdot s$]{ } ; ( childi) – ++(+:2 ); | the_species $\ct{k}$_of $k$-coding trees first,_then use_dissymmetry_to apply_the_results to unrooted_enumeration.
\[thm:funcdecompct\] The species_$\ctx{k}$ of $X$-rooted $k$-coding_trees, $\cty{k}$ of_$Y$-rooted_$k$-coding trees, and $\ctxy{k}$ of edge-rooted $k$-coding trees satisfy the functional equations
\[eq:ctfunc\] $$\begin{aligned}
__ __\ctx{k}_&= X \cdot \specname{C}_{k+1} \pbrac[\big]{\cty{k}}_\label{eq:ctxfunc} \\
_ _\cty{k} &= Y \cdot \specname{E} \pbrac*{X \cdot \specname{L}_{k}_\pbrac[\big]{\cty{k}}}_\label{eq:ctyfunc} \\
_ \ctxy{k} &= \cty{k} \cdot X_\cdot \specname{L}_{k} \pbrac[\big]{\cty{k}} = X \cdot_\specname{L}_{k+1} \pbrac[\big]{\cty{k}} \label{eq:ctxyfunc}
___ \end{aligned}$$
as isomorphisms of_species.
By construction, a $\ctx{k}$-structure consists of_a single $X$-label and a cyclically-ordered_$\pbrac{k+1}$-set of $\cty{k}$-structures. This gives \[eq:ctxfunc\]. See_\[fig:ctxconst\] for an example of this_construction.
(root) at (0,0) [$X$]{}; ;
(180/:1)_node [$\specname{C}_{\numfronts}$]{};
iin_[0,..., ]{}
(childi) at (:3) [$Y$]{};_(root) – (childi);_(childi) ++(+90:1)_node [$\cty{\kval}$]{};
(childi) ++(:1cm)_++(180+-:2cm) arc (180+-:180++:2cm);
(childi) – ++(:2) node_\[rotate=,fill=white\] [$\cdots$]{}; (childi)_– ++(+:2); |
-branes on internal cycles. Asymptotically $AdS_4$ black holes with more general transverse space can be found in [@Donos:2008ug] and [@Donos2012d] where the solutions were studied directly in M-theory. These include the M-theory lift of the solutions we give in Sections \[sec:Q111Simp\] and \[numericalQ111\].
The BPS black holes we construct in this paper are asymptotically $AdS_4$ and as such they are states in particular (deformed) three-dimensional superconformal field theories on $S^2\times \mathbb{R}$. The solution in [@Cacciatori:2009iz] can be considered as a state in the twisted ABJM theory [@Aharony:2008ug]. The solutions we have found in this paper can be seen as states in (twisted and deformed) three dimensional Chern-Simons matter theory dual to the M-theory compactifications of homogeneous Sasaki-Einstein manifolds[^3]. One feature of these theories compared to ABJM is the presence of many baryonic symmetries that couple to the vector multiplets arising from non trivial two-cycles in the Sasaki-Einstein manifold. In terms of the worldvolume theory, the black holes considered in this paper are then electrically charged states of a Chern-Simons matter theory in a monopole background for $U(1)_R$ symmetry and other global symmetries, including the baryonic ones[^4].
Gauged $\N=2$ supergravity with hypermultiplets is the generic low-energy theory arising from a Kaluza-Klein reduction of string/M-theory on a flux background. The hypermultiplet scalars interact with the vector-multiplet scalars through the scalar potential: around a generic $AdS_4$ vacuum the eigenmodes mix the hypers and vectors. In the models we study, we employ a particular simplification on the hypermultiplet scalar manifold and find solutions where only one real hypermultiplet scalar has a non-trivial profile. Given that the simplification is so severe it is quite a triumph that solutions exist within this ansatz. It would be interesting to understand if this represents a general feature of black holes in gauged supergravity.\
The paper is organized as follows. In Section 2 we summarize the ansatz we use and the resulting BPS equations for an arbitrary electrically gauged $\N=2$ supergravity theory. The restriction | -branes on internal cycles. Asymptotically $ AdS_4 $ black holes with more cosmopolitan cross outer space can be found in [ @Donos:2008ug ] and [ @Donos2012d ] where the solutions were study directly in M - hypothesis. These include the M - theory lift of the solution we give in Sections \[sec: Q111Simp\ ] and \[numericalQ111\ ].
The BPS black hole we construct in this paper are asymptotically $ AdS_4 $ and as such they are state in especial (deformed) three - dimensional superconformal field theories on $ S^2\times \mathbb{R}$. The solution in [ @Cacciatori:2009iz ] can be considered as a state in the distorted ABJM theory [ @Aharony:2008ug ]. The solutions we have found in this newspaper can be seen as states in (distorted and deformed) three dimensional Chern - Simons topic theory dual to the M - theory compactifications of homogeneous Sasaki - Einstein manifolds[^3 ]. One feature of these theories compared to ABJM is the bearing of many baryonic symmetries that couple to the vector multiplets arising from non trivial two - cycles in the Sasaki - Einstein manifold. In terms of the worldvolume hypothesis, the bootleg trap considered in this paper are then electrically charged states of a Chern - Simons matter theory in a monopole background for $ U(1)_R$ symmetry and early ball-shaped symmetries, including the baryonic ones[^4 ].
Gauged $ \N=2 $ supergravity with hypermultiplets is the generic low - department of energy hypothesis arising from a Kaluza - Klein reduction of string / M - theory on a flux background. The hypermultiplet scalars interact with the vector - multiplet scalars through the scalar potential: around a generic $ AdS_4 $ vacuum the eigenmodes shuffle the hypers and vectors. In the models we study, we employ a particular simplification on the hypermultiplet scalar manifold and witness solutions where only one real hypermultiplet scalar has a non - trivial profile. establish that the simplification is so severe it is quite a triumph that solutions exist within this ansatz. It would be interesting to sympathize if this represents a general feature of black holes in gauged supergravity.\
The paper is organized as follows. In department 2 we summarize the ansatz we use and the result BPS equations for an arbitrary electrically estimate $ \N=2 $ supergravity theory. The limitation | -brajes on internal cycles. Arymptotically $AbW_4$ blacn holea with mure general transverse space cqn be found in [@Donos:2008ug] and [@Aonos2012d] whvre the silutmons were studiev directly in J-bheorv. Vhese include tme M-theory nift of the sonugilns we give in Sections \[sec:Q111Simp\] anq \[numerocwlQ111\].
The BPS blask hpjes sv gonstruct in this paper are asymltoticanly $AdS_4$ and ax such they are states in oartlcular (deformed) thgee-dimensiobal fyperconformau field thtoxies on $S^2\tijes \mathbb{R}$. The solution in [@Cacziatoxi:2009iz] can be cinslgered as a wtate in the twisbvd ABJM theory [@Aharony:2008ug]. The sonutuons we have found in this paper can be seen as vtctes in (twisted and dwfirmed) thrae dknenriohak Dhern-Slmois matter tgeory dual ro the M-theory compscepgications of homogqnqous Sasaki-Einstein manifolds[^3]. One featuge or these theories comparwd to ABJM is the predence of iany baryonic symmetries that couple to the vectos mulvioleus ariskbg from non trivial two-cycles in the Sasaki-Einsesim kanifold. In tevms of the worldvokule jheory, the blazk holza donsidered in this paper wre tyen electwicakly charged states of a Cheen-Simons matner rheory in a monopope backgrouud for $U(1)_R$ summetry and other globau syjmetries, influding tgd baryonic ones[^4].
Gxugvd $\N=2$ supergravity with hypermujtiplets ms thz generiz loe-energr theory agisinn from a Kaluza-Kleij redoction of string/L-theory on a flux background. Thx hypermultipked sbalars inceract with the veceor-multiplet sealars tkrough the scalag potentiel: around a deneric $AdS_4$ vdfuum the eigxnmodes myx tye htpers ava vectors. In tne models we study, qe employ a particmlar rjmplification ou uhw hypermultiplrt rcajag menifojg and find smlutkonr wherd only one veau hylermultiplet scalar vas z non-trivial profike. Given tyat the fimplificatiom is so severe it ls qumte a vriumpn trat solutions exist within thia ansatz. Lt rould be intqresbing to underscand if this represents a general featurx of black holes in gaoged supergravity.\
The kaptr is organizxd as sollows. It Section 2 we summaruze the ansatz we use and the resultinf BPS aquatlons for an arbitrary electrically gauged $\N=2$ supergravity theory. The restruction | -branes on internal cycles. Asymptotically $AdS_4$ black more transverse space be found in solutions studied directly in These include the lift of the solutions we give Sections \[sec:Q111Simp\] and \[numericalQ111\]. The BPS black holes we construct in this paper asymptotically $AdS_4$ and as such they are states in particular (deformed) three-dimensional superconformal theories $S^2\times The in [@Cacciatori:2009iz] can be considered as a state in the twisted ABJM theory [@Aharony:2008ug]. The solutions have found in this paper can be seen states in (twisted and three dimensional Chern-Simons matter theory to M-theory compactifications homogeneous manifolds[^3]. feature of these compared to ABJM is the presence of many baryonic symmetries that couple to the vector multiplets arising non trivial the Sasaki-Einstein In of worldvolume theory, the considered in this paper are then of a Chern-Simons matter theory in a monopole for $U(1)_R$ and other global symmetries, including the ones[^4]. Gauged $\N=2$ supergravity with hypermultiplets is the low-energy theory arising from a Kaluza-Klein reduction of string/M-theory on a flux background. The hypermultiplet with the vector-multiplet scalars the scalar potential: a $AdS_4$ the mix the and vectors. In the models we study, we employ a particular on the hypermultiplet scalar manifold and find solutions where only hypermultiplet has a non-trivial Given that the simplification so it is quite a solutions within would interesting understand if this represents general feature of black holes gauged supergravity.\ The paper Section 2 we summarize the ansatz we use the resulting BPS equations for an arbitrary gauged $\N=2$ supergravity theory. The restriction | -branes on internal cycles. AsyMptoticallY $AdS_4$ bLacK hoLeS witH morE general transvERse sPace can be found in [@Donos:2008uG] and [@DOnOS2012d] whERe The soLutions WErE STudIeD dIreCtLY iN M-theOry. these inClude the M-tHeoRy Lift of the solUTiOns we give iN SeCtions \[sec:Q111SiMp\] aNd \[numeRiCalq111\].
the BPs blAck hoLes we cONstrucT in this paPeR Are asyMPtoticaLLY $ADS_4$ anD as such they are staTEs IN particular (defOrmed) tHrEE-dIMEnsIonAl superconFoRmal fIEld theoRIeS ON $s^2\tiMEs \mathbb{R}$. The sOlution in [@CaCCiaTori:2009iz] CaN be COnsideRed as A sTAte In the twisteD ABJm theory [@AhArony:2008uG]. the soluTIons we hAve fouNd iN thIs paPEr CaN be SeEN as STaTes IN (twIsted and DeFoRmed) tHree DIMENsioNal cherN-SimoNs matter theorY duAl to THe M-TheorY compActiFiCatioNs of hoMogenEoUs Sasaki-EinsteiN manIfolds[^3]. One FeaTuRe oF tHese tHEories ComParEd to ABJm is the pREseNcE OF MaNy baryonic symmetriEs THAt Couple to The vecTOr MuLTiplets aRiSinG froM NOn triVial TWo-Cycles in The SasAKi-eiNstein mAnIfold. IN tErmS of The woRLdvoLume thEory, the bLack hOLes considered iN This paper are tHEn ELEcTRicaLly Charged statEs of A cherN-SimONs MatTEr theOry in A mONoPOle background for $U(1)_R$ sYmMetry aNd othEr global symmeTries, incluDING the baryOnic ONeS[^4].
gauged $\N=2$ supergrAvity With hypermULtiplets Is the Generic lOw-energy tHEOry arisiNg fRom A KaLuzA-kLeIn reduction of STRing/m-tHeory on A flUx backgRouNd. THe hYpeRmUltiplet sCalars inTeRaCt WiTh tHe vecTOr-multipLeT scAlArs ThrouGH the scAlar pOtenTiAl: ARouNd a geneRIc $aDs_4$ vacUuM tHe eiGenMoDes miX the HYpeRs and veCtors. In thE moDEls wE sTuDy, we empLoy a particulaR sImplificatIoN on The hypERMultipleT scalar manifold and find sOLutions WheRe onlY one Real hyperMulTiplet ScaLAr has a Non-triVial pRoFilE. gIven tHAT tHe sImPlificatioN IS so SeverE iT is qUite a trIumph that solutions EXisT within this anSatZ. It wOULd Be iNTeREstInG To uNDErstand if this rePresents a gEnERaL feature of BLacK hOles in gAuged suPergrAVity.\
The Paper is orGanized as FoLlowS. iN SeCtion 2 we sumMarize thE ansatz we USe and THe ResulTinG BPS eqUaTioNs for An arbiTRarY elecTricalLy Gauged $\n=2$ supeRgRavity thEory. The restriction | -branes on internal cycles . Asymptot icall y $ AdS _4 $ bl ackholes with mor e gen eral transverse spacecan b ef ound in [@Do nos:200 8 ug ] and [ @D ono s2 0 12 d] wh ere the so lutions we rest udied direct l yin M-theor y.These includ e t he M-t he ory liftofthe s olutio n s we g ive in Se ct i ons \[ s ec:Q111 S i mp \] a nd \[numericalQ11 1 \] .
The BPS blac k hole sw ec o nst ruc t in thispa per a r e asymp t ot i c a lly $AdS_4$ and a s such they are state sinp articu lar ( de f orm ed) three-d imen sional su percon f ormal f i eld the orieson$S^ 2\ti m es \ mat hb b {R} $ .The sol ution in [ @C accia tori : 2 0 0 9iz] ca n be cons idered as a s tat e in the twis ted A BJMth eory[@Ahar ony:2 00 8ug]. The solut ions we havefou nd in t his p a per ca n b e s een asstatesi n ( tw i s t ed and deformed) thr ee d im ensional Chern - Si mo n s matter t heo ry d u a l totheM -t heory co mpacti f ic at ions of h omogen eo usSas aki-E i nste in man ifolds[^ 3]. O n e feature of t h ese theoriesc om p a re d toABJ M is the pr esen c e of man y b ary o nic s ymmet ri e st hat couple to the v ec tor mu ltipl ets arising f rom non tr i v i al two-c ycle s i n the Sasaki-Ei nstei n manifold . In term s ofthe worl dvolume t h e ory, the bl ack ho les c on sidered in th i s pap er are th enelectri cal lycha rge dstates of a Chern -S im on smat ter t h eory inamon op ole back g roundfor $ U(1) _R $s ymm etry an d o t h er g lo ba l sy mme tr ies,incl u din g the b aryonic o nes [ ^4].
Ga uged $\ N=2$ supergra vi ty with hy pe rmu ltiple t s is thegeneric low-energy theo r y arisi ngfroma Ka luza-Klei n r educti ono f stri ng/M-t heory o n a f lux b a c kg rou nd . The hype r m ult iplet s cala rs inte ract with the vect o r-m ultiplet scal ars thr o u gh th e s c ala rp ote n t ial: around a g eneric $Ad S_ 4 $vacuum the eig en modes m ix thehyper s and ve ctors. In the mode ls wes t udy , we emplo y a part icular si m plifi c at ion o n t he hyp er mul tiple t scal a r m anifo ld and f ind so lutio ns where o nly one real hypermulti plet s calar ha s a non-t riv i alprofile.Give n that the si mpl ifica tio n is s o se v er e i t is q uite a triumph th ats o lu tions exist w i thi n thi s a n satz.It w ould be interesti n g to understan d if t his re p rese nt s a general fe atu re o f blackho les in gaug ed super gr a vity. \
Thepaperis orga n i ze d as fo llow s.In Sectio n 2 w e summar iz et he ans atzwe use a nd the resu l t ing BPS equation s for a n arb i tra ry el ec tricall y gau ged $\N=2$ supergravi ty the ory. Therestric ti on | -branes on_internal cycles._Asymptotically $AdS_4$ black holes_with more_general_transverse space_can_be found in_[@Donos:2008ug] and [@Donos2012d]_where the solutions were_studied directly in_M-theory._These include the M-theory lift of the solutions we give in Sections \[sec:Q111Simp\] and_\[numericalQ111\].
The_BPS black_holes_we_construct in this paper are_asymptotically $AdS_4$ and as such_they are_states in particular (deformed) three-dimensional superconformal field theories_on_$S^2\times \mathbb{R}$. The_solution in [@Cacciatori:2009iz] can be considered as a state_in the twisted ABJM theory [@Aharony:2008ug]._The solutions we_have_found_in this paper can_be seen as states in (twisted_and deformed) three dimensional Chern-Simons matter_theory dual to the M-theory compactifications of_homogeneous Sasaki-Einstein manifolds[^3]. One feature of_these theories compared to ABJM_is the_presence of many baryonic symmetries_that couple to_the vector_multiplets arising from_non trivial two-cycles in the Sasaki-Einstein_manifold. In terms_of the worldvolume theory, the black_holes_considered in this_paper_are_then electrically_charged states of_a_Chern-Simons matter_theory_in a monopole background for $U(1)_R$_symmetry_and other global symmetries, including the baryonic_ones[^4].
Gauged $\N=2$ supergravity with_hypermultiplets_is the generic low-energy_theory arising from a Kaluza-Klein_reduction of string/M-theory on a flux_background. The_hypermultiplet scalars_interact with the vector-multiplet scalars through the scalar potential: around a_generic $AdS_4$ vacuum the eigenmodes mix_the hypers and vectors._In the_models_we study, we_employ_a particular_simplification on the hypermultiplet scalar manifold and_find solutions_where only one real hypermultiplet scalar_has a non-trivial profile._Given_that the simplification is so severe_it is quite a triumph that_solutions exist within this ansatz._It_would_be interesting to understand if_this represents a general feature of_black holes in_gauged supergravity.\
The paper is organized as follows._In_Section 2 we summarize the ansatz_we_use and the resulting BPS equations_for_an_arbitrary electrically gauged $\N=2$ supergravity_theory. The restriction |
ions order ferromagnetically, while the corresponding Ni and Co ions order antiferromagnetically. Therefore, as seen in figures 1c and 1d, there is not much contribution of the magnetic field into the measured specific heat.
>From resistivity measurements the LaMnO$_{3}$, La$_{1.5}$Sr$_{0.5}$NiO$_{4}$ and La$_{1.5}$Sr$_{0.5}$CoO$_{4}$ samples are electrical insulators at low temperatures, and applied magnetic fields up to 9 T are not strong enough to destroy this characteristic[@Zaliznyak],[@Wochner],[@Myron]. Therefore, we should not expect for the previous three crystals the linear contribution from free electrons to the specific heat. However, other kind of many-body excitations could also lead to a linear contribution[@Smolyaninova1]. On the other hand, the La$_{1.35}$Sr$_{1.65}$Mn$_{2}$O$_{7}$ is metallic below about 100 K[@Okuda]. Our fitting shows that $\beta
_{1} $ values are big for La$_{1.35}$Sr$_{1.65}$Mn$_{2}$O$_{7}$ and La$_{1.5}
$Sr$_{0.5}$NiO$_{4}$. To facilitate even more the comparison, we have re-plotted in figure 2 the data for the La$_{1.35}$Sr$_{1.65}$Mn$_{2}$O$_{7}$ and La$_{1.5}$Sr$_{0.5}$NiO$_{4}$ crystals at temperatures below 10 K. Closed symbols represent the measurements with [*H=0*]{} and open ones with [*9 T*]{}.
Okuda et al.[@Okuda] found that the decrease of specific heat at low temperatures, due to an applied magnetic field of 9 T, was about ten times larger in a La$_{1.35}$Sr$_{1.65}$Mn$_{2}$O$_{7}$ sample than the observed values in La$_{1-x}$Sr$_{x}$MnO$_{3}$ samples (with x=0.3 and 0.4). They also calculated the theoretical reduction in specific heat upon application of a magnetic field for the ideal simple-cubic (3D) and simple-square (2D) lattices and concluded that the observed change in specific heat for the bilayered manganite was large, but still less than that | ions order ferromagnetically, while the corresponding Ni and Co ions decree antiferromagnetically. consequently, as seen in figures 1c and 1d, there be not much contribution of the charismatic field into the measured specific heat.
& gt;From resistivity measurement the LaMnO$_{3}$, La$_{1.5}$Sr$_{0.5}$NiO$_{4}$ and La$_{1.5}$Sr$_{0.5}$CoO$_{4}$ sample distribution are electrical insulators at low temperatures, and apply magnetic fields up to 9 T are not strong enough to destroy this characteristic[@Zaliznyak],[@Wochner],[@Myron ]. Therefore, we should not ask for the previous three crystals the linear contribution from barren electrons to the specific heat. However, other kind of many - body excitement could also lead to a linear contribution[@Smolyaninova1 ]. On the other hand, the La$_{1.35}$Sr$_{1.65}$Mn$_{2}$O$_{7}$ is metallic below approximately 100 K[@Okuda ]. Our fitting shows that $ \beta
_ { 1 } $ values are big for La$_{1.35}$Sr$_{1.65}$Mn$_{2}$O$_{7}$ and La$_{1.5 }
$ Sr$_{0.5}$NiO$_{4}$. To facilitate even more the comparison, we have re - plotted in figure 2 the datum for the La$_{1.35}$Sr$_{1.65}$Mn$_{2}$O$_{7}$ and La$_{1.5}$Sr$_{0.5}$NiO$_{4}$ crystal at temperatures below 10 K. Closed symbols represent the measurements with [ * H=0 * ] { } and open ones with [ * 9 T * ] { }.
Okuda et al.[@Okuda ] found that the decrease of specific heat at low temperature, due to an applied magnetic sphere of 9 metric ton, was about ten times larger in a La$_{1.35}$Sr$_{1.65}$Mn$_{2}$O$_{7}$ sample than the observed values in La$_{1 - x}$Sr$_{x}$MnO$_{3}$ samples (with x=0.3 and 0.4). They besides calculated the theoretical reduction in specific heat upon application of a magnetic field for the ideal simple - cubic (3D) and simple - square (2D) lattices and concluded that the ascertained change in specific hotness for the bilayered manganite was large, but still less than that | iond order ferromagneticalln, while the corrgspondinj Ni ans Co ionr order antiferromagneticallb. Thwrefoee, as seen in figures 1z and 1d, tjere is bot nych contriuhtion on the lagnztmc field into tme measured specific heat.
&ct;Wrlm resistivity measurements the LaMgO$_{3}$, La$_{1.5}$Sr$_{0.5}$MiL$_{4}$ and La$_{1.5}$Sr$_{0.5}$CoO$_{4}$ fampkqs adv tlectrical insulators at low tempsraturev, and applied magnetic fields up to 9 T wre jot strong enough ho destroy jgis xharacteristkc[@Zaliznyak],[@Wochner],[@Myroh]. Therefore, we should not expecg for the previiuw tjtee crystals the jinear contrlnution from ftee electrons bo thx spwcific heat. However, ovher kind of many-bodi excitatimna could also lead ti a lhneas covrricutjoi[@Smklyanijove1]. On the otger hand, thw La$_{1.35}$Sr$_{1.65}$Mn$_{2}$O$_{7}$ is metalloc velow about 100 I[@Okuda]. Otr fitting shows that $\beta
_{1} $ values are tig for La$_{1.35}$Sr$_{1.65}$Mn$_{2}$O$_{7}$ and La$_{1.5}
$Sr$_{0.5}$NiI$_{4}$. To facilitate even lore the somparison, we have re-plotted in figure 2 the data xor tie La$_{1.35}$Wr$_{1.65}$Iv$_{2}$I$_{7}$ wnd La$_{1.5}$Sr$_{0.5}$NiO$_{4}$ crystals at temperatures below 10 K. Dlpsvd symbols represcnt the measuremenys wojh [*H=0*]{} and open unes wnfh [*9 T*]{}.
Okuda et al.[@Okudw] found that the decrtase pf specific heat at low temperatures, duv to an applied magnetnc field of 9 T, wss abput ten times larger in a Lz$_{1.35}$Sr$_{1.65}$Mn$_{2}$O$_{7}$ samppe than tgd observed valuer ik Ld$_{1-x}$Sr$_{x}$MnO$_{3}$ samples (with x=0.3 anq 0.4). They aoso ealculatdd tne thejretical rfductlmn in specific heah upou appnication ov a magnetic field for the ideal simple-cubic (3C) dnd simple-szuare (2D) lattices agd concluded tkat the jbserxed change in sperific heat fjr the bilayesgd manganite xas large, but stiol less ghan that | ions order ferromagnetically, while the corresponding Ni ions antiferromagnetically. Therefore, seen in figures not contribution of the field into the specific heat. >From resistivity measurements the La$_{1.5}$Sr$_{0.5}$NiO$_{4}$ and La$_{1.5}$Sr$_{0.5}$CoO$_{4}$ samples are electrical insulators at low temperatures, and applied magnetic up to 9 T are not strong enough to destroy this characteristic[@Zaliznyak],[@Wochner],[@Myron]. Therefore, should expect the three crystals the linear contribution from free electrons to the specific heat. However, other kind of excitations could also lead to a linear contribution[@Smolyaninova1]. the other hand, the is metallic below about 100 Our shows that _{1} values big for La$_{1.35}$Sr$_{1.65}$Mn$_{2}$O$_{7}$ La$_{1.5} $Sr$_{0.5}$NiO$_{4}$. To facilitate even more the comparison, we have re-plotted in figure 2 the data for La$_{1.35}$Sr$_{1.65}$Mn$_{2}$O$_{7}$ and at temperatures 10 Closed represent the measurements and open ones with [*9 T*]{}. found that the decrease of specific heat at temperatures, due an applied magnetic field of 9 was about ten times larger in a La$_{1.35}$Sr$_{1.65}$Mn$_{2}$O$_{7}$ than the observed values in La$_{1-x}$Sr$_{x}$MnO$_{3}$ samples (with x=0.3 and 0.4). They also calculated the in specific heat upon of a magnetic for ideal (3D) simple-square (2D) and concluded that the observed change in specific heat for the manganite was large, but still less than that | ions order ferromagneticallY, while the cOrresPonDinG NI and co ioNs order antiferROmagNetically. Therefore, as seEn in fIgURes 1c ANd 1D, therE is not mUCh CONtrIbUtIon Of THe MagneTic Field inTo the measuRed SpEcific heat.
>fRoM resistiviTy mEasurements tHe LAMnO$_{3}$, La$_{1.5}$sr$_{0.5}$niO$_{4}$ ANd La$_{1.5}$SR$_{0.5}$Coo$_{4}$ sampLes are ELectriCal insulaToRS at low TEmperatUREs, And aPplied magnetic fieLDs UP to 9 T are not stroNg enouGh TO dESTroY thIs characteRiStic[@ZALiznyak],[@wOcHNER],[@MyROn]. Therefore, we Should not exPEct For the PrEviOUs threE crysTaLS thE linear contRibuTion from fRee eleCTrons to THe speciFic heaT. HoWevEr, otHEr KiNd oF mANy-bODy ExcITatIons coulD aLsO lead To a lINEAR conTriButiOn[@SmoLyaninova1]. On thE otHer hANd, tHe La$_{1.35}$SR$_{1.65}$Mn$_{2}$O$_{7}$ iS metAlLic beLow aboUt 100 K[@OkUdA]. Our fitting showS thaT $\beta
_{1} $ valuEs aRe Big FoR La$_{1.35}$Sr$_{1.65}$mN$_{2}$O$_{7}$ and LA$_{1.5}
$Sr$_{0.5}$niO$_{4}$. to facilItate evEN moRe THE CoMparison, we have re-plOtTED iN figure 2 tHe data FOr ThE la$_{1.35}$Sr$_{1.65}$Mn$_{2}$O$_{7}$ aNd la$_{1.5}$SR$_{0.5}$NiO$_{4}$ CRYstalS at tEMpEratures Below 10 K. cLoSeD symbolS rEpreseNt The MeaSuremENts wIth [*H=0*]{} anD open oneS with [*9 t*]{}.
okuda et al.[@Okuda] FOund that the deCReASE oF SpecIfiC heat at low tEmpeRAturEs, duE To An aPPlied MagneTiC FiELd of 9 T, was about ten timEs Larger In a La$_{1.35}$sr$_{1.65}$Mn$_{2}$O$_{7}$ sample thAn the obserVED Values in la$_{1-x}$SR$_{X}$MNo$_{3}$ samples (with x=0.3 aNd 0.4). TheY also calcuLAted the tHeoreTical redUction in sPECific heaT upOn aPplIcaTIOn Of a magnetic fiELD for ThE ideal sImpLe-cubic (3d) anD siMplE-sqUaRe (2D) latticEs and conClUdEd ThAt tHe obsERved chanGe In sPeCifIc heaT For the BilayEred MaNgANitE was larGE, bUT StilL lEsS thaN thAt | ions order ferromagnetica lly, while thecor res po ndin g Ni and Co ions o r derantiferromagnetically. Ther ef o re,a sseenin figu r es 1 c a nd 1 d,th e re is n otmuch co ntribution of t he magneticf ie ld into th e m easured spec ifi c heat .
&g t ;From re sisti vity m e asurem ents theLa M nO$_{3 } $, La$_ { 1 .5 }$Sr $_{0.5}$NiO$_{4}$ an d La$_{1.5}$Sr$ _{0.5} $C o O$ _ { 4}$ sa mples areel ectri c al insu l at o r s at low temperatu res, and ap p lie d magn et icf ieldsup to 9 T a re not stro ng e nough todestro y this c h aracter istic[ @Za liz nyak ] ,[ @W och ne r ],[ @ My ron ] . T herefore ,we shou ld n o t e xpec t f or t he pr evious threecry stal s th e lin ear c ontr ib ution fromfreeel ectrons to thespec ific heat . H ow eve r, othe r kindofman y-bodyexcitat i ons c o u l dalso lead to a lin ea r co ntributi on[@Sm o ly an i nova1].On th e ot h e r han d, t h eLa$_{1.3 5}$Sr$ _ {1 .6 5}$Mn$_ {2 }$O$_{ 7} $ i s m etall i c be low ab out 100K[@Ok u da]. Our fitti n g shows that$ \b e t a_ {1}$ v alues are b ig f o r La $_{1 . 35 }$S r $_{1. 65}$M n$ _ {2 } $O$_{7}$ and La$_{1 .5 }
$Sr$ _{0.5 }$NiO$_{4}$.To facilit a t e even mo re t h ec omparison, wehavere-plotted in figur e 2 t he datafor the L a $ _{1.35}$ Sr$ _{1 .65 }$M n $ _{ 2}$O$_{7}$ an d La$_ {1 .5}$Sr$ _{0 .5}$NiO $_{ 4}$ cr yst al s at temp eratures b el ow 1 0 K . Clo s ed symbo ls re pr ese nt th e measu remen ts w it h[ *H= 0*]{} a n do p en o ne swith [* 9T*]{} .
O k uda et al. [@Okuda]fou n d th at t he decr ease of speci fi c heat atlo w t empera t u res, due to an applied magnetic field o f 9 T, w as a bout tentim es lar ger in a L a$_{1. 35}$S r$ _{1 . 6 5}$Mn $ _ {2 }$O $_ {7}$ sampl e tha n the o bser ved val ues in La$_{1-x}$S r $_{ x}$MnO$_{3}$sam ples ( wi thx =0 . 3 a nd 0.4 ) . They also calc ulated the t h eo retical re d uct io n in sp ecificheatu pon app licationof a magn et ic f i e ldfor the id eal simp le-cubic( 3D) a n dsimpl e-s quare(2 D)latti ces an d co nclud ed tha tthe ob serve dchange i n specific heat for the bilay eredman ganite wa s l a rge , but sti ll l ess than t hat | ions order_ferromagnetically, while_the corresponding Ni and_Co ions_order_antiferromagnetically. Therefore,_as_seen in figures_1c and 1d,_there is not much_contribution of the_magnetic_field into the measured specific heat.
>From resistivity measurements the LaMnO$_{3}$, La$_{1.5}$Sr$_{0.5}$NiO$_{4}$ and La$_{1.5}$Sr$_{0.5}$CoO$_{4}$ samples_are_electrical insulators_at_low_temperatures, and applied magnetic fields_up to 9 T are_not strong_enough to destroy this characteristic[@Zaliznyak],[@Wochner],[@Myron]. Therefore, we should_not_expect for the_previous three crystals the linear contribution from free electrons_to the specific heat. However, other_kind of many-body_excitations_could_also lead to a_linear contribution[@Smolyaninova1]. On the other hand,_the La$_{1.35}$Sr$_{1.65}$Mn$_{2}$O$_{7}$ is metallic below about_100 K[@Okuda]. Our fitting shows that $\beta
_{1}_$ values are big for La$_{1.35}$Sr$_{1.65}$Mn$_{2}$O$_{7}$_and La$_{1.5}
$Sr$_{0.5}$NiO$_{4}$. To facilitate even_more the_comparison, we have re-plotted in_figure 2 the_data for_the La$_{1.35}$Sr$_{1.65}$Mn$_{2}$O$_{7}$ and_La$_{1.5}$Sr$_{0.5}$NiO$_{4}$ crystals at temperatures below 10_K. Closed symbols_represent the measurements with [*H=0*]{} and_open_ones with [*9_T*]{}.
Okuda_et_al.[@Okuda] found_that the decrease_of_specific heat_at_low temperatures, due to an applied_magnetic_field of 9 T, was about ten_times larger in a_La$_{1.35}$Sr$_{1.65}$Mn$_{2}$O$_{7}$_sample than the observed_values in La$_{1-x}$Sr$_{x}$MnO$_{3}$ samples (with_x=0.3 and 0.4). They also calculated_the theoretical_reduction in_specific heat upon application of a magnetic field for the ideal_simple-cubic (3D) and simple-square (2D) lattices_and concluded that the_observed change_in_specific heat for_the_bilayered manganite_was large, but still less than that |
molecules as the glass transition is approached [@weeks1; @dawson1].
![ Glassy diffusion in the noise-perturbed logistic map. (a) Repeated-cell map (thick dashed line) and trajectory (full line). (b) Time evolution of the mean square displacement $\left\langle
x_{t}^{2}\right\rangle-\left\langle x_{t}\right\rangle ^{2}$ for an ensemble of $1000$ trajectories with initial conditions randomly distributed inside $[-1,1]$. Curves are labeled by the value of the noise amplitude. []{data-label="fig_diff"}](diff_cell.eps "fig:"){width="0.95\columnwidth"}![ Glassy diffusion in the noise-perturbed logistic map. (a) Repeated-cell map (thick dashed line) and trajectory (full line). (b) Time evolution of the mean square displacement $\left\langle
x_{t}^{2}\right\rangle-\left\langle x_{t}\right\rangle ^{2}$ for an ensemble of $1000$ trajectories with initial conditions randomly distributed inside $[-1,1]$. Curves are labeled by the value of the noise amplitude. []{data-label="fig_diff"}](diff_ensemble.eps "fig:"){width="0.95\columnwidth"}
To investigate this aspect of vitrification in the map at $\mu _{c}(\sigma )$, we constructed a periodic map via repetition of a single (cell) map. This setting has being used to study deterministic diffusion in nonlinear maps, in which the trajectories migrate into neighboring cells due to chaotic motion. For fully chaotic maps diffusion is normal [@schuster1] but for marginally chaotic maps it is anomalous [@barkai1]. In our case we design the map in such a way that diffusion is due only to the random noise term, otherwise motion is confined to a single cell. So, we have the periodic map $x_{t+1}=F(x_{t})$, $F(l+x)=l+F(x)$, $l=...-1,0,1,...$, where $$F(x)=\left\{
\begin{array}{c}
-\left\vert 1-\mu _{c}x^{2}\right\vert +\sigma \xi,\;-1\leq x<0, \\
\left\vert 1-\mu _{c}x^{2}\right\vert +\sigma \xi | molecules as the glass transition is approached [ @weeks1; @dawson1 ].
! [ Glassy dissemination in the randomness - perturbed logistic map. (a) recur - cellular telephone map (thick dashed pipeline) and trajectory (full line). (b) Time development of the base square translation $ \left\langle
x_{t}^{2}\right\rangle-\left\langle x_{t}\right\rangle ^{2}$ for an ensemble of $ 1000 $ trajectories with initial condition randomly distributed inside $ [ -1,1]$. Curves are labeled by the value of the noise amplitude. [ ] { data - label="fig_diff"}](diff_cell.eps " fig:"){width="0.95\columnwidth " }! [ Glassy dissemination in the noise - perturbed logistic map. (a) repeat - cell map (thick dashed argumentation) and trajectory (full line). (b) Time evolution of the mean square translation $ \left\langle
x_{t}^{2}\right\rangle-\left\langle x_{t}\right\rangle ^{2}$ for an ensemble of $ 1000 $ trajectories with initial conditions randomly distributed inside $ [ -1,1]$. Curves are labeled by the value of the noise amplitude. [ ] { data - label="fig_diff"}](diff_ensemble.eps " fig:"){width="0.95\columnwidth " }
To investigate this aspect of vitrification in the map at $ \mu _ { c}(\sigma) $, we constructed a periodic map via repetition of a single (cell) map. This setting has being use to study deterministic diffusion in nonlinear function, in which the trajectory migrate into neighboring cell due to chaotic motion. For fully chaotic maps diffusion is normal [ @schuster1 ] but for marginally chaotic maps it is anomalous [ @barkai1 ]. In our case we design the function in such a way that diffusion is due only to the random noise term, otherwise motion is confined to a single cell. So, we give birth the periodic function $ x_{t+1}=F(x_{t})$, $ F(l+x)=l+F(x)$, $ l=... -1,0,1,... $, where $ $ F(x)=\left\ {
\begin{array}{c }
-\left\vert 1-\mu _ { c}x^{2}\right\vert + \sigma \xi,\;-1\leq x<0, \\
\left\vert 1-\mu _ { c}x^{2}\right\vert + \sigma \xi | mopecules as the glass traksition is approcxhed [@wxeks1; @dasson1].
![ Glarsy diffusion in the noise-pecturved ligistic map. (a) Repeated-zell map (nhick dasyed oune) and trenectory (full mlne). (b) Vime evolution pf the meat square displdcdmznt $\left\langle
x_{t}^{2}\right\rangle-\left\langlq x_{t}\rignt\gangle ^{2}$ for an enstmbje or $1000$ trajectories with initial condifions rendomly distribited inside $[-1,1]$. Curves are lahelef by the value of hhe noise anplieyde. []{data-labeu="fig_diff"}](diff_cell.eps "fif:"){width="0.95\columnwidth"}![ Glassy diffuskon iu the noise-kzetugted logistir map. (w) Repeated-cell map (tvick daxhed line) and brajertort (full line). (b) Time evmlution of the meag square giaplacement $\left\labgoe
x_{t}^{2}\rhght\sanguw-\lewt\lznjle x_{t}\rigjt\rengle ^{2}$ for zn ensemble of $1000$ trajectories wotr initial condjtions rwndomly distributed inside $[-1,1]$. Curves are nabsled by the value of thw noise amplitude. []{datw-label="fig_qiff"}](diff_ensemble.eps "fig:"){width="0.95\columnwidth"}
To investicate vhks cwpect if vitrification in the map at $\mu _{c}(\sigma )$, we cogattubted a periodic mcp via repetitiom lf w single (cell) map. Tkjs setting has being used tj stuey determynisyic diffusion in nonlinear naps, in whick tye trajectories miyrate into nzighboting crlls due to chaotic motnon. Fod fully chaltic maps aiffusion is noroal [@swhuster1] but for marginally chaotic naps it is xnomslous [@farkai1]. In lur case we design the mwp in sgch a way hhat diffusion is due only to tix random noisg tarm, otherwife mobion is confineq to a single eell. So, ce havd the perikdic ma' $x_{t+1}=F(x_{t})$, $F(l+x)=l+S(x)$, $l=...-1,0,1,...$, where $$F(x)=\ngft\{
\begin{arrab}{c}
-\left\vere 1-\mu _{c}x^{2}\rught\verg +\sigma \xi,\;-1\leq x<0, \\
\left\vern 1-\iu _{c}z^{2}\right\vert +\sigma \xl | molecules as the glass transition is approached ![ diffusion in noise-perturbed logistic map. line) trajectory (full line). Time evolution of mean square displacement $\left\langle x_{t}^{2}\right\rangle-\left\langle x_{t}\right\rangle for an ensemble of $1000$ trajectories with initial conditions randomly distributed inside $[-1,1]$. are labeled by the value of the noise amplitude. []{data-label="fig_diff"}](diff_cell.eps "fig:"){width="0.95\columnwidth"}![ Glassy diffusion the logistic (a) map (thick dashed line) and trajectory (full line). (b) Time evolution of the mean square displacement x_{t}^{2}\right\rangle-\left\langle x_{t}\right\rangle ^{2}$ for an ensemble of $1000$ with initial conditions randomly inside $[-1,1]$. Curves are labeled the of the amplitude. "fig:"){width="0.95\columnwidth"} investigate this aspect vitrification in the map at $\mu _{c}(\sigma )$, we constructed a periodic map via repetition of a (cell) map. has being to deterministic in nonlinear maps, the trajectories migrate into neighboring cells motion. For fully chaotic maps diffusion is normal but for chaotic maps it is anomalous [@barkai1]. our case we design the map in such way that diffusion is due only to the random noise term, otherwise motion is confined single cell. So, we the periodic map $F(l+x)=l+F(x)$, where \begin{array}{c} 1-\mu _{c}x^{2}\right\vert \xi,\;-1\leq x<0, \\ \left\vert 1-\mu _{c}x^{2}\right\vert +\sigma \xi | molecules as the glass transiTion is apprOacheD [@weEks1; @DaWson1].
![ glasSy diffusion in tHE noiSe-perturbed logistic map. (A) RepeAtED-celL MaP (thicK dashed LInE) ANd tRaJeCtoRy (FUlL line). (B) TiMe evoluTion of the mEan SqUare displaceMEnT $\left\langlE
x_{t}^{2}\Right\rangle-\lEft\Langle X_{t}\RigHT\rangLe ^{2}$ fOr an eNsemblE Of $1000$ trajEctories wItH InitiaL ConditiONS rAndoMly distributed insIDe $[-1,1]$. cUrves are labeleD by the VaLUe OF The NoiSe amplitudE. []{dAta-laBEl="fig_diFF"}](dIFF_CelL.Eps "fig:"){width="0.95\coLumnwidth"}![ GlASsy DiffusIoN in THe noisE-pertUrBEd lOgistic map. (a) repeAted-cell mAp (thicK Dashed lINe) and trAjectoRy (fUll Line). (B) tiMe EvoLuTIon OF tHe mEAn sQuare disPlAcEment $\Left\LANGLe
x_{t}^{2}\RigHt\raNgle-\lEft\langle x_{t}\riGht\RangLE ^{2}$ foR an enSemblE of $1000$ tRaJectoRies wiTh iniTiAl conditions ranDomlY distribuTed InSidE $[-1,1]$. CUrves ARe labeLed By tHe value Of the noISe aMpLITUdE. []{data-label="fig_diff"}](dIfF_ENsEmble.eps "Fig:"){widTH="0.95\cOlUMnwidth"}
TO iNveStigATE this AspeCT oF vitrifiCation IN tHe Map at $\mu _{C}(\sIgma )$, we CoNstRucTed a pERiodIc map vIa repetiTion oF A single (cell) map. tHis setting has BEiNG UsED to sTudY determinisTic dIFfusIon iN NoNliNEar maPs, in wHiCH tHE trajectories migratE iNto neiGhborIng cells due to Chaotic motION. for fully ChaoTIc MAps diffusion is NormaL [@schuster1] bUT for margInallY chaotic Maps it is aNOMalous [@baRkaI1]. In Our CasE WE dEsign the map in SUCh a wAy That difFusIon is duE onLy tO thE raNdOm noise teRm, otherwIsE mOtIoN is ConfiNEd to a sinGlE ceLl. so, wE have THe periOdic mAp $x_{t+1}=f(x_{T})$, $F(L+X)=l+F(X)$, $l=...-1,0,1,...$, where $$f(X)=\lEFT\{
\begIn{ArRay}{c}
-\LefT\vErt 1-\mu _{C}x^{2}\riGHt\vErt +\sigmA \xi,\;-1\leq x<0, \\
\leFt\vERt 1-\mu _{C}x^{2}\RiGht\vert +\Sigma \xi | molecules as the glass tr ansition i s app roa che d[@we eks1 ; @dawson1].
! [ Gl assy diffusion in thenoise -p e rtur b ed logi stic ma p .( a ) R ep ea ted -c e ll map(th ick das hed line)and t rajectory (f u ll line). (b ) T ime evolutio n o f theme ans quare di splac ement$ \left\ langle
x_ {t } ^{2}\r i ght\ran g l e- \lef t\langle x_{t}\ri g ht \ rangle ^{2}$ f or anen s em b l e o f $ 1000$ traj ec torie s with i n it i a l co n ditions rando mly distrib u ted insid e$[- 1 ,1]$.Curve sa relabeled bythevalue ofthe no i se ampl i tude. [ ]{data -la bel ="fi g _d if f"} ]( d iff _ ce ll. e ps"fig:"){ wi dt h="0. 95\c o l u m nwid th" }![Glass y diffusion i n t he n o ise -pert urbed log is tic m ap. (a ) Rep ea ted-cell map (t hick dashed l ine )and t rajec t ory (f ull li ne). (b ) Timee vol ut i o n o f the mean squaredi s p la cement $ \left\ l an gl e
x_{t}^{ 2} \ri ght\ r a ngle- \lef t \l angle x_ {t}\ri g ht \r angle ^ {2 }$ for a n e nse mbleo f $1 000$ t rajector ies w i th initial con d itions random l yd i st r ibut edinside $[-1 ,1]$ . Cur vesa re la b eledby th ev al u e of the noise ampl it ude. [ ]{dat a-label="fig_ diff"}](di f f _ ensemble .eps "f i g:"){width="0. 95\co lumnwidth" }
To inv estig ate this aspect o f vitrific ati oninthe m ap at $\mu _{c} ( \ sigm a)$, wecon structe d a pe rio dic m ap via re petition o fasi ngl e (ce l l) map.Th isse tti ng ha s being used tost ud y de termini s ti c diff us io n in no nl inear map s , i n which the traj ect o ries m ig rate in to neighborin gcells dueto ch aoticm o tion. Fo r fully chaotic maps di f fusionisnorma l [@ schuster1 ] b ut for ma r ginall y chao tic m ap s i t is an o m al ous [ @barkai1]. I n o ur ca se wedesignthe map in such aw aythat diffusio n i s du e on lyt ot hera n dom n oise term, othe rwise moti on is confinedt o a s ingle c ell. So , weh ave the periodic map $x_{ t+ 1}=F ( x _{t })$, $F(l+ x)=l+F(x )$, $l=.. . -1,0, 1 ,. ..$,whe re $$F (x )=\ left\ {
\be g in{ array }{c}
- \l eft\ve rt 1- \m u _{c}x^ {2}\right\vert +\sigma\xi,\; -1\le q x <0, \\
\ lef t \ve rt 1-\mu_{c} x^{2}\righ t\v ert +\si gma \xi | molecules_as the_glass transition is approached_[@weeks1; @dawson1].
![_Glassy_diffusion in_the_noise-perturbed logistic map._(a) Repeated-cell map_(thick dashed line) and_trajectory (full line)._(b)_Time evolution of the mean square displacement $\left\langle
x_{t}^{2}\right\rangle-\left\langle x_{t}\right\rangle ^{2}$ for an ensemble of_$1000$_trajectories with_initial_conditions_randomly distributed inside $[-1,1]$. Curves_are labeled by the value_of the_noise amplitude. []{data-label="fig_diff"}](diff_cell.eps "fig:"){width="0.95\columnwidth"}![ Glassy diffusion in the_noise-perturbed_logistic map. (a)_Repeated-cell map (thick dashed line) and trajectory (full line)._(b) Time evolution of the mean_square displacement $\left\langle
x_{t}^{2}\right\rangle-\left\langle_x_{t}\right\rangle_^{2}$_for an ensemble of_$1000$ trajectories with initial conditions randomly_distributed inside $[-1,1]$. Curves are labeled_by the value of the noise amplitude._[]{data-label="fig_diff"}](diff_ensemble.eps "fig:"){width="0.95\columnwidth"}
To investigate this aspect of_vitrification in the map at_$\mu _{c}(\sigma_)$, we constructed a periodic_map via repetition_of a_single (cell) map._This setting has being used to_study deterministic diffusion_in nonlinear maps, in which the_trajectories_migrate into neighboring_cells_due_to chaotic_motion. For fully_chaotic_maps diffusion_is_normal [@schuster1] but for marginally chaotic_maps_it is anomalous [@barkai1]. In our case_we design the map_in_such a way that_diffusion is due only to_the random noise term, otherwise motion_is confined_to a_single cell. So, we have the periodic map $x_{t+1}=F(x_{t})$, $F(l+x)=l+F(x)$, $l=...-1,0,1,...$,_where $$F(x)=\left\{
\begin{array}{c}
-\left\vert 1-\mu _{c}x^{2}\right\vert +\sigma_\xi,\;-1\leq x<0, \\
\left\vert_1-\mu _{c}x^{2}\right\vert_+\sigma_\xi |
kinds of the $\Lambda N$ interaction, YNG-ND and -JA are used.[]{data-label="tab1"}
![Contour maps of the energy surfaces for $J^\pi=0^+$(left top), $2^+$(right top), and $4^+$(bottom) states in the two-parameter space $\beta_x=\beta_y(\equiv \beta_\perp), \beta_z$, given as $E(\beta_\perp,\beta_z)$ in the calculation of Eq. (\[eq:ghweq\]). Two minima are denoted by $\times$ and $+$. YNG-ND is adopted for the $\Lambda N$ interaction.[]{data-label="fig:energy"}](2almj0esfc_kpdg_def_pot6_lv1.eps "fig:") \[fig:energy1\]![Contour maps of the energy surfaces for $J^\pi=0^+$(left top), $2^+$(right top), and $4^+$(bottom) states in the two-parameter space $\beta_x=\beta_y(\equiv \beta_\perp), \beta_z$, given as $E(\beta_\perp,\beta_z)$ in the calculation of Eq. (\[eq:ghweq\]). Two minima are denoted by $\times$ and $+$. YNG-ND is adopted for the $\Lambda N$ interaction.[]{data-label="fig:energy"}](2almj2esfc_kpdg_def_pot6_lv1.eps "fig:") \[fig:energy2\]![Contour maps of the energy surfaces for $J^\pi=0^+$(left top), $2^+$(right top), and $4^+$(bottom) states in the two-parameter space $\beta_x=\beta_y(\equiv \beta_\perp), \beta_z$, given as $E(\beta_\perp,\beta_z)$ in the calculation of Eq. (\[eq:ghweq\]). Two minima are denoted by $\times$ and $+$. YNG-ND is adopted for the $\Lambda N$ interaction.[]{data-label="fig:energy"}](2almj4esfc_kpdg_def_pot6_l5_lv1.eps "fig:") \[fig:energy3\]
In order to compare the single Hyper-THSR wave function with the Brink-GCM wave functions obtained above, we first solve the Hill | kinds of the $ \Lambda N$ interaction, YNG - ND and -JA are used.[]{data - label="tab1 " }
! [ Contour maps of the energy surfaces for $ J^\pi=0^+$(left circus tent), $ 2^+$(right top), and $ 4^+$(bottom) state in the two - parameter space $ \beta_x=\beta_y(\equiv \beta_\perp), \beta_z$, given as $ E(\beta_\perp,\beta_z)$ in the calculation of Eq. (\[eq: ghweq\ ]). Two minima are denote by $ \times$ and $ + $. YNG - ND is adopted for the $ \Lambda N$ interaction.[]{data - label="fig: energy"}](2almj0esfc_kpdg_def_pot6_lv1.eps " fig: ") \[fig: energy1\]![Contour function of the department of energy surfaces for $ J^\pi=0^+$(left top), $ 2^+$(right top), and $ 4^+$(bottom) states in the two - parameter space $ \beta_x=\beta_y(\equiv \beta_\perp), \beta_z$, give as $ E(\beta_\perp,\beta_z)$ in the calculation of Eq. (\[eq: ghweq\ ]). Two minima are denoted by $ \times$ and $ + $. YNG - ND is adopted for the $ \Lambda N$ interaction.[]{data - label="fig: energy"}](2almj2esfc_kpdg_def_pot6_lv1.eps " libyan islamic fighting group: ") \[fig: energy2\]![Contour map of the energy surfaces for $ J^\pi=0^+$(left top), $ 2^+$(right top), and $ 4^+$(bottom) states in the two - parameter outer space $ \beta_x=\beta_y(\equiv \beta_\perp), \beta_z$, given as $ E(\beta_\perp,\beta_z)$ in the calculation of Eq. (\[eq: ghweq\ ]). Two minima are denoted by $ \times$ and $ + $. YNG - ND is adopted for the $ \Lambda N$ interaction.[]{data - label="fig: energy"}](2almj4esfc_kpdg_def_pot6_l5_lv1.eps " libyan islamic fighting group: ") \[fig: energy3\ ]
In order to compare the single Hyper - THSR wave function with the Brink - GCM wave functions obtain above, we first solve the Hill | kijds of the $\Lambda N$ intevaction, YNG-ND anb -JA arx used.[]{dzta-label="gab1"}
![Contour maps of the energb suefacew for $J^\pi=0^+$(left top), $2^+$(righg top), and $4^+$(bottom) wtatts in the two-parameter space $\befw_x=\beca_b(\equiv \beta_\perp), \beta_z$, givan as $E(\beta_\per[,\bdtc_z)$ in the calculation of Eq. (\[eq:ghweq\]). Tro minika are denoted bi $\timts$ wnd $+$. JNN-ND is adopted for the $\Lambda N$ jnteracuion.[]{data-label="fig:enrrgy"}](2almj0esfc_kpdg_def_pot6_lv1.epd "fih:") \[fig:energy1\]![Contour maps of thg enqegy surfaces for $J^\pi=0^+$(left top), $2^+$(right top), and $4^+$(bottom) states in the tdo-parcmeter spacg $\gehd_x=\beta_y(\equit \beta_\ierp), \beta_z$, giyvn as $E(\teta_\perl,\beta_z)$ in the galcunatuon of Eq. (\[eq:ghweq\]). Two kinima are denoted by $\times$ aud $+$. YNG-ND is adopted doe the $\Lamtda V$ ingerzcvioh.[]{data-lwbem="fig:energy"}](2zlmj2esfc_kpdt_def_pot6_lv1.eps "fig:") \[fib:egvtgy2\]![Contour mzps of tre energy surfaces for $J^\pi=0^+$(left top), $2^+$(righu top), and $4^+$(bottom) states in tye two-parameter space $\beta_x=\betw_y(\equiv \beta_\perp), \beta_z$, given as $E(\beta_\perp,\beta_z)$ in the raucuoabion if Eq. (\[eq:ghweq\]). Two minima are denoted by $\times$ anq $+$. UNN-ND is adopted fjr the $\Lambca N$ ynteraction.[]{daja-label="yjg:snergy"}](2almj4esfc_kpdg_fef_pot6_l5_jv1.eps "fig:") \[fig:egergu3\]
In order to compare the sibgle Hyper-THFE wave function wich the Brink-YCM wafe fumctions obtained above, ce firat solve thf Hill | kinds of the $\Lambda N$ interaction, YNG-ND are ![Contour maps the energy surfaces and states in the space $\beta_x=\beta_y(\equiv \beta_\perp), given as $E(\beta_\perp,\beta_z)$ in the calculation Eq. (\[eq:ghweq\]). Two minima are denoted by $\times$ and $+$. YNG-ND is adopted the $\Lambda N$ interaction.[]{data-label="fig:energy"}](2almj0esfc_kpdg_def_pot6_lv1.eps "fig:") \[fig:energy1\]![Contour maps of the energy surfaces for $J^\pi=0^+$(left $2^+$(right and states the two-parameter space $\beta_x=\beta_y(\equiv \beta_\perp), \beta_z$, given as $E(\beta_\perp,\beta_z)$ in the calculation of Eq. (\[eq:ghweq\]). Two are denoted by $\times$ and $+$. YNG-ND is for the $\Lambda N$ "fig:") \[fig:energy2\]![Contour maps of the surfaces $J^\pi=0^+$(left top), top), $4^+$(bottom) in the two-parameter $\beta_x=\beta_y(\equiv \beta_\perp), \beta_z$, given as $E(\beta_\perp,\beta_z)$ in the calculation of Eq. (\[eq:ghweq\]). Two minima are denoted by and $+$. adopted for $\Lambda interaction.[]{data-label="fig:energy"}](2almj4esfc_kpdg_def_pot6_l5_lv1.eps \[fig:energy3\] In order the single Hyper-THSR wave function with functions obtained above, we first solve the Hill | kinds of the $\Lambda N$ interactIon, YNG-ND anD -JA arE usEd.[]{dAtA-labEl="taB1"}
![Contour maps of THe enErgy surfaces for $J^\pi=0^+$(left Top), $2^+$(riGhT Top), aND $4^+$(bOttom) States iN ThE TWo-pArAmEteR sPAcE $\beta_X=\beTa_y(\equiV \beta_\perp), \bEta_Z$, gIven as $E(\beta_\pERp,\Beta_z)$ in the CalCulation of Eq. (\[Eq:gHweq\]). TwO mIniMA are dEnoTed by $\Times$ aND $+$. YNG-ND Is adopted FoR The $\LamBDa N$ inteRACtIon.[]{dAta-label="fig:energy"}](2ALmJ0Esfc_kpdg_def_pot6_Lv1.eps "fIg:") \[FIg:ENErgY1\]![CoNtour maps oF tHe eneRGy surfaCEs FOR $j^\pi=0^+$(LEft top), $2^+$(right toP), and $4^+$(bottom) sTAteS in the TwO-paRAmeter Space $\BeTA_x=\bEta_y(\equiv \beTa_\peRp), \beta_z$, giVen as $E(\BEta_\perp,\BEta_z)$ in tHe calcUlaTioN of EQ. (\[Eq:GhWeq\]). twO MinIMa Are DEnoTed by $\timEs$ AnD $+$. YNG-Nd is aDOPTEd foR thE $\LamBda N$ iNteraction.[]{datA-laBel="fIG:enErgy"}](2aLmj2esFc_kpDg_Def_poT6_lv1.eps "Fig:") \[fiG:eNergy2\]![Contour mapS of tHe energy sUrfAcEs fOr $j^\pi=0^+$(leFT top), $2^+$(riGht Top), And $4^+$(bottOm) stateS In tHe TWO-PaRameter space $\beta_x=\bEtA_Y(\EqUiv \beta_\pErp), \betA_Z$, gIvEN as $E(\beta_\PeRp,\bEta_z)$ IN The caLculATiOn of Eq. (\[eq:Ghweq\]). TWO mInIma are dEnOted by $\TiMes$ And $+$. yNG-ND IS adoPted foR the $\LambDa N$ inTEraction.[]{data-laBEl="fig:energy"}](2alMJ4eSFC_kPDg_deF_poT6_l5_lv1.eps "fig:") \[fIg:enERgy3\]
IN ordER tO coMPare tHe sinGlE hyPEr-THSR wave function wItH the BrInk-GCm wave functionS obtained aBOVE, we first SolvE ThE hill | kinds of the $\Lambda N$interactio n, YN G-N D a nd -JA are used.[]{data- l abel ="tab1"}
![Contour ma ps of t h e en e rg y sur faces f o r$ J ^\p i= 0^ +$( le f ttop), $2 ^+$(rig ht top), a nd$4 ^+$(bottom)s ta tes in the tw o-parameterspa ce $\b et a_x = \beta _y( \equi v \bet a _\perp ), \beta_ z$ , given as $E(\ b e ta _\pe rp,\beta_z)$ in t h ec alculation ofEq. (\ [e q :g h w eq\ ]). Two minim aare d e noted b y $ \ t i mes $ and $+$. YNG -ND is adop t edfor th e$\L a mbda N $ int er a cti on.[]{data- labe l="fig:en ergy"} ] (2almj0 e sfc_kpd g_def_ pot 6_l v1.e p s"f ig: ") \[f i g: ene r gy1 \]![Cont ou rmapsof t h e e nerg y s urfa ces f or $J^\pi=0^+ $(l eftt op) , $2^ +$(ri ghtto p), a nd $4^ +$(bo tt om) states in t he t wo-parame ter s pac e$\bet a _x=\be ta_ y(\ equiv \ beta_\p e rp) ,\ b e ta _z$, given as $E(\ be t a _\ perp,\be ta_z)$ in t h e calcul at ion ofE q . (\[ eq:g h we q\]). Tw o mini m aar e denot ed by $\ ti mes $ a nd $+ $ . YN G-ND i s adopte d for the $\Lambda N $ interaction. [ ]{ d a ta - labe l=" fig:energy" }](2 a lmj2 esfc _ kp dg_ d ef_po t6_lv 1. e ps "fig:") \[fig:energ y2 \]![Co ntour maps of theenergy sur f a c es for $ J^\p i =0 ^ +$(left top),$2^+$ (right top ) , and $4 ^+$(b ottom) s tates int h e two-pa ram ete r s pac e $\ beta_x=\beta_ y ( \equ iv \beta_ \pe rp), \b eta _z$ , g ive nas $E(\be ta_\perp ,\ be ta _z )$in th e calcula ti onof Eq . (\[ e q:ghwe q\]). Two m in i maare den o te d by $ \t im es$and $ +$. Y NG-N D is adopte d for the $\ L ambd aN$ intera ction.[]{data -l abel="fig: en erg y"}](2 a l mj4esfc_ kpdg_def_pot6_l5_lv1.ep s "fig:" ) \ [fig: ener gy3\]
In or der to co m pare t he sin gle H yp er- T H SR wa v e f unc ti on with th e Bri nk-GC Mwave functi ons obtained above , we first solvethe Hil l | kinds_of the_$\Lambda N$ interaction, YNG-ND_and -JA_are_used.[]{data-label="tab1"}
![Contour maps_of_the energy surfaces_for $J^\pi=0^+$(left top),_$2^+$(right top), and $4^+$(bottom)_states in the_two-parameter_space $\beta_x=\beta_y(\equiv \beta_\perp), \beta_z$, given as $E(\beta_\perp,\beta_z)$ in the calculation of Eq. (\[eq:ghweq\]). Two minima_are_denoted by_$\times$_and_$+$. YNG-ND is adopted for_the $\Lambda N$ interaction.[]{data-label="fig:energy"}](2almj0esfc_kpdg_def_pot6_lv1.eps "fig:")_\[fig:energy1\]![Contour maps_of the energy surfaces for $J^\pi=0^+$(left top), $2^+$(right_top),_and $4^+$(bottom) states_in the two-parameter space $\beta_x=\beta_y(\equiv \beta_\perp), \beta_z$, given as_$E(\beta_\perp,\beta_z)$ in the calculation of Eq. (\[eq:ghweq\])._Two minima are_denoted_by_$\times$ and $+$. YNG-ND_is adopted for the $\Lambda N$_interaction.[]{data-label="fig:energy"}](2almj2esfc_kpdg_def_pot6_lv1.eps "fig:") \[fig:energy2\]![Contour maps of the_energy surfaces for $J^\pi=0^+$(left top), $2^+$(right top),_and $4^+$(bottom) states in the two-parameter_space $\beta_x=\beta_y(\equiv \beta_\perp), \beta_z$, given_as $E(\beta_\perp,\beta_z)$_in the calculation of Eq. (\[eq:ghweq\])._Two minima are_denoted by_$\times$ and $+$._YNG-ND is adopted for the $\Lambda_N$ interaction.[]{data-label="fig:energy"}](2almj4esfc_kpdg_def_pot6_l5_lv1.eps "fig:")_\[fig:energy3\]
In order to compare the single_Hyper-THSR_wave function with_the_Brink-GCM_wave functions_obtained above, we_first_solve the_Hill |
-bound*]{} of a subset $M \subseteq H$ is defined as $$\label{eq:ell-eins-bdd}
\abs{M}_1 := \inf \; \norm{L} \sup_{x \in M} \sum_{\alpha\in I}
\abs{\sprod{Rx}{e_\alpha}} \}$$ where the infimum is taken over all pairs of operators $(L, R)$ with $R: H \to \ell_2(I)$ and $L: \ell_2(I) \to H$ such that $L R =
\Id_H$. Let $X$ be a Banach space. An operator $T: X \to
H$, called [*-frame-bounded*]{} if $T$ maps the unit ball of $X$ into an -frame-bounded subset of $X$. In this case, $$\abs{T}_{\ell_1} := \bigl| \{ Tx \suchthat \norm{x}_X\le 1\} \bigr|_1$$ is called the [*-frame-bound*]{} of $T$.
\[lob.r.l1-bdd\]
\[item:l1-bdd-compact\] -frame-bounded sets need not be compact.
\[item:l1-bdd-composition\] Let $X, Y$ be Banach spaces. If $U: X \to H$ is -frame-bounded and $V: Y \to X$ is bounded, then $UV: Y \to H$ is -frame-bounded and $$\abs{UV}_{\ell_1} \le \abs{U}_{\ell_1} \norm{V}.$$
We point out that we do not know yet whether finite unions or simple translates of -frame-bounded sets are again -frame-bounded, something one would certainly expect to hold for a “good” boundedness concept. Consequently, we do not know whether the set of -frame-bounded operators $X \to H$ form a vector space.
\[lob.l.l1-bdd-basic\] Let $H$ be any Hilbert space and $M \subseteq H$. Then the following assertions hold.
\[item:label-lemma-a\] If $M$ is -frame-bounded, then it is norm-bounded, with $$\sup_{x\in M} \norm{x | -bound * ] { } of a subset $ M \subseteq H$ is defined as $ $ \label{eq: ell - eins - bdd }
\abs{M}_1: = \inf \; \norm{L } \sup_{x \in M } \sum_{\alpha\in I }
\abs{\sprod{Rx}{e_\alpha } } \}$$ where the infimum is taken over all pairs of operators $ (L, R)$ with $ radius: H \to \ell_2(I)$ and $ lambert: \ell_2(I) \to H$ such that $ L R =
\Id_H$. Let $ X$ be a Banach distance. An operator $ T: X \to
H$, call [ * -frame - bounce * ] { } if $ T$ maps the unit of measurement ball of $ X$ into an -frame - bounded subset of $ X$. In this character, $ $ \abs{T}_{\ell_1 }: = \bigl| \ { Tx \suchthat \norm{x}_X\le 1\ } \bigr|_1$$ is called the [ * -frame - bound * ] { } of $ T$.
\[lob.r.l1 - bdd\ ]
\[item: l1 - bdd - compact\ ] -frame - bound set need not be compact.
\[item: l1 - bdd - composition\ ] Let $ X, Y$ be Banach spaces. If $ U: ten \to H$ is -frame - bounded and $ V: Y \to X$ is bounded, then $ ultraviolet: Y \to H$ is -frame - bounded and $ $ \abs{UV}_{\ell_1 } \le \abs{U}_{\ell_1 } \norm{V}.$$
We point out that we do not acknowledge yet whether finite union or simple translates of -frame - bounded sets are again -frame - bounded, something one would certainly ask to hold for a “ good ” boundedness concept. Consequently, we do not know whether the set of -frame - bounded operators $ X \to H$ form a vector space.
\[lob.l.l1 - bdd - basic\ ] Let $ H$ be any Hilbert space and $ M \subseteq H$. Then the following assertions hold.
\[item: label - lemma - a\ ] If $ M$ is -frame - bounded, then it is norm - jump, with $ $ \sup_{x\in M } \norm{x | -boujd*]{} of a subset $M \subseted H$ is defined cw $$\laben{eq:ell-sins-bdd}
\xbs{M}_1 := \inf \; \norm{L} \sup_{x \in M} \sym_{\alpya\in I}
\abs{\sprod{Rx}{e_\alphx}} \}$$ where nhe infimym iw taken ovec all palxs of lperctirs $(L, R)$ with $R: H \to \ell_2(I)$ and $L: \ell_2(I) \to H$ sbch that $L R =
\Id_H$. Let $X$ be a Banach s[ace. An ooerator $T: X \to
R$, cakjed [*-rgane-bounded*]{} if $T$ maps the unit gall of $X$ into an -frsme-bounded subset of $X$. In hhis case, $$\abs{T}_{\ell_1} := \bigp| \{ Tx \suchtyat \girm{x}_X\le 1\} \bigf|_1$$ is called the [*-frame-bkund*]{} of $T$.
\[lob.r.l1-bdd\]
\[item:l1-bdd-compacg\] -frake-bounded wers jged not be cimpacn.
\[item:l1-bdd-composition\] Net $X, Y$ be Banach spages. Ih $U: Z \to H$ is -frame-boundev and $V: Y \to X$ is boonded, then $UR: Y \to H$ is -frame-bouneee and $$\abs{GV}_{\elu_1} \le \aba{U}_{\xll_1} \norm{V}.$$
Ae 'oint out tgat we do nit know yet whether fybite unions od simpje translates of -frame-bounded sets are agdin -frame-bounded, something one would certainly gxpect to rold for a “good” boundedness concept. Consequently, fe do vot kkow dyehher the set of -frame-bounded operators $X \to H$ rotm a vector spacc.
\[lob.l.l1-bdd-basic\] Let $H$ br any Hilbert rpace chd $M \subseteq H$. Then the fojlowibg assertyons hold.
\[item:label-lemma-a\] If $M$ iw -frame-boundvd, tyen it is norm-bounbed, with $$\sup_{r\in M} \morm{x | -bound*]{} of a subset $M \subseteq H$ as \abs{M}_1 := \; \norm{L} \sup_{x \}$$ the infimum is over all pairs operators $(L, R)$ with $R: H \ell_2(I)$ and $L: \ell_2(I) \to H$ such that $L R = \Id_H$. Let be a Banach space. An operator $T: X \to H$, called [*-frame-bounded*]{} if maps unit of into an -frame-bounded subset of $X$. In this case, $$\abs{T}_{\ell_1} := \bigl| \{ Tx \suchthat \norm{x}_X\le \bigr|_1$$ is called the [*-frame-bound*]{} of $T$. \[lob.r.l1-bdd\] -frame-bounded sets need not compact. \[item:l1-bdd-composition\] Let $X, Y$ Banach If $U: \to is and $V: Y X$ is bounded, then $UV: Y \to H$ is -frame-bounded and $$\abs{UV}_{\ell_1} \le \abs{U}_{\ell_1} \norm{V}.$$ We point that we know yet finite or translates of -frame-bounded again -frame-bounded, something one would certainly for a “good” boundedness concept. Consequently, we do know whether set of -frame-bounded operators $X \to form a vector space. \[lob.l.l1-bdd-basic\] Let $H$ be Hilbert space and $M \subseteq H$. Then the following assertions hold. \[item:label-lemma-a\] If $M$ is it is norm-bounded, with M} \norm{x | -bound*]{} of a subset $M \subseteq H$ iS defined as $$\Label{Eq:eLl-eInS-bdd}
\Abs{M}_1 := \Inf \; \norm{L} \sup_{x \in m} \Sum_{\aLpha\in I}
\abs{\sprod{Rx}{e_\alphA}} \}$$ wherE tHE infIMuM is taKen over ALl PAIrs Of OpEraToRS $(L, r)$ with $r: H \tO \ell_2(I)$ anD $L: \ell_2(I) \to H$ sUch ThAt $L R =
\Id_H$. Let $X$ bE A BAnach space. an oPerator $T: X \to
H$, CalLed [*-fraMe-BouNDed*]{} if $t$ maPs the Unit baLL of $X$ inTo an -frame-BoUNded suBSet of $X$. IN THiS casE, $$\abs{T}_{\ell_1} := \bigl| \{ Tx \sucHThAT \norm{x}_X\le 1\} \bigr|_1$$ iS calleD tHE [*-fRAMe-bOunD*]{} of $T$.
\[lob.r.l1-bDd\]
\[Item:l1-BDd-compaCT\] -fRAME-boUNded sets need nOt be compact.
\[ITem:L1-bdd-coMpOsiTIon\] Let $x, Y$ be BAnACh sPaces. If $U: X \to h$ is -fRame-boundEd and $V: y \To X$ is boUNded, theN $UV: Y \to h$ is -FraMe-boUNdEd And $$\AbS{uV}_{\eLL_1} \lE \abS{u}_{\elL_1} \norm{V}.$$
We PoInT out tHat wE DO NOt knOw yEt whEther Finite unions oR siMple TRanSlateS of -frAme-bOuNded sEts are Again -FrAme-bounded, sometHing One would cErtAiNly ExPect tO Hold foR a “gOod” BoundedNess conCEpt. coNSEQuEntly, we do not know whEtHER tHe set of -fRame-boUNdEd OPerators $x \tO H$ fOrm a VECtor sPace.
\[LOb.L.l1-bdd-basIc\] Let $H$ BE aNy hilbert SpAce and $m \sUbsEteQ H$. TheN The fOllowiNg assertIons hOLd.
\[item:label-lemMA-a\] If $M$ is -frame-bOUnDED, tHEn it Is nOrm-bounded, wIth $$\sUP_{x\in m} \norM{X | -bound*]{} of a subset $M\subseteqH$ is de fin ed as$$\l abel{eq:ell-ei n s-bd d}
\abs{M}_1 := \inf \; \n o rm{L } \ sup_{ x \in M } \ s u m_{ \a lp ha\ in I}
\ab s{\ sprod{R x}{e_\alph a}} \ }$$ where th e i nfimum istak en over allpai rs ofop era t ors $ (L, R)$with $ R : H \t o \ell_2( I) $ and $ L : \ell_ 2 ( I) \to H$ such that $LR = \Id_H$. Let $X $ be a B a na c h sp ace . An opera to r $T: X \to
H $ ,c a l led [*-frame-boun ded*]{} if$ T$maps t he un i t ball of $ X$ int o an -frame -bou nded subs et of$ X$. Int his cas e, $$\ abs {T} _{\e l l_ 1} := \ b igl | \ { T x \s uchthat\n or m{x}_ X\le 1 \ } \bi gr| _1$$ is c alled the [*- fra me-b o und *]{}of $T $.
\[ lob.r .l1-bd d\]
\[ item:l1-bdd-com pact \] -frame -bo un ded s ets n e ed not be co mpact.
\[item : l1- bd d - c om position\] Let $X, Y $ be Banachspaces . I f$ U: X \to H $ i s -f r a me-bo unde d a nd $V: Y \to X $ i sbounded ,then $ UV : Y \t o H$i s -f rame-b ounded a nd $$ \ abs{UV}_{\ell_ 1 } \le \abs{U} _ {\ e l l_ 1 } \n orm {V}.$$
Wepoin t out tha t w e d o notknowye t w h ether finite unions o r simp le tr anslates of - frame-boun d e d sets ar e ag a in -frame-bounded , som ething one would ce rtain ly expec t to hold f or a “go od” bo und edn e s sconcept. Cons e q uent ly , we do no t knowwhe the r t hese t of -fra me-bound ed o pe ra tor s $X\ to H$ fo rm ave cto r spa c e.
\[ lob.l .l1- bd d- b asi c\] Let $H $ be a ny H ilbe rtsp ace a nd $ M \s ubseteq H$. Then th e fol lo wi ng asse rtions hold.
\ [item:labe l- lem ma-a\] I f $M$ is -frame-bounded, then i t is nor m-b ounde d, w ith $$\su p_{ x\in M } \ n orm{x | -bound*]{} of_a subset_$M \subseteq H$ is_defined as_$$\label{eq:ell-eins-bdd}
_\abs{M}_1 :=_\inf_ \; _\norm{L} \sup_{x \in_M} \sum_{\alpha\in I}
\abs{\sprod{Rx}{e_\alpha}}_\}$$ where the_infimum_is taken over all pairs of operators $(L, R)$ with $R: H \to \ell_2(I)$_and_$L: \ell_2(I)_\to_H$_such that $L R =
\Id_H$._Let $X$ be a Banach_space. An_operator $T: X \to
H$, called [*-frame-bounded*]{} if $T$_maps_the unit ball_of $X$ into an -frame-bounded subset of $X$. In_this case, $$\abs{T}_{\ell_1} := \bigl| \{_Tx \suchthat \norm{x}_X\le_1\}_\bigr|_1$$_is called the [*-frame-bound*]{}_of $T$.
\[lob.r.l1-bdd\]
\[item:l1-bdd-compact\] -frame-bounded sets need not_be compact.
\[item:l1-bdd-composition\] Let $X, Y$ be_Banach spaces. If $U: X \to H$_is -frame-bounded and $V: Y \to_X$ is bounded, then $UV:_Y \to_H$ is -frame-bounded and $$\abs{UV}_{\ell_1}_\le \abs{U}_{\ell_1} \norm{V}.$$
We_point out_that we do_not know yet whether finite unions_or simple translates_of -frame-bounded sets are again -frame-bounded,_something_one would certainly_expect_to_hold for_a “good” boundedness_concept._Consequently, we_do_not know whether the set of_-frame-bounded_operators $X \to H$ form a vector_space.
\[lob.l.l1-bdd-basic\] Let $H$ be_any_Hilbert space and $M_\subseteq H$. Then the following_assertions hold.
\[item:label-lemma-a\] If $M$ is -frame-bounded,_then it_is norm-bounded,_with $$\sup_{x\in M} \norm{x |
Comparison of alternative QUBO formulations for Maximum Clique
The QUBO formulation was used for all previous investigations. This is due to the fact that, in contrast to, the former does not need the (usually unknown) clique size $K$ as an input parameter. Nevertheless, this section investigates whether the choice of formulation has an effect on the persistencies found and whether changing the QUBO formulation can result in more or fewer persistencies. If this is the case, then it may be worth investing time to search for formulations that are more suitable for the persistency algorithm.
In order to find the clique size that is needed for, we first run an alternative algorithm to find the maximum clique size $K$ for each tested graph. We then compute the QUBO coefficients for with parameter $K$ and identify strong and weak persistencies using Qpbo.
Table \[tab:qubo\_comparison\] shows results for a Hamming graph with parameters $n=8$ and $d=2$ and a c-fat graph with parameters $n=200$ and $c=1$. As can be seen from the table, formulation results in much larger QUBO problems (which are likely to also have a more complex structure, thus being more difficult to solve for Qpbo) and thus in fewer persistencies compared to, so clearly the latter formulation is preferable, at least for the type of graphs tested.
### Combining Qpbo with the decomposition algorithm {#sec:splitting_random}
We apply the graph splitting algorithm described in Section \[sec:decomp\] to solve the Maximum Clique problem on graphs of various sizes. However, before splitting a graph, we look for persistencies (in particular, for strong and weak persistencies since applying probing in each time step took prohibitively long) and, if found, reduce the size of the graph by removing the corresponding variables. We stop splitting when all generated (sub-)graphs are of size small enough to fit DW, which in our case is $45$ vertices.
We test the resulting algorithm on random graphs with $n=500$ vertices and an expected number of edges in the interval $[10000,40000]$.
![Proportional saving as a function of the expected number of edges of the random graph when using Qpbo in the graph splitting routine, computed as $(n_\text{qpbo}-n_\text{no-qpbo})/n_\text{ | Comparison of alternative QUBO formulations for Maximum Clique
The QUBO formulation was used for all former probe. This is due to the fact that, in contrast to, the erstwhile does not want the (usually unknown) clique size $ K$ as an input signal parameter. Nevertheless, this part investigates whether the choice of conceptualization has an effect on the persistencies found and whether exchange the QUBO formulation can result in more or fewer persistencies. If this is the case, then it may be deserving investing time to search for conceptualization that are more suitable for the persistency algorithm.
In holy order to find the clique size that is necessitate for, we first run an alternative algorithm to find the maximum clique size $ K$ for each test graph. We then compute the QUBO coefficients for with parameter $ K$ and identify strong and weak persistencies using Qpbo.
Table \[tab: qubo\_comparison\ ] shows results for a Hamming graph with parameters $ n=8 $ and $ d=2 $ and a c - fat graph with parameters $ n=200 $ and $ c=1$. As can be seen from the table, formulation results in a lot bigger QUBO problems (which are probable to besides have a more complex structure, thus being more difficult to resolve for Qpbo) and thus in fewer persistencies compared to, therefore clearly the latter formulation is preferable, at least for the type of graphs tested.
# # # Combining Qpbo with the decay algorithm { # sec: splitting_random }
We apply the graph splitting algorithm described in Section \[sec: decomp\ ] to solve the Maximum Clique problem on graphs of various sizes. However, before burst a graph, we look for persistencies (in particular, for strong and weak doggedness since applying probing in each time step took prohibitively long) and, if found, reduce the size of the graph by removing the corresponding variable. We stop separate when all generated (sub-)graphs are of size minor enough to fit DW, which in our event is $ 45 $ vertices.
We screen the resulting algorithm on random graph with $ n=500 $ vertices and an expected number of edges in the time interval $ [ 10000,40000]$.
! [ Proportional saving as a function of the expected number of edges of the random graph when using Qpbo in the graph splitting routine, computed as $ (n_\text{qpbo}-n_\text{no - qpbo})/n_\text { | Colparison of alternative DUBO formulatiouw for Kaximuj Clique
Ghe QUBO formulation was usev foe all previous investigatiovs. This id due to the dact that, mh contrast to, bhe fmcmer does not nged the (usuanly unknown) clhqje size $K$ as an input parameter. Nevereheless, tjis section indestpgwtes whether the choice of formulation has an effect on thr persistencies found and ahetjer changing the QKBO formulajjon xan result iv more or fewer persisjencies. If this is the case, then it mcy be worth ibvedjing time to searbh for formulations tvat are more suitable foc thw persistency algoritim.
In order to find tre clique snze that is needed foe, qe fitst rgn av algerhavivs algogitim to find fhe maximum clique size $K$ for tacr tested graph. We thqn compute the QUBO coefficients for with padameter $K$ and identify wtrong and weak persidtencies tsing Qpbo.
Table \[tab:qubo\_comparison\] shows results fos a Hemoiny grapf wlth parameters $n=8$ and $d=2$ and a c-fat graph with [zrsmvters $n=200$ and $c=1$. As gan be seen from tne tsfle, formulatiun resbmta in much larger QKBO proflems (which art likrly to also have a more complex structuge, tyus being more difyicult to souve gor Qlbo) and thus in fewer pzrsistsncies compwred to, sk clearly the latger fmrmulation is preferable, ae least fir tke type uf gtaphs tqsted.
### Comblning Qpbo with the decolposijion angorithm {#sfc:splitting_random}
We apply the gceph splitting angogithm deseribed in Section \[ses:decomp\] to solre the Mcximum Clique prkblem oi graphs of darious sizes. Jowever, befoce splittyng q grqph, we uuok for persisyencies (iu parricular, for strong ana weak persistenenew since applyinb pfobynh mn easv time step dook prunibitkvely long) cna, if found, reduce the sise or the graph by rempvlng the cirresponqing variablex. We stop splittinh whei all jenerayed (sub-)graphs are of size small ehough to vit DW, which in our case is $45$ vextices.
We test the resulting algorithm on random graphs wijh $n=500$ vertices and an gxptcted number pf eddes in tha interval $[10000,40000]$.
![Proportiobal saving as a fmnction of the expectes numbar of edges of the random graph when using Qpbo in the graph splitting routine, compuveq as $(n_\text{qlbo}-n_\yext{nm-q'bo})/u_\text{ | Comparison of alternative QUBO formulations for Maximum QUBO was used all previous investigations. fact in contrast to, former does not the (usually unknown) clique size $K$ an input parameter. Nevertheless, this section investigates whether the choice of formulation has effect on the persistencies found and whether changing the QUBO formulation can result more fewer If is the case, then it may be worth investing time to search for formulations that are suitable for the persistency algorithm. In order to the clique size that needed for, we first run alternative to find maximum size for each tested We then compute the QUBO coefficients for with parameter $K$ and identify strong and weak persistencies using Table \[tab:qubo\_comparison\] for a graph parameters and $d=2$ and graph with parameters $n=200$ and $c=1$. seen from the table, formulation results in much QUBO problems are likely to also have a complex structure, thus being more difficult to solve Qpbo) and thus in fewer persistencies compared to, so clearly the latter formulation is preferable, for the type of tested. ### Combining with decomposition {#sec:splitting_random} apply the splitting algorithm described in Section \[sec:decomp\] to solve the Maximum Clique on graphs of various sizes. However, before splitting a graph, for (in particular, for and weak persistencies since probing each time step took and, found, of graph removing the corresponding variables. stop splitting when all generated are of size small in our case is $45$ vertices. We test resulting algorithm on random graphs with $n=500$ and an expected number of edges in the interval $[10000,40000]$. ![Proportional saving a function expected number of edges of the random graph using Qpbo in the splitting routine, computed as $(n_\text{qpbo}-n_\text{no-qpbo})/n_\text{ | Comparison of alternative QUbO formulatIons fOr MAxiMuM CliQue
THe QUBO formulatIOn waS used for all previous invEstigAtIOns. THIs Is due To the faCT tHAT, in CoNtRasT tO, ThE formEr dOes not nEed the (usuaLly UnKnown) clique sIZe $k$ as an input ParAmeter. NevertHelEss, thiS sEctIOn invEstIgateS whethER the chOice of forMuLAtion hAS an effeCT On The pErsistencies found ANd WHether changing The QUBo fORmULAtiOn cAn result in MoRe or fEWer persIStENCIes. iF this is the casE, then it may bE WorTh inveStIng TIme to sEarch FoR ForMulations thAt arE more suitAble foR The persIStency aLgoritHm.
IN orDer tO FiNd The ClIQue SIzE thAT is Needed foR, wE fIrst rUn an ALTERnatIve AlgoRithm To find the maxiMum CliqUE siZe $K$ foR each TestEd Graph. we then CompuTe The QUBO coefficiEnts For with paRamEtEr $K$ AnD idenTIfy strOng And Weak perSistencIEs uSiNG qPbO.
Table \[tab:qubo\_compaRiSON\] sHows resuLts for A haMmINg graph wItH paRameTERs $n=8$ anD $d=2$ anD A c-Fat graph With paRAmEtErs $n=200$ and $C=1$. AS can be SeEn fRom The taBLe, foRmulatIon resulTs in mUCh larger QUBO prOBlems (which are LIkELY tO Also HavE a more complEx stRUctuRe, thUS bEinG More dIfficUlT To SOlve for Qpbo) and thus iN fEwer peRsistEncies compareD to, so clearLY THe latter FormULaTIon is preferablE, at leAst for the tYPe of grapHs tesTed.
### CombiNing Qpbo wITH the decoMpoSitIon AlgORItHm {#sec:splittinG_RAndoM}
WE apply tHe gRaph splIttIng AlgOriThM describeD in SectiOn \[SeC:dEcOmp\] To solVE the MaxiMuM ClIqUe pRobleM On grapHs of vArioUs SiZEs. HOwever, bEFoRE SpliTtInG a grAph, We Look fOr peRSisTencies (In particuLar, FOr stRoNg And weak Persistencies SiNce applyinG pRobIng in eACH time steP took prohibitively long) aND, if founD, reDuce tHe siZe of the grAph By remoVinG The corResponDing vArIabLES. We stOP SpLitTiNg when all gENEraTed (suB-)gRaphS are of sIze small enough to fiT dW, wHich in our case Is $45$ vErtiCES.
WE teST tHE reSuLTinG ALgorithm on randoM graphs witH $n=500$ VErTices and an EXpeCtEd numbeR of edgeS in thE IntervaL $[10000,40000]$.
![ProportiOnal savinG aS a fuNCTioN of the expeCted numbEr of edges OF the rANdOm graPh wHen usiNg qpbO in thE graph SPliTting RoutinE, cOmputeD as $(n_\tExT{qpbo}-n_\teXt{no-qpbo})/n_\text{ | Comparison of alternative QUBO form ulati ons fo rMaxi mumClique
The QU B O fo rmulation was used for allpr e viou s i nvest igation s .T h isis d ueto th e fac t t hat, in contrastto, t he former do e snot need t he(usually unk now n) cli qu e s i ze $K $ a s aninputp aramet er. Never th e less,t his sec t i on inv estigates whether th e choice of for mulati on ha s aneff ect on the p ersis t enciesf ou n d and whether chang ing the QUB O fo rmulat io n c a n resu lt in m o reor fewer pe rsis tencies.If thi s is the case, t hen it ma y b e wo r th i nve st i ngt im e t o se arch for f or mulat ions t h a t ar e m oresuita ble for the p ers iste n cyalgor ithm.
In o rderto fin d the c lique size that isneeded fo r,we fi rs t run an alt ern ati ve algo rithm t o fi nd t h emaximum clique siz e$ K $for each teste d g ra p h. We th en co mput e the Q UBOc oe fficient s forw it hparamet er $K$ a nd id ent ify s t rong and w eak pers isten c ies using Qpbo .
Table \[tab : qu b o \_ c ompa ris on\] showsresu l ts f or a Ha mmi n g gra ph wi th pa r ameters $n=8$ and $ d= 2$ and a c- fat graph wit h paramete r s $n=200$and$ c= 1 $. As can be s een f rom the ta b le, form ulati on resul ts in muc h larger Q UBO pr obl ems ( wh ich are likel y to a ls o havea m ore com ple x s tru ctu re , thus be ing more d if fi cu ltto so l ve for Q pb o)an d t hus i n fewer pers iste nc ie s co mparedt o, s o cl ea rl y th e l at ter f ormu l ati on is p referable , a t lea st f or thetype of graph stested.
# ## Co mbinin g Qpbo wit h the decomposition alg o rithm { #se c:spl itti ng_random }
We app lyt he gra ph spl ittin galg o r ithmd e sc rib ed in Sectio n \[s ec:de co mp\] to sol ve the Maximum Cli q ueproblem on gr aph s of v ar iou s s i zes .H owe v e r, before split ting a gra ph , w e look for per si stencie s (in p artic u lar, fo r strongand weakpe rsis t e nci es since a pplyingprobing i n each ti me st eptook p ro hib itive ly lon g ) a nd, i f foun d, reduc e the s ize of t he graph by removing th e corr espon din g variabl es. Westop spli ttin g when all ge ner ated(su b -)gra phsa re of sizesmal l enough t o f itD W ,which in ou r c ase is $ 45$ vertic es.
We test the resu l ting algorithm onr a ndo m g r aphs w ith $n=500$ ve rti ce s and an e xp ected numbe r of edg es in th e inte rval $ [10000, 4 0 00 0 ]$.
! [Pro por tional sa vin ga s a fun ct io n of th e ex pe cted n umbero f ed g e s of the randomgraph w hen u s ing Qpbo i n the g r aphsplittingroutine, co mputed as$(n_\ text{qp bo }-n_\t ext {n o-qpbo})/n _ \text{ | Comparison_of alternative_QUBO formulations for Maximum_Clique
The QUBO_formulation_was used_for_all previous investigations._This is due_to the fact that,_in contrast to,_the_former does not need the (usually unknown) clique size $K$ as an input parameter._Nevertheless,_this section_investigates_whether_the choice of formulation has_an effect on the persistencies_found and_whether changing the QUBO formulation can result in_more_or fewer persistencies._If this is the case, then it may be_worth investing time to search for_formulations that are_more_suitable_for the persistency algorithm.
In_order to find the clique size_that is needed for, we first_run an alternative algorithm to find the_maximum clique size $K$ for each_tested graph. We then compute_the QUBO_coefficients for with parameter $K$_and identify strong_and weak_persistencies using Qpbo.
Table \[tab:qubo\_comparison\]_shows results for a Hamming graph_with parameters $n=8$_and $d=2$ and a c-fat graph_with_parameters $n=200$ and_$c=1$._As_can be_seen from the_table,_formulation results_in_much larger QUBO problems (which are_likely_to also have a more complex structure,_thus being more difficult_to_solve for Qpbo) and_thus in fewer persistencies compared_to, so clearly the latter formulation_is preferable,_at least_for the type of graphs tested.
### Combining Qpbo with the decomposition_algorithm {#sec:splitting_random}
We apply the graph splitting_algorithm described in Section \[sec:decomp\]_to solve_the_Maximum Clique problem_on_graphs of_various sizes. However, before splitting a graph,_we look_for persistencies (in particular, for strong_and weak persistencies since_applying_probing in each time step took_prohibitively long) and, if found, reduce_the size of the graph_by_removing_the corresponding variables. We stop_splitting when all generated (sub-)graphs are_of size small_enough to fit DW, which in our_case_is $45$ vertices.
We test the resulting_algorithm_on random graphs with $n=500$ vertices_and_an_expected number of edges in_the interval $[10000,40000]$.
![Proportional saving as a_function of the expected number of edges of the_random graph when_using Qpbo in the graph_splitting_routine,_computed as $(n_\text{qpbo}-n_\text{no-qpbo})/n_\text{ |
omega}_{1}^-
\equiv
\mathcal{E}\,k_{\|}-\sqrt{k_{\bot}^2+(1+\mathcal{E}^2)\,k_{\|}^2}\,,$$ the scalar part of the propagator is $$\frac{1}{k\cdot\xi\,k\cdot\zeta+k^2+\mathrm{i}\epsilon}=\frac{1}{(k^0-\widetilde{\omega}_1^++\mathrm{i}\epsilon)(k^0-\widetilde{\omega}_1^--\mathrm{i}\epsilon)}\,.$$ The pole with positive real part can be cast in the following form: $$\frac{1}{k^0-\widetilde{\omega}_{1}^++\mathrm{i}\epsilon}=\mathcal{P}\frac{1}{k^0-\widetilde{\omega}_{1}^+}-\mathrm{i}\pi\,\delta(k^0-\widetilde{\omega}_{1}^+)\,.$$ Because of energy conservation only $\widetilde{\omega}_1^+$ and not $\widetilde{\omega}_1^-$ contributes to the imaginary part. We define $\widehat{\mathcal{M}}_{\widehat{1}}\equiv \mathcal{M}(e_L^-e_R^+\rightarrow\widetilde{\gamma})$ and obtain: $$\begin{aligned}
\label{eq:Optical-theorem-parity-odd}
2\,\mathrm{Im}(\mathcal{M}_{\,\widehat{1}})\big|_{\lambda=1}&=\int
\frac{\mathrm{d}^3k}{(2\pi)^{3}\,2\widetilde{\omega}_{1}^+}\,\delta^{(4)}(k_1+k_2-k) \notag \\
&\hspace{1.5cm}\times e^2\;\overline{u}(k_1)\gamma^{\nu}\frac{\mathds{1}-\gamma_5}{2}v(k_2)\;
\overline{v}(k_2)\gamma^{\mu}\frac{\mathds{1}
-\gamma_5}{2}u(k_1) \notag \\
&\hspace{1.5cm}\times
\frac{1}{N'}\;\big(-\eta_{\mu\nu}-\widehat{c}\,\xi_{\mu}\xi_{\nu}
-\widehat{e}\,\zeta_{\mu}\zeta_{\nu}
-\widehat{g}\,(\xi_{\mu}\zeta_{\nu}+\zeta_{\mu}\xi_{\nu})\big)
\notag \displaybreak[0]\\
| omega}_{1}^-
\equiv
\mathcal{E}\,k_{\|}-\sqrt{k_{\bot}^2+(1+\mathcal{E}^2)\,k_{\|}^2}\,,$$ the scalar part of the propagator is $ $ \frac{1}{k\cdot\xi\,k\cdot\zeta+k^2+\mathrm{i}\epsilon}=\frac{1}{(k^0-\widetilde{\omega}_1^++\mathrm{i}\epsilon)(k^0-\widetilde{\omega}_1 ^ --\mathrm{i}\epsilon)}\,.$$ The pole with positive real character can be roll in the following form: $ $ \frac{1}{k^0-\widetilde{\omega}_{1}^++\mathrm{i}\epsilon}=\mathcal{P}\frac{1}{k^0-\widetilde{\omega}_{1}^+}-\mathrm{i}\pi\,\delta(k^0-\widetilde{\omega}_{1}^+)\,.$$ Because of energy conservation only $ \widetilde{\omega}_1^+$ and not $ \widetilde{\omega}_1 ^ -$ contribute to the imaginary part. We specify $ \widehat{\mathcal{M}}_{\widehat{1}}\equiv \mathcal{M}(e_L^-e_R^+\rightarrow\widetilde{\gamma})$ and prevail: $ $ \begin{aligned }
\label{eq: Optical - theorem - parity bit - curious }
2\,\mathrm{Im}(\mathcal{M}_{\,\widehat{1}})\big|_{\lambda=1}&=\int
\frac{\mathrm{d}^3k}{(2\pi)^{3}\,2\widetilde{\omega}_{1}^+}\,\delta^{(4)}(k_1+k_2 - k) \notag \\
& \hspace{1.5cm}\times e^2\;\overline{u}(k_1)\gamma^{\nu}\frac{\mathds{1}-\gamma_5}{2}v(k_2)\;
\overline{v}(k_2)\gamma^{\mu}\frac{\mathds{1 }
-\gamma_5}{2}u(k_1) \notag \\
& \hspace{1.5cm}\times
\frac{1}{N'}\;\big(-\eta_{\mu\nu}-\widehat{c}\,\xi_{\mu}\xi_{\nu }
-\widehat{e}\,\zeta_{\mu}\zeta_{\nu }
-\widehat{g}\,(\xi_{\mu}\zeta_{\nu}+\zeta_{\mu}\xi_{\nu})\big)
\notag \displaybreak[0]\\ | omeha}_{1}^-
\equiv
\mathcal{E}\,k_{\|}-\sqrt{k_{\bot}^2+(1+\oathcal{E}^2)\,k_{\|}^2}\,,$$ the seqlar pert of fhe propxgator is $$\frac{1}{k\cdot\xi\,k\cdot\zeva+k^2+\mqthrm{u}\epsilon}=\frac{1}{(k^0-\widetilde{\umega}_1^++\mathgm{i}\epsilob)(k^0-\wivetilde{\omega}_1^--\mathcj{i}\epsilon)}\,.$$ The iole civh positive reak part can be cast in tha woplowing form: $$\frac{1}{k^0-\widetilde{\omega}_{1}^++\matrrm{i}\epsolln}=\mathcal{P}\frac{1}{h^0-\widttijde{\ojvgc}_{1}^+}-\mathrm{i}\pi\,\delta(k^0-\widetilde{\omega}_{1}^+)\,.$$ Gecause of energy comservation only $\widetilde{\olega}_1^+$ and not $\widetilde{\lmega}_1^-$ contrubutqw to the imaeinary paru. Ce define $\wjdehat{\mathcal{M}}_{\widehat{1}}\equiv \mathzal{M}(e_K^-e_R^+\rightartos\wlgetilde{\gamme})$ and jbtain: $$\begin{aligned}
\ldbel{eq:Oltical-theorem-psrivy-ode}
2\,\mathrm{Im}(\mathcal{M}_{\,\wideiat{1}})\big|_{\lambda=1}&=\int
\frac{\mwthrm{d}^3k}{(2\pi)^{3}\,2\fibetilde{\omega}_{1}^+}\,\delta^{(4)}(k_1+k_2-k) \borag \\
&\hvpaca{1.5cm}\tknes e^2\;\obeclihe{u}(k_1)\galma^{\iu}\frac{\mathda{1}-\gamma_5}{2}v(k_2)\;
\oveeline{v}(k_2)\gamma^{\mu}\frac{\mstres{1}
-\gamma_5}{2}u(k_1) \notzg \\
&\hspwcq{1.5cm}\times
\frac{1}{N'}\;\big(-\eta_{\mu\nu}-\widehat{c}\,\xi_{\mu}\xi_{\nu}
-\fidshat{e}\,\zeta_{\mu}\zeta_{\nu}
-\widehar{g}\,(\xi_{\mu}\zeta_{\nu}+\zeta_{\mu}\xi_{\no})\big)
\notag \qisplaybreak[0]\\
| omega}_{1}^- \equiv \mathcal{E}\,k_{\|}-\sqrt{k_{\bot}^2+(1+\mathcal{E}^2)\,k_{\|}^2}\,,$$ the scalar part of is The pole positive real part following $$\frac{1}{k^0-\widetilde{\omega}_{1}^++\mathrm{i}\epsilon}=\mathcal{P}\frac{1}{k^0-\widetilde{\omega}_{1}^+}-\mathrm{i}\pi\,\delta(k^0-\widetilde{\omega}_{1}^+)\,.$$ Because of conservation only $\widetilde{\omega}_1^+$ not $\widetilde{\omega}_1^-$ contributes to the imaginary We define $\widehat{\mathcal{M}}_{\widehat{1}}\equiv \mathcal{M}(e_L^-e_R^+\rightarrow\widetilde{\gamma})$ and obtain: $$\begin{aligned} \label{eq:Optical-theorem-parity-odd} 2\,\mathrm{Im}(\mathcal{M}_{\,\widehat{1}})\big|_{\lambda=1}&=\int \frac{\mathrm{d}^3k}{(2\pi)^{3}\,2\widetilde{\omega}_{1}^+}\,\delta^{(4)}(k_1+k_2-k) \notag \\ &\hspace{1.5cm}\times \overline{v}(k_2)\gamma^{\mu}\frac{\mathds{1} -\gamma_5}{2}u(k_1) \notag \\ &\hspace{1.5cm}\times \frac{1}{N'}\;\big(-\eta_{\mu\nu}-\widehat{c}\,\xi_{\mu}\xi_{\nu} -\widehat{e}\,\zeta_{\mu}\zeta_{\nu} -\widehat{g}\,(\xi_{\mu}\zeta_{\nu}+\zeta_{\mu}\xi_{\nu})\big) \notag \displaybreak[0]\\ | omega}_{1}^-
\equiv
\mathcal{E}\,k_{\|}-\sqrt{k_{\bOt}^2+(1+\mathcal{E}^2)\,K_{\|}^2}\,,$$ the sCalAr pArT of tHe prOpagator is $$\frac{1}{K\Cdot\Xi\,k\cdot\zeta+k^2+\mathrm{i}\epsIlon}=\fRaC{1}{(K^0-\widETiLde{\omEga}_1^++\mathRM{i}\EPSilOn)(K^0-\wIdeTiLDe{\Omega}_1^--\MatHrm{i}\epsIlon)}\,.$$ The polE wiTh Positive real PArT can be cast In tHe following fOrm: $$\Frac{1}{k^0-\wIdEtiLDe{\omeGa}_{1}^++\mAthrm{I}\epsilON}=\mathcAl{P}\frac{1}{k^0-\wIdETilde{\oMEga}_{1}^+}-\mathRM{I}\pI\,\delTa(k^0-\widetilde{\omega}_{1}^+)\,.$$ bEcAUse of energy conServatIoN OnLY $\WidEtiLde{\omega}_1^+$ anD nOt $\widETilde{\omEGa}_1^-$ CONTriBUtes to the imagInary part. We DEfiNe $\wideHaT{\maTHcal{M}}_{\wIdehaT{1}}\eQUiv \Mathcal{M}(e_L^-e_r^+\rigHtarrow\wiDetildE{\Gamma})$ anD Obtain: $$\bEgin{alIgnEd}
\lAbel{EQ:OPtIcaL-tHEorEM-pAriTY-odD}
2\,\mathrm{IM}(\mAtHcal{M}_{\,\WideHAT{1}})\BIg|_{\laMbdA=1}&=\int
\Frac{\mAthrm{d}^3k}{(2\pi)^{3}\,2\wideTilDe{\omEGa}_{1}^+}\,\dElta^{(4)}(k_1+K_2-k) \notAg \\
&\hsPaCe{1.5cm}\tImes e^2\;\oVerliNe{U}(k_1)\gamma^{\nu}\frac{\maThds{1}-\Gamma_5}{2}v(k_2)\;
\ovErlInE{v}(k_2)\GaMma^{\mu}\FRac{\matHds{1}
-\GamMa_5}{2}u(k_1) \notAg \\
&\hspacE{1.5Cm}\tImES
\FRaC{1}{N'}\;\big(-\eta_{\mu\nu}-\widehaT{c}\,\XI_{\Mu}\Xi_{\nu}
-\wideHat{e}\,\zeTA_{\mU}\zETa_{\nu}
-\wideHaT{g}\,(\xI_{\mu}\zETA_{\nu}+\zeTa_{\mu}\XI_{\nU})\big)
\notaG \displAYbReAk[0]\\
| omega}_{1}^-
\equiv
\mathc al{E}\,k_{ \|}-\ sqr t{k _{ \bot }^2+ (1+\mathcal{E} ^ 2)\, k_{\|}^2}\,,$$ the sca lar p ar t oft he prop agatori s$ $ \fr ac {1 }{k \c d ot \xi\, k\c dot\zet a+k^2+\mat hrm {i }\epsilon}=\ f ra c{1}{(k^0- \wi detilde{\ome ga} _1^++\ ma thr m {i}\e psi lon)( k^0-\w i detild e{\omega} _1 ^ --\mat h rm{i}\e p s il on)} \,.$$ The pole wi t hp ositive real p art ca nb ec a stinthe follow in g for m : $$\fr a c{ 1 } { k^0 - \widetilde{\o mega}_{1}^+ + \ma thrm{i }\ eps i lon}=\ mathc al { P}\ frac{1}{k^0 -\wi detilde{\ omega} _ {1}^+}- \ mathrm{ i}\pi\ ,\d elt a(k^ 0 -\ wi det il d e{\ o me ga} _ {1} ^+)\,.$$ B ec auseof e n e r g y co nse rvat ion o nly $\widetil de{ \ome g a}_ 1^+$and n ot $ \w ideti lde{\o mega} _1 ^-$ contributes tothe imagi nar ypar t. We d e fine $ \wi deh at{\mat hcal{M} } _{\ wi d e h at {1}}\equiv \mathca l{ M } (e _L^-e_R^ +\righ t ar ro w \widetil de {\g amma } ) $ and obt a in : $$\beg in{ali g ne d}
\label {e q:Opti ca l-t heo rem-p a rity -odd}2\,\math rm{Im } (\mathcal{M}_{ \ ,\widehat{1}} ) \b i g |_ { \lam bda =1}&=\int
\ frac { \mat hrm{ d }^ 3k} { (2\pi )^{3} \, 2 \w i detilde{\omega}_{1} ^+ }\,\de lta^{ (4)}(k_1+k_2- k) \notag\ \ &\hspace {1.5 c m} \ times e^2\;\ov erlin e{u}(k_1)\ g amma^{\n u}\fr ac{\math ds{1}-\ga m m a_5}{2}v (k_ 2)\ ;
\ ove r l in e{v}(k_2)\gam m a ^{\m u} \frac{\ mat hds{1}-\g amm a_5 }{2 }u (k_1) \no tag \\
& \h sp ac e{ 1.5 cm}\t i mes
\fra c{ 1}{ N' }\; \big( - \eta_{ \mu\n u}-\ wi de h at{ c}\,\xi _ {\ m u }\xi _{ \n u}
- \wi de hat{e }\,\ z eta _{\mu}\ zeta_{\nu }
- \ wide ha t{ g}\,(\x i_{\mu}\zeta_ {\ nu}+\zeta_ {\ mu} \xi_{\ n u })\big)\notag \displaybreak[0] \ \
| omega}_{1}^-
\equiv
\mathcal{E}\,k_{\|}-\sqrt{k_{\bot}^2+(1+\mathcal{E}^2)\,k_{\|}^2}\,,$$ the_scalar part_of the propagator is_$$\frac{1}{k\cdot\xi\,k\cdot\zeta+k^2+\mathrm{i}\epsilon}=\frac{1}{(k^0-\widetilde{\omega}_1^++\mathrm{i}\epsilon)(k^0-\widetilde{\omega}_1^--\mathrm{i}\epsilon)}\,.$$ The_pole_with positive_real_part can be_cast in the_following form: $$\frac{1}{k^0-\widetilde{\omega}_{1}^++\mathrm{i}\epsilon}=\mathcal{P}\frac{1}{k^0-\widetilde{\omega}_{1}^+}-\mathrm{i}\pi\,\delta(k^0-\widetilde{\omega}_{1}^+)\,.$$ Because_of energy conservation_only_$\widetilde{\omega}_1^+$ and not $\widetilde{\omega}_1^-$ contributes to the imaginary part. We define $\widehat{\mathcal{M}}_{\widehat{1}}\equiv \mathcal{M}(e_L^-e_R^+\rightarrow\widetilde{\gamma})$ and_obtain:_$$\begin{aligned}
\label{eq:Optical-theorem-parity-odd}
2\,\mathrm{Im}(\mathcal{M}_{\,\widehat{1}})\big|_{\lambda=1}&=\int
\frac{\mathrm{d}^3k}{(2\pi)^{3}\,2\widetilde{\omega}_{1}^+}\,\delta^{(4)}(k_1+k_2-k) \notag_\\
&\hspace{1.5cm}\times_e^2\;\overline{u}(k_1)\gamma^{\nu}\frac{\mathds{1}-\gamma_5}{2}v(k_2)\;
\overline{v}(k_2)\gamma^{\mu}\frac{\mathds{1}
-\gamma_5}{2}u(k_1)_\notag \\
&\hspace{1.5cm}\times
\frac{1}{N'}\;\big(-\eta_{\mu\nu}-\widehat{c}\,\xi_{\mu}\xi_{\nu}
-\widehat{e}\,\zeta_{\mu}\zeta_{\nu}
-\widehat{g}\,(\xi_{\mu}\zeta_{\nu}+\zeta_{\mu}\xi_{\nu})\big)
\notag \displaybreak[0]\\
|
display of all past [*chameleon* ]{}posts.
During (5) *delivery* and (6) the second *maturation* phase, the refreshed [*chameleon* ]{}profiles (avatars) contact the target and build trust with it. The (7) *execution* phase in this misuse case depends on the attacker’s goals. The avatars that already engaged in an attack will likely be discarded.
#### Potential Impact
The adversary does not have to create an OSN account and build an appropriate agenda for each avatar long before executing an attack. *Chameleon* profiles and posts are created and maintained as a general resource suitable for various attack campaigns. As a result, the cost of maintaining such avatars is dramatically reduced. Moreover, if an avatar is detected and blocked during the attack campaign, its replacement can be *weaponized* and released very quickly.
### Evading Censorship
OSNs maintain millions of entities, such as pages, groups, communities, etc. For example, Facebook groups unite users based on shared interests [@casteleyn2009use]. To ensure proper language, avoid trolling and abuse, or allow in only users with a very specific agenda, moderators inspect the users who ask to join the groups and review the published posts. *Chameleon* attack can help in evading censorship, as well as a shallow screening of OSN profiles. See Section \[sec:mitigation\] for specific recommendations on profile screening to detect [*chameleon* ]{}profiles.
For example, assume two Facebook groups, uniting Democrat and Republican activists during U.S. elections. Assume a dishonest activist from one political extreme that would like to join a Facebook group of the rivals. Reasons may vary from trolling to spying. Assume, that this activist would like to spread propaganda within the rival group. But pages that exhibit an agenda that is not appropriate for the group would not be allowed by the group owner. The next procedure would allow the rival activist to bypass the censorship of the group moderator.
#### Execution Flow
During the (1) *reconnaissance* phase, the adversary learns the censorship rules of the target. (2) The *weaponizing* phase includes establishing a [*chameleon* ]{}profile with agenda appropriate to the censorship. During the (3) *maturation* phase, the adversary publ | display of all past [ * chameleon * ] { } posts.
During (5) * delivery * and (6) the second * growth * phase, the freshen [ * chameleon * ] { } profiles (avatars) touch the target and build reliance with it. The (7) * execution * phase in this misuse case depend on the attacker ’s goals. The avatars that already hire in an attack will likely be discarded.
# # # # Potential Impact
The adversary does not have to create an OSN account and construct an appropriate agenda for each avatar long before executing an attack. * chamaeleon * profiles and posts are create and maintained as a cosmopolitan resource suitable for versatile attack campaigns. As a result, the monetary value of maintaining such avatars is dramatically reduced. Moreover, if an avatar is detected and blocked during the attack campaign, its replacement can be * weaponized * and released very quickly.
# # # Evading Censorship
OSNs maintain millions of entity, such as page, groups, community, etc. For exercise, Facebook groups unite drug user based on shared interests [ @casteleyn2009use ]. To ensure proper linguistic process, avoid trolling and abuse, or allow in only users with a very specific agenda, moderator inspect the users who ask to join the group and review the published position. * chamaeleon * attack can help in evading censorship, as well as a shallow screening of OSN profiles. See incision \[sec: mitigation\ ] for specific recommendations on profile screening to detect [ * chameleon * ] { } visibility.
For example, simulate two Facebook groups, uniting Democrat and Republican militant during U.S. elections. Assume a dishonest activist from one political extreme point that would like to join a Facebook group of the rival. Reasons may vary from trolling to spying. Assume, that this activist would wish to spread propaganda within the rival group. But pages that exhibit an agenda that is not appropriate for the group would not be allow by the group owner. The next procedure would allow the rival activist to bypass the censorship of the group moderator.
# # # # performance Flow
During the (1) * reconnaissance * phase, the adversary learns the censoring rules of the target. (2) The * weaponizing * phase admit establishing a [ * chameleon * ] { } profile with agenda appropriate to the censoring. During the (3) * maturation * phase, the adversary publ | didplay of all past [*chamelton* ]{}posts.
During (5) *bwliverb* and (6) fhe secovd *maturation* phase, the refrxshee [*chaneleon* ]{}profiles (avatarr) contact the tartet end build trust xjth it. Bke (7) *esccutimi* phase in this misuse cave depends on dhd cttacker’s goals. The avatars that alrqady enbahed in an attask wplj liivln be discarded.
#### Potential Impact
Tge advegsary does not hafe to create an OSN accounh anf build an approprlate agenda for wach avatar uong befort zxecuting ah attack. *Chameleon* profiles and postx are creajze ajg maintainev as a general resource suhtable gor various atback rampqigns. As a result, the cost of maintainigg such aeacars is dramatically eeeuced. Moraovef, if an atatzr is fetxcted and bmocked duribg the attack campaogg, its replacemsnt cag fe *weaponized* and released very quickly.
### Evzding Censorship
OSNs mauntain millions of enjities, sucr as pages, groups, communities, etc. For example, Facabook erobif ynlte users based on shared interests [@casteleyn2009ufs]. Uo vnsure proper lannuage, avoid trollimg amq abuse, or aluow in onmy users with a vegy specyfic qgenda, moqerayors inspect the users who qsk to join nhe troups and review che publisheb postx. *Chakeleon* attack can help nn evasing censordhip, as wsul as a shallow rcrvenitg of OSN profiles. See Seceion \[sec:mivigatnon\] for rpecofic rqcommendatlons on profile screeninh to betecd [*chameleoj* ]{}profiles.
For example, assume two Facebook groops, unpting Demjcrat and Republicwn activists doring U.S. zlectiuns. Assume a dishmnest activyst from one [llitical extceme that wouod luke to guin a Facebook group of the rivalw. Reasons may vary fruj trolling to s'vibg. Assume, that yhir astpvixt fould like tm spfeaa propxganda withln ghe tival group. But pagev thzt exhibit an agenca that is not appwopriate for yhe group would nou be anloxed by thg group owner. The next procedurs would aplor the rival wctiyist to bypass the censorship of the group moderator.
#### Xxecution Flow
During thw (1) *reconnaissance* pkaxe, the advecsary jearns tha censorship rules od the target. (2) The *weaponizing* phase indludes estahlishing a [*chameleon* ]{}profile with agenda appropriate to the censorship. Dueing tie (3) *maturatioh* phsse, tve cdversawy pnbu | display of all past [*chameleon* ]{}posts. During and the second phase, the refreshed target build trust with The (7) *execution* in this misuse case depends on attacker’s goals. The avatars that already engaged in an attack will likely be #### Potential Impact The adversary does not have to create an OSN account build appropriate for avatar long before executing an attack. *Chameleon* profiles and posts are created and maintained as a resource suitable for various attack campaigns. As a the cost of maintaining avatars is dramatically reduced. Moreover, an is detected blocked the campaign, its replacement be *weaponized* and released very quickly. ### Evading Censorship OSNs maintain millions of entities, such as pages, communities, etc. Facebook groups users on interests [@casteleyn2009use]. To language, avoid trolling and abuse, or users with a very specific agenda, moderators inspect users who to join the groups and review published posts. *Chameleon* attack can help in evading as well as a shallow screening of OSN profiles. See Section \[sec:mitigation\] for specific recommendations screening to detect [*chameleon* For example, assume Facebook uniting and activists during elections. Assume a dishonest activist from one political extreme that would to join a Facebook group of the rivals. Reasons may trolling spying. Assume, that activist would like to propaganda the rival group. But exhibit agenda appropriate the would not be allowed the group owner. The next would allow the rival of the group moderator. #### Execution Flow During (1) *reconnaissance* phase, the adversary learns the rules of the target. (2) The *weaponizing* phase includes establishing a [*chameleon* with agenda the censorship. During the (3) *maturation* phase, the publ | display of all past [*chameleon* ]{}Posts.
DurinG (5) *deliVerY* anD (6) tHe seCond *Maturation* phasE, The rEfreshed [*chameleon* ]{}profiLes (avAtARs) coNTaCt the Target aND bUILd tRuSt WitH iT. thE (7) *execUtiOn* phase In this misuSe cAsE depends on thE AtTacker’s goaLs. THe avatars thaT alReady eNgAgeD In an aTtaCk wilL likelY Be discArded.
#### PoteNtIAl ImpaCT
The advERSaRy doEs not have to create AN Osn account and buiLd an apPrOPrIATe aGenDa for each aVaTar loNG before EXeCUTIng AN attack. *ChamelEon* profiles ANd pOsts arE cReaTEd and mAintaInED as A general resOurcE suitable For varIOus attaCK campaiGns. As a ResUlt, The cOSt Of MaiNtAIniNG sUch AVatArs is draMaTiCally ReduCED. mOreoVer, If an AvataR is detected anD blOckeD DurIng thE attaCk caMpAign, iTs replAcemeNt Can be *weaponized* And rEleased veRy qUiCklY.
### EVadinG censorShiP
OSns maintAin millIOns Of ENTItIes, such as pages, grouPs, COMmUnities, eTc. For eXAmPlE, facebook GrOupS uniTE Users BaseD On Shared inTerestS [@CaStEleyn2009usE]. TO ensurE pRopEr lAnguaGE, avoId trolLing and aBuse, oR Allow in only useRS with a very speCIfIC AgENda, mOdeRators inspeCt thE UserS who ASk To jOIn the GroupS aND rEView the published posTs. *chamelEon* atTack can help in Evading cenSORShip, as weLl as A ShALlow screening oF OSN pRofiles. See sEction \[seC:mitiGation\] foR specific RECommendaTioNs oN prOfiLE ScReening to deteCT [*ChamElEon* ]{}profIleS.
For exaMplE, asSumE twO FAcebook grOups, unitInG DEmOcRat And RePUblican aCtIviStS duRing U.s. ElectiOns. AsSume A dIsHOneSt activISt FROm onE pOlIticAl eXtReme tHat wOUld Like to jOin a FacebOok GRoup Of ThE rivals. reasons may varY fRom trollinG tO spYing. AsSUMe, that thIs activist would like to spREad propAgaNda wiThin The rival gRouP. But paGes THat exhIbit an AgendA tHat IS Not apPROpRiaTe For the grouP WOulD not bE aLlowEd by the Group owner. The next pROceDure would alloW thE rivAL AcTivISt TO byPaSS thE CEnsorship of the gRoup moderaToR.
#### exEcution FloW
durInG the (1) *recOnnaissAnce* pHAse, the aDversary lEarns the cEnSorsHIP ruLes of the taRget. (2) The *wEaponizinG* Phase INcLudes EstAblishInG a [*cHamelEon* ]{}proFIle With aGenda aPpRopriaTe to tHe CensorshIp. During the (3) *maturation* phAse, the AdverSarY publ | display of all past [*cha meleon* ]{ }post s.
Du ri ng ( 5) * delivery* and( 6) t he second *maturation* phas e, ther ef reshe d [*cha m el e o n*]{ }p rof il e s(avat ars ) conta ct the tar get a nd build tru s twith it. T he(7) *executi on* phase i n t h is mi sus e cas e depe n ds onthe attac ke r ’s goa l s. Thea v at arsthat already enga g ed in an attack w ill li ke l yb e di sca rded.
### #Poten t ial Imp a ct T hea dversary does not have t o cr eate a nOSN accoun t and b u ild an appropr iate agenda f or eac h avatar long be fore e xec uti ng a n a tt ack .* Cha m el eon * pr ofiles a nd p ostsarec r e a tedand mai ntain ed as a gener alreso u rce suit ableforva rious attac k cam pa igns. As a resu lt,the costofma int ai nings uch av ata rsis dram aticall y re du c e d .Moreover, if an av at a r i s detect ed and bl oc k ed durin gthe att a c k cam paig n ,its repl acemen t c an be *we ap onized *and re lease d ver y quic kly.
## # Eva d ing Censorship
OSNs maintai n m i l li o ns o f e ntities, su ch a s pag es,g ro ups , comm uniti es , e t c. For example, Fac eb ook gr oupsunite users b ased on sh a r e d intere sts[ @c a steleyn2009use ]. To ensure pr o per lang uage, avoid t rolling a n d abuse,orall owino n ly users with a v erysp ecificage nda, mo der ato rsins pe ct the us ers whoas kto j oin theg roups an drev ie w t he pu b lished post s. * Ch am e leo n* atta c kc a n he lp i n ev adi ng cens orsh i p,as well as a sha llo w scr ee ni ng of O SN profiles.Se e Section\[ sec :mitig a t ion\] fo r specific recommendati o ns on p rof ile s cree ning to d ete ct [*c ham e leon*]{}pro files .
Fo r examp l e ,ass um e two Face b o okgroup s, uni ting De mocrat and Republi c anactivists dur ing U.S . el ect i on s . A ss u mea dishonest activ ist from o ne po litical ex t rem ethat wo uld lik e toj oin a F acebook g roup of t he riv a l s.Reasons ma y vary f rom troll i ng to sp ying. As sume,th atthisactivi s t w ouldlike t ospread prop ag anda wit hin the rival group. Bu t page s tha t e xhibit an ag e nda that isnotappropriat e f orthe g rou p woul d no t b e a l lowed byt he groupo wn er. T he next proce d u r e w ouldall o w theriva l activist to byp a ss the censors hipo f th e g r oupmo derator.
#### Ex ec u t ion Flow
During the(1) *rec on n aissa nce* p hase,the adv e r sa r y lear ns t hecensorshi p r ul e s of th eta r get. ( 2) T he *weap onizin g * ph a s e includes estab lishi n g a [* c ham eleon *]{}prof i le w ith agenda appropriat e to t he c ensor ship. D ur ing th e ( 3) *maturati o n* phase, theadversa ry pub l | display_of all_past [*chameleon* ]{}posts.
During (5)_*delivery* and_(6)_the second_*maturation*_phase, the refreshed_[*chameleon* ]{}profiles (avatars)_contact the target and_build trust with_it._The (7) *execution* phase in this misuse case depends on the attacker’s goals. The_avatars_that already_engaged_in_an attack will likely be_discarded.
#### Potential Impact
The adversary does_not have_to create an OSN account and build an_appropriate_agenda for each_avatar long before executing an attack. *Chameleon* profiles and_posts are created and maintained as_a general resource_suitable_for_various attack campaigns. As_a result, the cost of maintaining_such avatars is dramatically reduced. Moreover,_if an avatar is detected and blocked_during the attack campaign, its replacement_can be *weaponized* and released_very quickly.
###_Evading Censorship
OSNs maintain millions of_entities, such as_pages, groups,_communities, etc. For_example, Facebook groups unite users based_on shared interests [@casteleyn2009use]._To ensure proper language, avoid trolling_and_abuse, or allow_in_only_users with_a very specific_agenda,_moderators inspect_the_users who ask to join the_groups_and review the published posts. *Chameleon* attack_can help in evading_censorship,_as well as a_shallow screening of OSN profiles._See Section \[sec:mitigation\] for specific recommendations on_profile screening_to detect_[*chameleon* ]{}profiles.
For example, assume two Facebook groups, uniting Democrat and Republican_activists during U.S. elections. Assume a_dishonest activist from one_political extreme_that_would like to_join_a Facebook_group of the rivals. Reasons may vary_from trolling_to spying. Assume, that this activist_would like to spread_propaganda_within the rival group. But pages_that exhibit an agenda that is_not appropriate for the group_would_not_be allowed by the group_owner. The next procedure would allow_the rival activist_to bypass the censorship of the group_moderator.
####_Execution Flow
During the (1) *reconnaissance* phase,_the_adversary learns the censorship rules of_the_target._(2) The *weaponizing* phase includes_establishing a [*chameleon* ]{}profile with agenda_appropriate to the censorship. During the (3) *maturation* phase,_the adversary publ |
)^{w_j}}{2},\label{frmcn}$$ where $n$ has the prime factorization $$n=2^{{\alpha}}3^{{\beta}}\prod_{i=1}^{r}p_i^{v_i}\prod_{j=1}^sq_j^{w_j},$$ where $p_i \equiv 1 {\,(\textup{mod}\,3)}$ and $2\neq q_i \equiv 2 {\,(\textup{mod}\,3)}$.
From, we have $$\label{sevdi}
a(n)={\dfrac}{c(n)-c(n/3)}{3}+c(n/9)+{\dfrac}{4}{3}d(n),$$ where we assume $c(n/l)=0$ if $l\not|\, n$. If $n \not \equiv 1
{\,(\textup{mod}\,6)}$, then $d(n)\equiv 0$ and $a(n)=c(n/9)$ if $3|n$, while $a(n)=c(n)/3$ if $3\not |n$. This proves the claim in for $a(n)$.
Now assume $n \equiv 1 {\,(\textup{mod}\,6)}$. If $p$ is a prime and $p \equiv 1
{\,(\textup{mod}\,3)}$, then by we see that $c(p)=4$. If $p$ is represented by the form $(1,0,27)$, then $a(p)=4$. This is because $4 \leq a(p)
\leq c(p)=4$. Using we see that $d(p)=2$. If $p$ is not represented by $(1,0,27)$ then, by, $d(p)=-1$. Gauss proved that if $p$ is a prime and $p \equiv 1 {\,(\textup{mod}\,3)}$, then $p$ is represented by $(1,0,27)$ iff $2$ is a cubic residue modulo $p$ or, equivalently, iff $2^{(p-1)/3} \equiv 1 {\,(\textup{mod}\,p)}$. Therefore,
$$\label{dtable}
d(p)=\left\{ \begin{array}{ll}
2 & \text{if}\;\; p \equiv 1 {\,(\textup{mod}\,3)} \;\;\text{and}\;\; 2^{( | ) ^{w_j}}{2},\label{frmcn}$$ where $ n$ has the prime factorization $ $ n=2^{{\alpha}}3^{{\beta}}\prod_{i=1}^{r}p_i^{v_i}\prod_{j=1}^sq_j^{w_j},$$ where $ p_i \equiv 1 { \,(\textup{mod}\,3)}$ and $ 2\neq q_i \equiv 2 { \,(\textup{mod}\,3)}$.
From, we have $ $ \label{sevdi }
a(n)={\dfrac}{c(n)-c(n/3)}{3}+c(n/9)+{\dfrac}{4}{3}d(n),$$ where we assume $ c(n / l)=0 $ if $ l\not|\, n$. If $ n \not \equiv 1
{ \,(\textup{mod}\,6)}$, then $ d(n)\equiv 0 $ and $ a(n)=c(n/9)$ if $ 3|n$, while $ a(n)=c(n)/3 $ if $ 3\not |n$. This testify the title in for $ a(n)$.
Now assume $ n \equiv 1 { \,(\textup{mod}\,6)}$. If $ p$ is a flower and $ p \equiv 1
{ \,(\textup{mod}\,3)}$, then by we see that $ c(p)=4$. If $ p$ is represented by the phase $ (1,0,27)$, then $ a(p)=4$. This is because $ 4 \leq a(p)
\leq c(p)=4$. Using we see that $ d(p)=2$. If $ p$ is not represent by $ (1,0,27)$ then, by, $ d(p)=-1$. Gauss proved that if $ p$ is a prime and $ phosphorus \equiv 1 { \,(\textup{mod}\,3)}$, then $ p$ is represented by $ (1,0,27)$ iff $ 2 $ is a cubic residue modulo $ p$ or, equivalently, iff $ 2^{(p-1)/3 } \equiv 1 { \,(\textup{mod}\,p)}$. Therefore,
$ $ \label{dtable }
d(p)=\left\ { \begin{array}{ll }
2 & \text{if}\;\; p \equiv 1 { \,(\textup{mod}\,3) } \;\;\text{and}\;\; 2^ { ( | )^{w_j}}{2},\lwbel{frmcn}$$ where $n$ has tht prime factorizajiin $$n=2^{{\al'ha}}3^{{\beta}}\lrod_{i=1}^{r}p_i^{x_i}\prod_{j=1}^sq_j^{w_j},$$ where $p_i \equiv 1 {\,(\twxtup{nod}\,3)}$ and $2\neq q_i \equiv 2 {\,(\gextup{mod}\,3)}$.
Vrom, we yave $$\oabel{sevdi}
e(h)={\dfrac}{c(k)-e(n/3)}{3}+c(n/9)+{\drvac}{4}{3}d(n),$$ xhere we assume $c(n/l)=0$ if $l\nmt|\, n$. If $n \not \aqjir 1
{\,(\textup{mod}\,6)}$, then $d(n)\equiv 0$ and $a(n)=c(n/9)$ is $3|n$, whike $a(n)=c(n)/3$ if $3\not |n$. Thix probvs the claim in for $a(n)$.
Now assums $n \equpv 1 {\,(\textup{mod}\,6)}$. If $l$ is a prime and $p \equiv 1
{\,(\tfxtuo{mod}\,3)}$, then by we sef that $c(p)=4$. Id $p$ yw representea by the form $(1,0,27)$, then $a(p)=4$. This is because $4 \leq a(p)
\leq c(p)=4$. Jsing we see thqt $d(p)=2$. Hf $p$ is not reprvsented by $(1,0,27)$ tmvn, by, $d([)=-1$. Gauss proved that in $p$ iv a prime and $p \equiv 1 {\,(\txxtup{mod}\,3)}$, then $p$ is rgpresented bv $(1,0,27)$ iff $2$ is a cubic rewieue mmdulm $p$ ue, eduibakehtly, ivf $2^{('-1)/3} \equiv 1 {\,(\testup{mod}\,p)}$. Thwrefore,
$$\label{dtable}
d(k)=\lesn\{ \begin{array}{lm}
2 & \text{if}\;\; p \equiv 1 {\,(\textup{mod}\,3)} \;\;\text{and}\;\; 2^{( | )^{w_j}}{2},\label{frmcn}$$ where $n$ has the prime factorization $p_i 1 {\,(\textup{mod}\,3)}$ $2\neq q_i \equiv $$\label{sevdi} where we assume if $l\not|\, n$. $n \not \equiv 1 {\,(\textup{mod}\,6)}$, then 0$ and $a(n)=c(n/9)$ if $3|n$, while $a(n)=c(n)/3$ if $3\not |n$. This proves the in for $a(n)$. Now assume $n \equiv 1 {\,(\textup{mod}\,6)}$. If $p$ is a and \equiv {\,(\textup{mod}\,3)}$, by we see that $c(p)=4$. If $p$ is represented by the form $(1,0,27)$, then $a(p)=4$. This because $4 \leq a(p) \leq c(p)=4$. Using we that $d(p)=2$. If $p$ not represented by $(1,0,27)$ then, $d(p)=-1$. proved that $p$ a and $p \equiv {\,(\textup{mod}\,3)}$, then $p$ is represented by $(1,0,27)$ iff $2$ is a cubic residue modulo $p$ or, equivalently, $2^{(p-1)/3} \equiv Therefore, $$\label{dtable} \begin{array}{ll} & p \equiv 1 2^{( | )^{w_j}}{2},\label{frmcn}$$ where $n$ has the pRime factorIzatiOn $$n=2^{{\AlpHa}}3^{{\Beta}}\Prod_{I=1}^{r}p_i^{v_i}\prod_{j=1}^sq_j^{W_J},$$ wheRe $p_i \equiv 1 {\,(\textup{mod}\,3)}$ and $2\nEq q_i \eQuIV 2 {\,(\texTUp{Mod}\,3)}$.
FrOm, we havE $$\LaBEL{seVdI}
a(N)={\dfRaC}{C(n)-C(n/3)}{3}+c(n/9)+{\dFraC}{4}{3}d(n),$$ wherE we assume $c(N/l)=0$ iF $l\Not|\, n$. If $n \not \eqUIv 1
{\,(\Textup{mod}\,6)}$, tHen $D(n)\equiv 0$ and $a(n)=C(n/9)$ iF $3|n$, whilE $a(N)=c(n)/3$ IF $3\not |n$. thiS provEs the cLAim in fOr $a(n)$.
Now asSuME $n \equiV 1 {\,(\Textup{mOD}\,6)}$. if $P$ is a Prime and $p \equiv 1
{\,(\texTUp{MOd}\,3)}$, then by we see tHat $c(p)=4$. IF $p$ IS rEPResEntEd by the forM $(1,0,27)$, tHen $a(p)=4$. tHis is beCAuSE $4 \LEq a(P)
\Leq c(p)=4$. Using we sEe that $d(p)=2$. If $p$ IS noT repreSeNteD By $(1,0,27)$ then, By, $d(p)=-1$. GAuSS prOved that if $p$ Is a pRime and $p \eQuiv 1 {\,(\teXTup{mod}\,3)}$, tHEn $p$ is rePresenTed By $(1,0,27)$ iFf $2$ is A CuBiC reSiDUe mODuLo $p$ OR, eqUivalentLy, IfF $2^{(p-1)/3} \equIv 1 {\,(\teXTUP{Mod}\,p)}$. theRefoRe,
$$\labEl{dtable}
d(p)=\lefT\{ \beGin{aRRay}{Ll}
2 & \texT{if}\;\; p \eQuiv 1 {\,(\TeXtup{mOd}\,3)} \;\;\text{And}\;\; 2^{( | )^{w_j}}{2},\label{frmcn}$ $ where $n $ has th e p ri me f acto rization $$n=2 ^ {{\a lpha}}3^{{\beta}}\prod _{i=1 }^ { r}p_ i ^{ v_i}\ prod_{j = 1} ^ s q_j ^{ w_ j}, $$ wh ere $ p_i \equiv 1 {\,(\te xtu p{ mod}\,3)}$ a n d$2\neq q_i \e quiv 2 {\,(\ tex tup{mo d} \,3 ) }$.
Fro m, we have$ $\labe l{sevdi}a( n )={\df r ac}{c(n ) - c( n/3) }{3}+c(n/9)+{\dfr a c} { 4}{3}d(n),$$ w here w ea ss u m e $ c(n /l)=0$ if$l \not| \ , n$. I f $ n \ not \equiv 1
{\,( \textup{mod } \,6 )}$, t he n $ d (n)\eq uiv 0 $a nd$a(n)=c(n/9 )$ i f $3|n$,while$ a(n)=c( n )/3$ if $3\no t | n$. Thi s p ro ves t h e c l ai m i n fo r $a(n)$ .
N ow as sume $ n \equ iv1 {\ ,(\te xtup{mod}\,6) }$. If$ p$is aprime and $ p \eq uiv 1{\,(\ te xtup{mod}\,3)}$ , th en by wesee t hat $ c(p)= 4 $. If$p$ is repres ented b y th ef o r m$(1,0,27)$, then $ a( p ) =4 $. Thisis bec a us e$ 4 \leq a (p )
\ leqc ( p)=4$ . Us i ng we seethat $ d (p )= 2$. If$p $ is n ot re pre sente d by$(1,0, 27)$ the n, by , $d(p)=-1$. Ga u ss proved tha t i f $p $ isa p rime and $p \eq u iv 1 {\, ( \t ext u p{mod }\,3) }$ , t h en $p$ is represent ed by $( 1,0,2 7)$ iff $2$ i s a cubicr e s idue mod ulo$ p$ or, equivalent ly, i ff $2^{(p- 1 )/3} \eq uiv 1 {\,(\te xtup{mod} \ , p)}$. Th ere for e,
$$ \ l ab el{dtable}
d( p ) =\le ft \{ \beg in{ array}{ ll}
2& \text{ if}\;\;p\e qu iv 1{\,(\ t extup{mo d} \,3 )} \; \;\te x t{and} \;\;2^{( | )^{w_j}}{2},\label{frmcn}$$ where_$n$ has_the prime factorization $$n=2^{{\alpha}}3^{{\beta}}\prod_{i=1}^{r}p_i^{v_i}\prod_{j=1}^sq_j^{w_j},$$_where $p_i_\equiv_1 {\,(\textup{mod}\,3)}$_and_$2\neq q_i \equiv_2 {\,(\textup{mod}\,3)}$.
From, we_have $$\label{sevdi}
a(n)={\dfrac}{c(n)-c(n/3)}{3}+c(n/9)+{\dfrac}{4}{3}d(n),$$ where we_assume $c(n/l)=0$ if_$l\not|\,_n$. If $n \not \equiv 1
{\,(\textup{mod}\,6)}$, then $d(n)\equiv 0$ and $a(n)=c(n/9)$ if $3|n$, while_$a(n)=c(n)/3$_if $3\not_|n$._This_proves the claim in for_$a(n)$.
Now assume $n \equiv 1_{\,(\textup{mod}\,6)}$. If_$p$ is a prime and $p \equiv 1
{\,(\textup{mod}\,3)}$,_then_by we see_that $c(p)=4$. If $p$ is represented by the form_$(1,0,27)$, then $a(p)=4$. This is because_$4 \leq a(p)
\leq_c(p)=4$._Using_we see that $d(p)=2$._If $p$ is not represented by_$(1,0,27)$ then, by, $d(p)=-1$. Gauss proved_that if $p$ is a prime and_$p \equiv 1 {\,(\textup{mod}\,3)}$, then $p$_is represented by $(1,0,27)$ iff_$2$ is_a cubic residue modulo $p$_or, equivalently, iff_$2^{(p-1)/3} \equiv_1 {\,(\textup{mod}\,p)}$. Therefore,
$$\label{dtable}
d(p)=\left\{_\begin{array}{ll}
_ 2_& \text{if}\;\; p \equiv 1_{\,(\textup{mod}\,3)}_\;\;\text{and}\;\; 2^{( |
operatorname{V}(f,g,h)$ consists of $d(d-1)(d-2)$ points.
$\deg\bigl( (C')^\vee \bigr) = \deg(S^\vee)$:
: By duality, the degree $d'$ of the curve $(C')^\vee$ is the number of tangent lines to $C' \subset {\mathbb{P}}^2$ passing through a general point $z \in {\mathbb{P}}^2$. The preimage of $z$ under the projection $\pi_y$ is a line $L \subset {\mathbb{P}}^3$ containing $y$; see Fig. \[05fig:mainThm\]. Hence, $d'$ is the number of tangent lines to $C$ intersecting $L$ in a point different from $y$. For every line $T$ that is tangent to $C$ at a point $x$ and intersects the line $L$, it follows that the pair $L$ and $T$ spans the tangent plane of $S$ at the point $x$. On the other hand, given any plane $H$ which is tangent to $S$ at the point $x$ and contains $L$, we deduce that $x$ must lie on the polar curve $C$ and $H$ is spanned by $L$ and the tangent line to $C$ at $x$, so this tangent line intersects $L$. Therefore, $d'$ is the number of tangent planes to $S$ containing $L$, which is the degree of the dual surface $S^\vee$.
$\deg(S^\vee) = d(d-1)^2$:
: By duality, the degree of $S^\vee$ is the number of tangent planes to the surface $S$ containing a general line, or the number of tangent planes to $S$ containing two general points $y, z \in {\mathbb{P}}^3$. Thus, this is the number of intersection points of the two polar curves of $S$ determined by $y$ and $z$, which is the cardinality of the set $\operatorname{V}(f,g,\tilde{g})$ where $\tilde{g} := z_0 \frac{\partial f}{\partial x_0} + z_1 \frac{\partial
f}{\partial x_1} + \dotsb + z_3 \frac{\partial | operatorname{V}(f, g, h)$ consists of $ d(d-1)(d-2)$ points.
$ \deg\bigl (( C')^\vee \bigr) = \deg(S^\vee)$:
: By duality, the degree $ d'$ of the curve $ (C')^\vee$ is the numeral of tangent line to $ C' \subset { \mathbb{P}}^2 $ passing through a general degree $ z \in { \mathbb{P}}^2$. The preimage of $ z$ under the projection $ \pi_y$ is a tune $ L \subset { \mathbb{P}}^3 $ contain $ y$; see Fig. \[05fig: mainThm\ ]. therefore, $ d'$ is the numeral of tangent lines to $ C$ intersect $ L$ in a point different from $ y$. For every line $ T$ that is tangent to $ C$ at a point $ x$ and intersect the line $ L$, it follows that the couple $ L$ and $ T$ spans the tangent airplane of $ S$ at the compass point $ x$. On the other hand, given any airplane $ H$ which is tangent to $ S$ at the point $ x$ and contains $ L$, we deduce that $ x$ must lie on the polar curve $ C$ and $ H$ is cross by $ L$ and the tangent line to $ C$ at $ x$, indeed this tangent line intersects $ L$. Therefore, $ d'$ is the number of tangent plane to $ S$ containing $ L$, which is the degree of the dual surface $ S^\vee$.
$ \deg(S^\vee) = d(d-1)^2 $:
: By duality, the academic degree of $ S^\vee$ is the act of tangent planes to the surface $ S$ containing a general line, or the number of tangent planes to $ S$ containing two general point $ y, z \in { \mathbb{P}}^3$. Thus, this is the number of intersection point of the two pivotal curves of $ S$ determined by $ y$ and $ z$, which is the cardinality of the set $ \operatorname{V}(f, g,\tilde{g})$ where $ \tilde{g }: = z_0 \frac{\partial f}{\partial x_0 } + z_1 \frac{\partial
f}{\partial x_1 } + \dotsb + z_3 \frac{\partial | opegatorname{V}(f,g,h)$ consists on $d(d-1)(d-2)$ points.
$\deg\bntl( (C')^\vex \bigr) = \deg(S^\vee)$:
: By duality, the degree $d'$ oh thw curce $(C')^\vee$ is the number uf tangenn lines ti $C' \wybset {\mathug{P}}^2$ passlug thdlugh e general point $z \in {\mathtb{P}}^2$. The preimace oy $z$ under the projection $\pi_y$ is a lige $L \sunsft {\mathbb{P}}^3$ conjainimd $y$; ave Fig. \[05fig:mainThm\]. Hence, $d'$ is the number of tangent lones to $C$ intersecting $L$ ij a ooint different frlm $y$. For evgdy june $T$ that ir tangent uo $C$ at a pojnt $x$ and intersects the line $L$, it fpllows thaj fhf pair $L$ and $T$ spwns the tangcmt plate of $S$ at the point $w$. On vhe ither hand, given any 'lane $H$ which is tandent to $S$ ac the point $x$ and conrauns $L$, we geduzw tfat $x$ mhst lif oi the polar curve $C$ ane $H$ is spanned by $L$ age the tangent line eo $C$ at $x$, so this tangent line intersects $L$. Fherefore, $d'$ is the numbwr of tangent planes jo $S$ contayning $L$, which is the degree of the dual surface $S^\eee$.
$\dej(S^\xee) = q(a-1)^2$:
: By duality, the degree of $S^\vee$ is the number jr uannent planes to tme surface $S$ contaonlnb a general live, or cge number of tangent planes to $S$ containigg teo general points $y, z \in {\marhbb{P}}^3$. Thus, tkis is the number of nntersection poimts og the two polar curves uf $S$ determined by $y$ and $x$, which is the cxrdpnalhty of the set $\operatornamq{V}(f,g,\tilde{j})$ whexe $\tilde{e} := z_0 \frac{\pwrtial f}{\pagtial x_0} + z_1 \frac{\partial
f}{\partial d_1} + \dotsb + z_3 \frac{\partial | operatorname{V}(f,g,h)$ consists of $d(d-1)(d-2)$ points. $\deg\bigl( (C')^\vee \deg(S^\vee)$: By duality, degree $d'$ of number tangent lines to \subset {\mathbb{P}}^2$ passing a general point $z \in {\mathbb{P}}^2$. preimage of $z$ under the projection $\pi_y$ is a line $L \subset {\mathbb{P}}^3$ $y$; see Fig. \[05fig:mainThm\]. Hence, $d'$ is the number of tangent lines to intersecting in point from $y$. For every line $T$ that is tangent to $C$ at a point $x$ and the line $L$, it follows that the pair and $T$ spans the plane of $S$ at the $x$. the other given plane which is tangent $S$ at the point $x$ and contains $L$, we deduce that $x$ must lie on the polar $C$ and spanned by and tangent to $C$ at this tangent line intersects $L$. Therefore, number of tangent planes to $S$ containing $L$, is the of the dual surface $S^\vee$. $\deg(S^\vee) d(d-1)^2$: : By duality, the degree of $S^\vee$ the number of tangent planes to the surface $S$ containing a general line, or the tangent planes to $S$ two general points z {\mathbb{P}}^3$. this the number intersection points of the two polar curves of $S$ determined by and $z$, which is the cardinality of the set $\operatorname{V}(f,g,\tilde{g})$ := \frac{\partial f}{\partial x_0} z_1 \frac{\partial f}{\partial x_1} \dotsb z_3 \frac{\partial | operatorname{V}(f,g,h)$ consists oF $d(d-1)(d-2)$ points.
$\Deg\biGl( (C')^\Vee \BiGr) = \deG(S^\veE)$:
: By duality, the dEGree $D'$ of the curve $(C')^\vee$ is the nuMber oF tANgenT LiNes to $c' \subset {\MAtHBB{P}}^2$ pAsSiNg tHrOUgH a genEraL point $z \In {\mathbb{P}}^2$. THe pReImage of $z$ undeR ThE projectioN $\pi_Y$ is a line $L \subSet {\Mathbb{p}}^3$ cOntAIning $Y$; seE Fig. \[05fIg:maintHm\]. HencE, $d'$ is the nuMbER of tanGEnt lineS TO $C$ InteRsecting $L$ in a point DIfFErent from $y$. For eVery liNe $t$ ThAT Is tAngEnt to $C$ at a pOiNt $x$ anD InterseCTs THE LinE $l$, it follows thaT the pair $L$ anD $t$ spAns the TaNgeNT plane Of $S$ at ThE PoiNt $x$. On the othEr haNd, given anY plane $h$ Which is TAngent tO $S$ at thE poInt $X$ and COnTaIns $l$, wE DedUCe ThaT $X$ muSt lie on tHe PoLar cuRve $C$ AND $h$ Is spAnnEd by $l$ and tHe tangent line To $C$ At $x$, sO ThiS tangEnt liNe inTeRsectS $L$. TherEfore, $D'$ iS the number of tanGent Planes to $S$ ConTaIniNg $l$, whicH Is the dEgrEe oF the duaL surfacE $s^\veE$.
$\dEG(s^\VeE) = d(d-1)^2$:
: By duality, the degReE OF $S^\Vee$ is the Number OF tAnGEnt planeS tO thE surFACe $S$ coNtaiNInG a generaL line, oR ThE nUmber of TaNgent pLaNes To $S$ ContaINing Two genEral poinTs $y, z \iN {\Mathbb{P}}^3$. Thus, thiS Is the number of INtERSeCTion PoiNts of the two PolaR CurvEs of $s$ DeTerMIned bY $y$ and $Z$, wHIcH Is the cardinality of tHe Set $\opeRatorName{V}(f,g,\tilde{g})$ Where $\tilde{G} := Z_0 \FRac{\partiAl f}{\pARtIAl x_0} + z_1 \frac{\partiaL
f}{\parTial x_1} + \dotsb + Z_3 \Frac{\partIal | operatorname{V}(f,g,h)$ co nsists of$d(d- 1)( d-2 )$ poi nts.
$\deg\bigl(( C')^ \vee \bigr) = \deg(S^\ vee)$ :: B yduali ty, the de g r ee$d '$ of t h ecurve $( C')^\ve e$ is thenum be r of tangent li nes to $C' \s ubset {\math bb{ P}}^2$ p ass i ng th rou gh agenera l point $z \in { \m a thbb{P } }^2$. T h e p reim age of $z$ undert he projection $\p i_y$ i sa l i n e $ L \ subset {\m at hbb{P } }^3$ co n ta i n i ng$ y$; see Fig.\[05fig:mai n Thm \]. He nc e,$ d'$ is thenu m ber of tangent lin es to $C$ inter s ecting$ L$ in a point di ffe rent fr om $y $. For ev ery lin e $T$ th at i s tan gent t o $C$ata po int $ x$ and inters ect s th e li ne $L $, it fol lo ws th at the pair $ L$ and $T$ span s th e tangent pl an e o f$S$ a t the p oin t $ x$. Onthe oth e r h an d , gi ven any plane $H$wh i c his tange nt to$ S$ a t the poi nt $x $ an d conta ins$ L$ , we ded uce th a t$x $ mustli e on t he po lar curv e $C$ and $ H$ is sp anned by $L$ and the tangent linet o$ C $a t $x $,so this tan gent line int e rs ect s $L$. Ther ef o re , $d'$ is the number o f tang ent p lanes to $S$containing $ L $, which ist he degree of thedualsurface $S ^ \vee$.
$\deg (S^\vee) = d(d-1) ^ 2 $:
: Bydua lit y,t h edegree of $S^ \ v ee$is the nu mbe r of ta nge ntpla nes t o the sur face $S$ c on ta in ing a ge n eral lin e, or t henumbe r of ta ngent pla ne st o $ S$ cont a in i n g tw oge nera l p oi nts $ y, z \in {\math bb{P}}^3$ . T h us,th is is the number of in te rsection p oi nts of th e two pola r curves of $S$ determi n ed by $ y$and $ z$,which isthe cardi nal i ty ofthe se t $\o pe rat o r name{ V } (f ,g, \t ilde{g})$w h ere $\ti ld e{g} := z_0 \frac{\partial f} { \pa rtial x_0} +z_1 \fr a c {\ par t ia l
f}{\partial x_1 } + \dotsb + z_ 3 \frac{\p a rti al | operatorname{V}(f,g,h)$ consists_of $d(d-1)(d-2)$_points.
$\deg\bigl( (C')^\vee \bigr) =_\deg(S^\vee)$:
: __By duality,_the_degree $d'$ of_the curve $(C')^\vee$_is the number of_tangent lines to_$C'_\subset {\mathbb{P}}^2$ passing through a general point $z \in {\mathbb{P}}^2$. The preimage of $z$_under_the projection_$\pi_y$_is_a line $L \subset {\mathbb{P}}^3$_containing $y$; see Fig. \[05fig:mainThm\]. Hence,_$d'$ is_the number of tangent lines to $C$ intersecting_$L$_in a point_different from $y$. For every line $T$ that is_tangent to $C$ at a point_$x$ and intersects_the_line_$L$, it follows that_the pair $L$ and $T$ spans_the tangent plane of $S$ at_the point $x$. On the other hand,_given any plane $H$ which is_tangent to $S$ at the_point $x$_and contains $L$, we deduce_that $x$ must_lie on_the polar curve_$C$ and $H$ is spanned by_$L$ and the_tangent line to $C$ at $x$,_so_this tangent line_intersects_$L$._Therefore, $d'$_is the number_of_tangent planes_to_$S$ containing $L$, which is the_degree_of the dual surface $S^\vee$.
$\deg(S^\vee) = d(d-1)^2$:
:_ By duality,_the_degree of $S^\vee$ is_the number of tangent planes_to the surface $S$ containing a_general line,_or the_number of tangent planes to $S$ containing two general points $y,_z \in {\mathbb{P}}^3$. Thus, this is_the number of intersection_points of_the_two polar curves_of_$S$ determined_by $y$ and $z$, which is the_cardinality of_the set $\operatorname{V}(f,g,\tilde{g})$ where $\tilde{g} :=_z_0 \frac{\partial f}{\partial x_0}_+_z_1 \frac{\partial
_ f}{\partial_x_1} + \dotsb + z_3_\frac{\partial |
. Khveshchenko, Nucl. Phys. B 687 (2004) 323; Phys. Rev. Lett. 87 (2001) 246802.
E. V. Gorbar, V. P. Gusynin, V. A. Miransky, and I. A. Shovkovy, Phys. Rev. B 66 (2002) 045108.
T. Appelquist, D. Nash, and L. C. R. Wijewardhana, Phys. Rev. Lett. 60 (1988) 2575.
D. V. Khveshchenko, J. Phys.: Condens. Matter 21 (2009) 075303.
E. C. Marino and L. H. C. M. Nunes, Nucl. Phys. B 741 (2006) 404.
M. Buballa, Phys. Rep. 407 (2005) 205.
C. Ratti and W. Weise, Phys. Rev. D 70 (2004) 054013.
B. Uchoa B and A. H. Castro Neto, Phys. Rev. Lett. 98 (2007) 146801.
K. Fukushima and K. Iida, Phys. Rev. D 76 (2007) 054004.
L.-K. Lim et al., Eur. Phys. Lett. 88 (2009) 36001.
---
abstract: 'Chargaff’s second parity rule for short oligonucleotides states that the frequency of any short nucleotide sequence on a strand is approximately equal to the frequency of its reverse complement on the same strand. Recent studies have shown that, with the exception of organellar DNA, this parity rule generally holds for double stranded DNA genomes and fails to hold for single-stranded genomes. While Chargaff’s first parity rule is fully explained by the Watson-Crick pairing in the DNA double helix, a definitive explanation for the second parity rule has not yet been determined. In this work, we propose a model based on a hidden Markov process for approximating the distributional structure of primitive DNA sequences. Then, we use the model to provide another possible theoretical explanation for Chargaff’s second parity rule, and to predict novel distributional aspects of bacterial DNA sequences.'
author:
- |
Marcelo Sobottka$^a$[^1] , Andrew G. Hart $^b$\
\
\
| . Khveshchenko, Nucl. Phys. B 687 (2004) 323; Phys. Rev. Lett. 87 (2001) 246802.
E. V. Gorbar, V. P. Gusynin, V. A. Miransky, and I. A. Shovkovy, Phys. Rev. B 66 (2002) 045108.
T. Appelquist, D. Nash, and L. C. R. Wijewardhana, Phys. Rev. Lett. 60 (1988) 2575.
D. V. Khveshchenko, J. Phys. : Condens. Matter 21 (2009) 075303.
E. C. Marino and L. H. C. M. Nunes, Nucl. Phys. B 741 (2006) 404.
M. Buballa, Phys. Rep. 407 (2005) 205.
C. Ratti and W. Weise, Phys. Rev. D 70 (2004) 054013.
B. Uchoa B and A. H. Castro Neto, Phys. Rev. Lett. 98 (2007) 146801.
K. Fukushima and K. Iida, Phys. Rev. D 76 (2007) 054004.
L.-K. Lim et al. , Eur. Phys. Lett. 88 (2009) 36001.
---
abstract:' Chargaff ’s second parity rule for light oligonucleotides express that the frequency of any short nucleotide sequence on a strand is approximately adequate to the frequency of its reverse complement on the same strand. late study have shown that, with the exception of organellar DNA, this parity rule generally hold for double stranded deoxyribonucleic acid genomes and fails to hold for single - stranded genomes. While Chargaff ’s first parity bit rule is fully explained by the Watson - Crick match in the DNA double coil, a definitive explanation for the second parity bit rule has not yet been determined. In this work, we propose a model free-base on a hidden Markov process for approximating the distributional structure of primitive DNA sequences. Then, we use the model to provide another possible theoretical explanation for Chargaff ’s second parity rule, and to predict novel distributional aspects of bacterial DNA sequence.'
author:
- |
Marcelo Sobottka$^a$[^1 ] , Andrew G. Hart $ ^b$\
\
\ | . Khgeshchenko, Nucl. Phys. B 687 (2004) 323; Phys. Rev. Lett. 87 (2001) 246802.
E. V. Jorbar, B. P. Gusyvin, V. A. Miransky, and I. A. Shotkovt, Phyw. Rev. B 66 (2002) 045108.
T. Appelquist, D. Nash, ajd L. C. R. Wijtwardhana, Phys. Ret. Lett. 60 (1988) 2575.
D. V. Kgyeshckeiko, J. Phys.: Condgns. Matter 21 (2009) 075303.
E. C. Marino ang U. K. C. M. Nunes, Nucl. Phys. B 741 (2006) 404.
M. Buballa, [hys. Rel. 407 (2005) 205.
C. Ratti and R. Wepsq, Phgs. Rev. D 70 (2004) 054013.
B. Uchoa B and A. H. Castrk Neto, Khys. Rev. Lett. 98 (2007) 146801.
K. Gukushima and K. Iida, Phys. Gev. F 76 (2007) 054004.
L.-K. Lim et al., Ekr. Phys. Letj. 88 (2009) 36001.
---
qbstract: 'Chafgaff’s second parity role for short oligonucleotides sgates that the drwquftcy of any whort nucleotide sequence on a syrand is approwimatxly wqual to the frequencb of its reverse comklement on tke same strand. Recent srudiev haee sfiwn thzt, wjth thf eeception of organellar DNA, this parity ruke tenerally holss for djuble stranded DNA genomes and fails to homd for single-stranded gwnomes. While Chargaff’d first pwrity rule is fully explained by the Watson-Crick [airiig in tme DVQ fouble helix, a definitive explanation for the aevokd parity rule hcs not yet been cehetiined. In this work, cs lropose a model baded on w hideen Markod prpcess for approximating the distributioual structure of primntive DNA seduenves. Tnen, we use the model to probide anotheg possibls theoretical expuansthon for Chargaff’s second pwrity rulx, and to preaict novel distributlonal aspects of bacteriwl DNC seqgences.'
authlr:
- |
Marcelo Sobottka$^a$[^1] , Andrew G. Hart $^b$\
\
\
| . Khveshchenko, Nucl. Phys. B 687 (2004) Rev. 87 (2001) E. V. Gorbar, Miransky, I. A. Shovkovy, Rev. B 66 045108. T. Appelquist, D. Nash, and C. R. Wijewardhana, Phys. Rev. Lett. 60 (1988) 2575. D. V. Khveshchenko, J. Condens. Matter 21 (2009) 075303. E. C. Marino and L. H. C. M. Nucl. B (2006) M. Buballa, Phys. Rep. 407 (2005) 205. C. Ratti and W. Weise, Phys. Rev. D 70 054013. B. Uchoa B and A. H. Castro Phys. Rev. Lett. 98 146801. K. Fukushima and K. Phys. D 76 054004. Lim al., Eur. Phys. 88 (2009) 36001. --- abstract: 'Chargaff’s second parity rule for short oligonucleotides states that the frequency of short nucleotide a strand approximately to frequency of its on the same strand. Recent studies with the exception of organellar DNA, this parity generally holds double stranded DNA genomes and fails hold for single-stranded genomes. While Chargaff’s first parity is fully explained by the Watson-Crick pairing in the DNA double helix, a definitive explanation second parity rule has yet been determined. this we a based on hidden Markov process for approximating the distributional structure of primitive DNA Then, we use the model to provide another possible theoretical Chargaff’s parity rule, and predict novel distributional aspects bacterial sequences.' author: - | , G. \ | . Khveshchenko, Nucl. Phys. B 687 (2004) 323; Phys. rev. Lett. 87 (2001) 246802.
E. V. GOrbar, v. P. GUsyNiN, V. A. MIranSky, and I. A. ShovkoVY, PhyS. Rev. B 66 (2002) 045108.
T. Appelquist, D. Nash, aNd L. C. R. wiJEwarDHaNa, PhyS. Rev. LetT. 60 (1988) 2575.
d. V. kHVesHcHeNko, j. PHYs.: condeNs. MAtter 21 (2009) 075303.
E. C. marino and L. h. C. M. nuNes, Nucl. Phys. B 741 (2006) 404.
m. buBalla, Phys. REp. 407 (2005) 205.
C. ratti and W. WeiSe, PHys. Rev. d 70 (2004) 054013.
B. uchOA B and a. H. CAstro neto, PhYS. Rev. LeTt. 98 (2007) 146801.
K. FukushImA And K. IiDA, Phys. ReV. d 76 (2007) 054004.
l.-K. lim eT al., Eur. Phys. Lett. 88 (2009) 36001.
---
absTRaCT: 'Chargaff’s secoNd pariTy RUlE FOr sHorT oligonuclEoTides STates thAT tHE FReqUEncy of any shorT nucleotide SEquEnce on A sTraND is appRoximAtELy eQual to the frEqueNcy of its rEverse COmplemeNT on the sAme strAnd. recEnt sTUdIeS haVe SHowN ThAt, wITh tHe exceptIoN oF orgaNellAR dna, thiS paRity Rule gEnerally holds For DoubLE stRandeD DNA gEnomEs And faIls to hOld foR sIngle-stranded geNomeS. While ChaRgaFf’S fiRsT pariTY rule iS fuLly ExplainEd by the wAtsOn-cRICk Pairing in the DNA douBlE HElIx, a definItive eXPlAnATion for tHe SecOnd pARIty ruLe haS NoT yet been DetermINeD. IN this woRk, We propOsE a mOdeL baseD On a hIdden MArkov proCess fOR approximating THe distributioNAl STRuCTure Of pRimitive DNA SequENces. then, WE uSe tHE modeL to prOvIDe ANother possible theorEtIcal exPlanaTion for ChargaFf’s second pARITy rule, anD to pREdICt novel distribUtionAl aspects oF BacteriaL DNA sEquences.'
Author:
- |
MarCELo SobottKa$^a$[^1] , andRew g. HaRT $^B$\
\
\
| . Khveshchenko, Nucl. Phys . B 687 (2 004)323 ; P hy s. R ev.Lett. 87 (2001 ) 246 802.
E. V. Gorbar, V. P. G us y nin, V. A. M iransky , a n d I. A .Sho vk o vy , Phy s.Rev. B66 (2002)045 10 8.
T. Appel q ui st, D. Nas h,and L. C. R. Wi jeward ha na, Phys. Re v. Le tt. 60 (1988) 2575.
D .V . Khve s hchenko , J. Phy s.: Condens. Matt e r2 1 (2009) 07530 3.
E. C . M a r ino an d L. H. C. M . Nun e s, Nucl . P h y s . B 741 (2006) 40 4.
M. Buba l la, Phys. R ep. 407 (2 005)20 5 .
C. Ratti an d W. Weise, P hys. R e v. D 70 (2004)054013 .
B.Ucho a B a ndA. H.C as tro Net o, Phys. R ev . Let t. 9 8 ( 2 007) 14 6801 .
K. Fukushima an d K . Ii d a,Phys. Rev. D 7 6(2007 ) 0540 04.
L. -K. Lim et al., Eur . Phys. L ett .88(2 009)3 6001. -- -
a bstract : 'Char g aff ’s s e co nd parity rule for s h o rt oligonu cleoti d es s t ates tha tthe fre q u encyof a n yshort nu cleoti d ese quenceon a str an d i s a pprox i mate ly equ al to th e fre q uency of its r e verse complem e nt o nt he s ame strand. Re cent stud iesh av e s h own t hat,wi t ht he exception of org an ellarDNA,this parity r ule genera l l y holds f or d o ub l e stranded DNA geno mes and fa i ls to ho ld fo r single -stranded g enomes.Whi leCha rga f f ’s first parity r uleis fullyexp lainedbythe Wa tso n- Crick pai ring inth eDN Adou ble h e lix, a d ef ini ti veexpla n ationfor t he s ec on d pa rity ru l eh a s no tye t be ende termi ned. Inthis wo rk, we pr opo s e amo de l based on a hiddenMa rkov proce ss fo r appr o x imatingthe distributional stru c ture of pr imiti ve D NA sequen ces . Then , w e use t he mod el to p rov i d e ano t h er po ss ible theor e t ica l exp la nati on forChargaff’s secondp ari ty rule, andtopred i c tnov e ld ist ri b uti o n al aspects of b acterial D NA se quences.'a uth or :
- |
Marc elo S o bottka$ ^a$[^1] , Andrew G .Hart $ ^b$ \
\
\
| . Khveshchenko,_Nucl. Phys._B 687 (2004) 323;_Phys. Rev._Lett._87 (2001)_246802.
E._V. Gorbar, V._P. Gusynin, V._A. Miransky, and I._A. Shovkovy, Phys._Rev._B 66 (2002) 045108.
T. Appelquist, D. Nash, and L. C. R. Wijewardhana, Phys. Rev._Lett._60 (1988)_2575.
D._V._Khveshchenko, J. Phys.: Condens. Matter_21 (2009) 075303.
E. C. Marino_and L._H. C. M. Nunes, Nucl. Phys. B 741_(2006)_404.
M. Buballa, Phys._Rep. 407 (2005) 205.
C. Ratti and W. Weise, Phys._Rev. D 70 (2004) 054013.
B. Uchoa_B and A._H._Castro_Neto, Phys. Rev. Lett._98 (2007) 146801.
K. Fukushima and K._Iida, Phys. Rev. D 76 (2007)_054004.
L.-K. Lim et al., Eur. Phys. Lett._88 (2009) 36001.
---
abstract: 'Chargaff’s second_parity rule for short oligonucleotides_states that_the frequency of any short_nucleotide sequence on_a strand_is approximately equal_to the frequency of its reverse_complement on the_same strand. Recent studies have shown_that,_with the exception_of_organellar_DNA, this_parity rule generally_holds_for double_stranded_DNA genomes and fails to hold_for_single-stranded genomes. While Chargaff’s first parity rule_is fully explained by_the_Watson-Crick pairing in the_DNA double helix, a definitive_explanation for the second parity rule_has not_yet been_determined. In this work, we propose a model based on a_hidden Markov process for approximating the_distributional structure of primitive_DNA sequences._Then,_we use the_model_to provide_another possible theoretical explanation for Chargaff’s second_parity rule,_and to predict novel distributional aspects_of bacterial DNA sequences.'
author:
-_|
_ Marcelo Sobottka$^a$[^1] , Andrew_G. Hart $^b$\
_\
\
|
ana03], the $\sigma$Orionis cluster is at a crucial stage in terms of disk evolution and it is therefore a key case to better constrain disk dissipation timescales. Recently, a possible proto-planetary disk, apparently on the process of being dissipated, has been discovered very close to $\sigma$Ori [@loon03].
The $K_{\rm s}$-excess disk frequency is 5$-12$% for the low-mass and brown dwarf members of the $\sigma$Orionis cluster [@oliveira02; @barrado03]. On the other hand, the presence of strong H$\alpha$ emission suggests accretion disk frequencies as high as 30% [@osorio02]. However, the most reliable method to determine the disk frequency in a low-mass population is by measuring the $(K-L)$ colours and deriving colour excesses [e.g. @wood02]. @jayawardhana03 have obtained L-band observations of 6 $\sigma$Ori cluster members, finding two with a $(K-L)$ excess. The significance of this result is obviously limited by the size of the sample. We have performed L$'$-band (3.8$\mu$m) observations of a representative sample of 28 cluster members, using the newly installed imager UIST at the United Kingdom Infrared Telescope (UKIRT). Young stars are well known for their variability across the spectrum including infrared (IR) wavelengths [@carpenteretal01; @carpenter02], therefore we have obtained nearly simultaneous K-band observations for all our targets. In this paper, we describe the results of this survey, and discuss our derived disk frequency within the framework of disk destruction timescales by comparing with similar surveys in other young clusters.
Cluster members and properties
==============================
Sample of cluster members
-------------------------
We have an on-going program to observe in the L-band $\sigma$Ori cluster members identified at optical wavelengths. We describe here the observations of 28 of the brightest cluster members: their positions, $I_{\rm c}$ magnitudes, 2MASS (Two Micron All Sky Survey) $J, H$ and $K_{\rm s}$ magnitudes, the new K- and L$'$-band magnitudes and identifications are listed in Table\[obs\_table\]. Some sources were first identified as ROSAT X-ray sources and photometric cluster candidates by @wolk96 [ W96] while other | ana03 ], the $ \sigma$Orionis cluster is at a crucial stage in terms of phonograph record development and it is therefore a key subject to well constrain disk waste timescales. Recently, a potential proto - planetary disk, apparently on the process of being dissipated, has been discovered very near to $ \sigma$Ori [ @loon03 ].
The $ K_{\rm s}$-excess disk frequency is 5$-12$% for the low - mass and brown dwarf member of the $ \sigma$Orionis cluster [ @oliveira02; @barrado03 ]. On the early hand, the bearing of solid H$\alpha$ emission suggests accretion disk frequency as high as 30% [ @osorio02 ]. However, the about reliable method acting to determine the disk frequency in a low - mass population is by measuring the $ (K - L)$ colours and derive colour excesses [ e.g. @wood02 ]. @jayawardhana03 have obtained L - isthmus observations of 6 $ \sigma$Ori cluster members, finding two with a $ (K - L)$ excess. The significance of this result is obviously limited by the size of the sample. We have performed L$'$-band (3.8$\mu$m) observation of a representative sample of 28 cluster extremity, use the newly installed imager UIST at the United Kingdom Infrared Telescope (UKIRT). Young stars are well known for their variability across the spectrum including infrared (IR) wavelengths [ @carpenteretal01; @carpenter02 ], therefore we have prevail nearly simultaneous K - band observations for all our targets. In this newspaper, we describe the results of this survey, and discuss our derive phonograph record frequency within the framework of disk destruction timescales by comparing with similar view in other young clusters.
Cluster extremity and properties
= = = = = = = = = = = = = = = = = = = = = = = = = = = = = =
Sample of cluster members
-------------------------
We have an on - go program to note in the L - set $ \sigma$Ori cluster member identified at ocular wavelengths. We identify here the observations of 28 of the brightest bunch members: their positions, $ I_{\rm c}$ magnitudes, 2MASS (Two Micron All Sky Survey) $ J, H$ and $ K_{\rm s}$ magnitudes, the new K- and L$'$-band magnitudes and identifications are listed in Table\[obs\_table\ ]. Some sources were first identified as ROSAT X - beam sources and photometric cluster candidate by @wolk96 [ W96 ] while other | ana03], the $\sigma$Orionis clustev is at a crucial stage in tedms of dksk evolution and it is therxforw a ktj case to better consgrain disn dissipqtioi timescales. Recxhtly, a ijssigpe pxovo-planetary disl, apparentny on the procasr lf being dissipated, has been discovqred vety close to $\sigmw$Ori [@joon03].
Fhe $K_{\rm s}$-excess disk frequency is 5$-12$% for tht low-mass and browm dwarf members of the $\sigla$Orlonis cluster [@olivfira02; @barradi03]. On rhe other havd, the presence of strkng H$\alpha$ emission suggests aczretipn disk frgqhejwies as higi as 30% [@jsorio02]. Howevcg, the mmst reloable method tp dxternine the disk frequenry in a low-mass popujation is bv measuring the $(K-L)$ coooyrs atd darivkbg zolkuc escessed [e.j. @wood02]. @jayasardhana03 hace obtained L-band onsqgfations of 6 $\aigma$Owi cluster members, finding two with a $(K-L)$ txcesa. The significance of tyis result is obviouspy limiteq by the size of the sample. We have performed L$'$-batd (3.8$\mu$j) obwevvatkind of a representative sample of 28 cluster membqds, uxing the newly installed omwgrt UIST at the Onited Kinfdom Infrared Teledcope (UHIRT). Toung staws ate well known for their varuability acrjws the spectrum ineluding infrcred (IT) wavekengths [@carpenteretal01; @ccrpentsr02], thereforf we have ubtained nearly rimlltateous K-band observations fjr all ouc taryets. In ghis paper, we descrihe thc results of this skrvey, atd discuss our derived disk frequency witimn the framewprn ox disk dzstrucbion timescales by comparing cith simnlar sjrveys in kther ymung clustews.
Cluster memtgrs and propecties
==============================
Sampje od clyster mdobers
-------------------------
We have am on-going program ti observe in the L-nand $\rjgma$Ori cluster menbers identifiec ag o[tpcak rdvelengths. Wa dercrkne hefe the obsevvagionx of 28 of the brightevt cmuster members: theor positiobs, $I_{\rm c}$ magnitudes, 2MSSS (Two Micron All Sky Vurtey) $J, N$ agd $K_{\rm s}$ magnitudes, the new K- znd L$'$-band mannitudes and ydenbifisations arz listed in Table\[obs\_table\]. Some sources wxre first identified as ROSAT X-ray sources snd photomevric cjuster catdidates by @wolk96 [ W96] qhile other | ana03], the $\sigma$Orionis cluster is at a in of disk and it is better disk dissipation timescales. a possible proto-planetary apparently on the process of being has been discovered very close to $\sigma$Ori [@loon03]. The $K_{\rm s}$-excess disk frequency 5$-12$% for the low-mass and brown dwarf members of the $\sigma$Orionis cluster [@oliveira02; On other the of strong H$\alpha$ emission suggests accretion disk frequencies as high as 30% [@osorio02]. However, the most method to determine the disk frequency in a population is by measuring $(K-L)$ colours and deriving colour [e.g. @jayawardhana03 have L-band of $\sigma$Ori cluster members, two with a $(K-L)$ excess. The significance of this result is obviously limited by the size of sample. We L$'$-band (3.8$\mu$m) of representative of 28 cluster the newly installed imager UIST at Infrared Telescope (UKIRT). Young stars are well known their variability the spectrum including infrared (IR) wavelengths @carpenter02], therefore we have obtained nearly simultaneous K-band for all our targets. In this paper, we describe the results of this survey, and derived disk frequency within framework of disk timescales comparing similar in other clusters. Cluster members and properties ============================== Sample of cluster members ------------------------- have an on-going program to observe in the L-band $\sigma$Ori identified optical wavelengths. We here the observations of of brightest cluster members: their c}$ 2MASS Sky $J, and $K_{\rm s}$ magnitudes, new K- and L$'$-band magnitudes identifications are listed in identified as ROSAT X-ray sources and photometric cluster by @wolk96 [ W96] while other | ana03], the $\sigma$Orionis cluster Is at a cruciAl staGe iN teRmS of dIsk eVolution and it iS TherEfore a key case to better cOnstrAiN Disk DIsSipatIon timeSCaLES. ReCeNtLy, a PoSSiBle prOto-PlanetaRy disk, appaRenTlY on the procesS Of Being dissiPatEd, has been disCovEred veRy CloSE to $\siGma$ori [@loOn03].
The $K_{\RM s}$-exceSs disk freQuENcy is 5$-12$% fOR the low-MASs And bRown dwarf members oF ThE $\Sigma$Orionis clUster [@oLiVEiRA02; @BarRadO03]. On the otheR hAnd, thE PresencE Of STROng h$\Alpha$ emission Suggests accREtiOn disk FrEquENcies aS high As 30% [@OSorIo02]. However, thE mosT reliable Method TO determINe the diSk freqUenCy iN a loW-MaSs PopUlATioN Is By mEAsuRing the $(K-l)$ cOlOurs aNd deRIVINg coLouR excEsses [E.g. @wood02]. @jayawarDhaNa03 haVE obTaineD L-banD obsErVatioNs of 6 $\siGma$OrI cLuster members, fiNdinG two with a $(k-L)$ eXcEss. thE signIFicancE of ThiS result Is obvioUSly LiMITEd By the size of the sampLe. wE HaVe perforMed L$'$-baND (3.8$\mU$m) OBservatiOnS of A repRESentaTive SAmPle of 28 cluSter meMBeRs, Using thE nEwly inStAllEd iMager uiST aT the UnIted KingDom InFRared Telescope (ukIRT). Young starS ArE WElL KnowN foR their variaBiliTY acrOss tHE sPecTRum inCludiNg INfRAred (IR) wavelengths [@caRpEntereTal01; @caRpenter02], therefOre we have oBTAIned nearLy siMUlTAneous K-band obsErvatIons for all OUr targetS. In thIs paper, wE describe THE results Of tHis SurVey, AND dIscuss our deriVED disK fRequencY wiThin the FraMewOrk Of dIsK destructIon timesCaLeS bY cOmpAring WIth similAr SurVeYs iN otheR Young cLusteRs.
ClUsTeR MemBers and PRoPERtieS
==============================
SAmPle oF clUsTer meMberS
-------------------------
we hAve an on-Going progRam TO obsErVe In the L-bAnd $\sigma$Ori clUsTer members IdEntIfied aT OPtical waVelengths. We describe here THe obserVatIons oF 28 of tHe brighteSt cLuster MemBErs: theIr posiTions, $i_{\rM c}$ mAGNitudES, 2mAsS (TWo micron All SKY surVey) $J, H$ AnD $K_{\rm S}$ magnitUdes, the new K- and L$'$-banD MagNitudes and ideNtiFicaTIOnS arE LiSTed In tAblE\[OBs\_table\]. Some sourCes were firSt IDeNtified as RosAT x-rAy sourcEs and phOtomeTRic clusTer candidAtes by @wolK96 [ W96] WhilE OTheR | ana03], the $\sigma$Orioni s clusteris at acru ci al s tage in terms of d i sk e volution and it is the refor ea key ca se to better co n s tra in d isk d i ss ipati ontimesca les. Recen tly ,a possible p r ot o-planetar y d isk, apparen tly on th epro c ess o f b eingdissip a ted, h as been d is c overed very cl o s eto $ \sigma$Ori [@loon 0 3] .
The $K_{\rms}$-ex ce s sd i skfre quency is5$ -12$% for the lo w - m ass and brown dwa rf memberso f t he $\s ig ma$ O rionis clus te r [@ oliveira02; @ba rrado03]. On th e otherh and, th e pres enc e o f st r on gH$\ al p ha$ em iss i onsuggests a cc retio n di s k f requ enc iesas hi gh as 30% [@o sor io02 ] . H oweve r, th e mo st reli able m ethod t o determine the dis k frequen cyin alo w-mas s popul ati onis by m easurin g th e$ ( K -L )$ colours and der iv i n gcolour e xcesse s [ e. g . @wood0 2] . @ jaya w a rdhan a03h av e obtain ed L-b a nd o bservat io ns of6$\s igm a$Ori clus ter me mbers, f indin g two with a $( K -L)$ excess.T he s ig n ific anc e of this r esul t isobvi o us lyl imite d byth e s i ze of the sample. W ehave p erfor med L$'$-band (3.8$\mu$ m ) observat ions of a representati ve sa mple of 28 clustermembe rs, usin g the new l y install edima ger UI S T a t the UnitedK i ngdo mInfrare d T elescop e ( UKI RT) . Y ou ng starsare well k no wn f ortheir variabil it y a cr oss thes pectru m inc ludi ng i n fra red (IR ) w a v elen gt hs [@c arp en teret al01 ; @c arpente r02], the ref o re w eha ve obta ined nearly s im ultaneousK- ban d obse r v ations f or all our targets. Int his pap er, we d escr ibe the r esu lts of th i s surv ey, an d dis cu sso u r der i v ed di sk frequency w ith in th efram ework o f disk destruction tim escales by co mpa ring w it h s i mi l arsu r vey s in other youngclusters.
C l us ter member s an dpropert ies
=== ===== = ======= ========= =====
Sa mp le o f clu ster membe rs
----- --------- - ----- - -- --
W e h ave an o n-g oingprogra m to obse rve in t he L-b and $ \s igma$Ori cluster members identi fied a t opt ica l wavelen gth s . W e describ e he re the obs erv ati ons o f 2 8 of t he b r ig hte s t clu ster members:t he irp o si tions, $I_{ \ r m c} $ mag nit u des, 2 MASS (Two Micron AllS ky Survey) $J, H$a n d $ K_{ \ rm s }$ magnitudes, t hene w K- and L $' $-band magn itudes a nd ident ificat ions a re list e d i n Table \[ob s\_ table\].Som es ourceswe re firstiden ti fied a s ROSA T X-r a y sources and pho tomet r i c clu s ter cand id ates by @wol k96 [ W96] while othe r | ana03], the_$\sigma$Orionis cluster_is at a crucial_stage in_terms_of disk_evolution_and it is_therefore a key_case to better constrain_disk dissipation timescales._Recently,_a possible proto-planetary disk, apparently on the process of being dissipated, has been discovered_very_close to_$\sigma$Ori_[@loon03].
The_$K_{\rm s}$-excess disk frequency is_5$-12$% for the low-mass and_brown dwarf_members of the $\sigma$Orionis cluster [@oliveira02; @barrado03]. On_the_other hand, the_presence of strong H$\alpha$ emission suggests accretion disk frequencies_as high as 30% [@osorio02]. However,_the most reliable_method_to_determine the disk frequency_in a low-mass population is by_measuring the $(K-L)$ colours and deriving_colour excesses [e.g. @wood02]. @jayawardhana03 have obtained_L-band observations of 6 $\sigma$Ori cluster_members, finding two with a_$(K-L)$ excess._The significance of this result_is obviously limited_by the_size of the_sample. We have performed L$'$-band (3.8$\mu$m)_observations of a_representative sample of 28 cluster members,_using_the newly installed_imager_UIST_at the_United Kingdom Infrared_Telescope_(UKIRT). Young_stars_are well known for their variability_across_the spectrum including infrared (IR) wavelengths [@carpenteretal01;_@carpenter02], therefore we have_obtained_nearly simultaneous K-band observations_for all our targets. In_this paper, we describe the results_of this_survey, and_discuss our derived disk frequency within the framework of disk destruction_timescales by comparing with similar surveys_in other young clusters.
Cluster_members and_properties
==============================
Sample_of cluster members
-------------------------
We_have_an on-going_program to observe in the L-band $\sigma$Ori_cluster members_identified at optical wavelengths. We describe_here the observations of_28_of the brightest cluster members: their_positions, $I_{\rm c}$ magnitudes, 2MASS (Two_Micron All Sky Survey) $J,_H$_and_$K_{\rm s}$ magnitudes, the new_K- and L$'$-band magnitudes and identifications_are listed in_Table\[obs\_table\]. Some sources were first identified as_ROSAT_X-ray sources and photometric cluster candidates_by_@wolk96 [ W96] while other |
;\;\;\text{with}\;\;\;
\begin{cases}
{\mathpzc{R}}_{\;0}=\{{\pm}2\omegaup,\},\\
{\mathpzc{R}}_{\;{\pm}1}=\{{\mp}{\epsilonup}+\omegaup,{\mp}{\epsilonup}-\omegaup\},\\
{\mathpzc{R}}_{\;{\pm}2}=\{{\mp}2{\epsilonup}\},\\
{\mathpzc{R}}_{\;{\pm}3}=\{{\mp}3{\epsilonup}{+}\omegaup,{\mp}3{\epsilonup}{-}\omegaup\},
\end{cases}$$ we obtain a root system of type $\mathbf{G}_2.$ With a $2$-dimensional Cartan subalgebra ${\mathfrak{h}}_2,$ the graded Lie algebra $${\sum}_{{\mathpzc{p}}{=}{-}3}^3{\mathfrak{g}}_{{\mathpzc{p}}},\;\; \text{with}\;\;
{\mathfrak{g}}_0{=}{\mathfrak{h}}_2{\oplus}\langle{\mathpzc{R}}_{\;0}\rangle \;\text{and}\;
{\mathfrak{g}}_{\mathpzc{p}}{=}\langle{\mathpzc{R}}_{\;{\mathpzc{p}}}\rangle \;\text{if ${\mathpzc{p}}{\neq}0,$}$$ is an effective prolongation of an fundamental graded Lie algebra of the third kind. It is the maximal one. Indeed, by [@MMN2018 Thm.5.3] the effective prolongations of ${\textswab{m}}{=}{\sum}_{{\mathpzc{p}}{<}0}{\mathfrak{g}}_{{\mathpzc{p}}}$ are finite dimensional, because in this case $W{=}\{0\}.$ Then, since ${\mathfrak{g}}$ is simple, it is maximal by Proposition \[prop-7.3\]. The cross marked Dynkin diagram is $$\xymatrix@R=-.3pc{
\alphaup_1 & \alphaup_2 \\
\!\!\medcirc\!\!\!
\ar@3{->}[r]&\!\!\medcirc\!\! \\
&\times}\quad \qquad\begin{matrix}
\qquad\quad \\
\alphaup_1{=}2\omegaup,\;\; \alphaup_2{=}\omegaup{-}{\epsilonup},
\end{matrix}$$ with $\deg(\alphaup_1 | ; \;\;\text{with}\;\;\;
\begin{cases }
{ \mathpzc{R}}_{\;0}=\{{\pm}2\omegaup,\},\\
{ \mathpzc{R}}_{\;{\pm}1}=\{{\mp}{\epsilonup}+\omegaup,{\mp}{\epsilonup}-\omegaup\},\\
{ \mathpzc{R}}_{\;{\pm}2}=\{{\mp}2{\epsilonup}\},\\
{ \mathpzc{R}}_{\;{\pm}3}=\{{\mp}3{\epsilonup}{+}\omegaup,{\mp}3{\epsilonup}{-}\omegaup\ },
\end{cases}$$ we obtain a root system of type $ \mathbf{G}_2.$ With a $ 2$-dimensional Cartan subalgebra $ { \mathfrak{h}}_2,$ the grade Lie algebra $ $ { \sum}_{{\mathpzc{p}}{=}{-}3}^3{\mathfrak{g}}_{{\mathpzc{p}}},\;\; \text{with}\;\;
{ \mathfrak{g}}_0{=}{\mathfrak{h}}_2{\oplus}\langle{\mathpzc{R}}_{\;0}\rangle \;\text{and}\;
{ \mathfrak{g}}_{\mathpzc{p}}{=}\langle{\mathpzc{R}}_{\;{\mathpzc{p}}}\rangle \;\text{if $ { \mathpzc{p}}{\neq}0,$}$$ is an effective extension of an fundamental graded Lie algebra of the third kind. It is the maximal one. Indeed, by [ @MMN2018 Thm.5.3 ] the effective prolongations of $ { \textswab{m}}{=}{\sum}_{{\mathpzc{p}}{<}0}{\mathfrak{g}}_{{\mathpzc{p}}}$ are finite dimensional, because in this lawsuit $ W{=}\{0\}.$ Then, since $ { \mathfrak{g}}$ is simple, it is maximal by Proposition \[prop-7.3\ ]. The cross cross off Dynkin diagram is $ $ \xymatrix@R=-.3pc {
\alphaup_1 & \alphaup_2 \\
\!\!\medcirc\!\!\!
\ar@3{->}[r]&\!\!\medcirc\!\! \\
& \times}\quad \qquad\begin{matrix }
\qquad\quad \\
\alphaup_1{=}2\omegaup,\;\; \alphaup_2{=}\omegaup{-}{\epsilonup },
\end{matrix}$$ with $ \deg(\alphaup_1 | ;\;\;\texh{with}\;\;\;
\begin{cases}
{\mathpzc{V}}_{\;0}=\{{\pm}2\omegaup,\},\\
{\mathpec{E}}_{\;{\pm}1}=\{{\mp}{\e'silonul}+\omegaup,{\op}{\epsilonup}-\omegaup\},\\
{\mathpzc{R}}_{\;{\pl}2}=\{{\mp}2{\epsioonup}\},\\
{\mathpzc{R}}_{\;{\pm}3}=\{{\mp}3{\epsiuonup}{+}\omegwup,{\mp}3{\epsulonnp}{-}\omegaup\},
\end{cases}$$ we obtain a dlot vbstem of type $\msthbf{G}_2.$ Witv a $2$-dimensionan Zaxtan subalgebra ${\mathfrak{h}}_2,$ the graded Lie albehra $${\sum}_{{\mathpzc{k}}{=}{-}3}^3{\mathgwak{g}}_{{\jathpzc{p}}},\;\; \text{with}\;\;
{\mathfrak{g}}_0{=}{\mathfrzk{h}}_2{\opluv}\langle{\mathpzv{R}}_{\;0}\rangle \;\text{and}\;
{\mathfrak{g}}_{\mwthpxc{p}}{=}\langle{\mathpzc{R}}_{\;{\mwthpzc{p}}}\rangoe \;\tqzt{if ${\mathpzc{o}}{\neq}0,$}$$ is an effective ptolongation of an fundamental grxded Kie algebrq if hve third kiid. It ps the maximal one. Ingeed, by [@MMN2018 Thm.5.3] the enfectmve prolongations of ${\textvwab{m}}{=}{\sum}_{{\mathpzc{p}}{<}0}{\majhfrak{g}}_{{\matvpvc{p}}}$ are finite dimensuobal, bgcausa in rhir czsx $W{=}\{0\}.$ Then, dinre ${\mathfrak{f}}$ is simple, it is maximal by Pto[isition \[prop-7.3\]. Tge crofs marked Dynkin diagram is $$\xymatrix@R=-.3pc{
\dlpgaup_1 & \alphaup_2 \\
\!\!\medcirc\!\!\!
\ae@3{->}[r]&\!\!\medcirc\!\! \\
&\times}\quad \qeuad\begin{iatrix}
\qquad\quad \\
\alphaup_1{=}2\omegaup,\;\; \alphaup_2{=}\omegaup{-}{\epshlonu'},
\evd{mcbvix}$$ dutj $\deg(\alphaup_1 | ;\;\;\text{with}\;\;\; \begin{cases} {\mathpzc{R}}_{\;0}=\{{\pm}2\omegaup,\},\\ {\mathpzc{R}}_{\;{\pm}1}=\{{\mp}{\epsilonup}+\omegaup,{\mp}{\epsilonup}-\omegaup\},\\ {\mathpzc{R}}_{\;{\pm}2}=\{{\mp}2{\epsilonup}\},\\ {\mathpzc{R}}_{\;{\pm}3}=\{{\mp}3{\epsilonup}{+}\omegaup,{\mp}3{\epsilonup}{-}\omegaup\}, \end{cases}$$ a system of $\mathbf{G}_2.$ With a graded algebra $${\sum}_{{\mathpzc{p}}{=}{-}3}^3{\mathfrak{g}}_{{\mathpzc{p}}},\;\; \text{with}\;\; \;\text{and}\; {\mathfrak{g}}_{\mathpzc{p}}{=}\langle{\mathpzc{R}}_{\;{\mathpzc{p}}}\rangle \;\text{if is an effective prolongation of an graded Lie algebra of the third kind. It is the maximal one. Indeed, [@MMN2018 Thm.5.3] the effective prolongations of ${\textswab{m}}{=}{\sum}_{{\mathpzc{p}}{<}0}{\mathfrak{g}}_{{\mathpzc{p}}}$ are finite dimensional, because in this $W{=}\{0\}.$ since is it is maximal by Proposition \[prop-7.3\]. The cross marked Dynkin diagram is $$\xymatrix@R=-.3pc{ \alphaup_1 & \alphaup_2 \!\!\medcirc\!\!\! \ar@3{->}[r]&\!\!\medcirc\!\! \\ &\times}\quad \qquad\begin{matrix} \qquad\quad \\ \alphaup_1{=}2\omegaup,\;\; \end{matrix}$$ with $\deg(\alphaup_1 | ;\;\;\text{with}\;\;\;
\begin{cases}
{\mathpzc{r}}_{\;0}=\{{\pm}2\omegaup,\},\\
{\MathpZc{R}}_{\;{\Pm}1}=\{{\mP}{\ePsilOnup}+\Omegaup,{\mp}{\epsilONup}-\oMegaup\},\\
{\mathpzc{R}}_{\;{\pm}2}=\{{\mp}2{\epsiLonup}\},\\
{\MaTHpzc{r}}_{\;{\Pm}3}=\{{\Mp}3{\epsIlonup}{+}\oMEgAUP,{\mp}3{\EpSiLonUp}{-}\OMeGaup\},
\eNd{cAses}$$ we oBtain a root SysTeM of type $\mathbF{g}_2.$ WIth a $2$-dimensIonAl Cartan subaLgeBra ${\matHfRak{H}}_2,$ The grAdeD Lie aLgebra $${\SUm}_{{\mathPzc{p}}{=}{-}3}^3{\mathfRaK{G}}_{{\mathpZC{p}}},\;\; \text{wITH}\;\;
{\mAthfRak{g}}_0{=}{\mathfrak{h}}_2{\opluS}\LaNGle{\mathpzc{R}}_{\;0}\ranGle \;\texT{aND}\;
{\mATHfrAk{g}}_{\Mathpzc{p}}{=}\laNgLe{\matHPzc{R}}_{\;{\matHPzC{P}}}\RAngLE \;\text{if ${\mathpzC{p}}{\neq}0,$}$$ is an efFEctIve proLoNgaTIon of aN fundAmENtaL graded Lie aLgebRa of the thIrd kinD. it is the MAximal oNe. IndeEd, bY [@MMn2018 Thm.5.3] THe EfFecTiVE prOLoNgaTIonS of ${\textsWaB{m}}{=}{\Sum}_{{\maThpzC{P}}{<}0}{\MAThfrAk{g}}_{{\MathPzc{p}}}$ aRe finite dimenSioNal, bECauSe in tHis caSe $W{=}\{0\}.$ THeN, sincE ${\mathfRak{g}}$ iS sImple, it is maximaL by PRopositioN \[prOp-7.3\]. the CrOss maRKed DynKin DiaGram is $$\xYmatrix@r=-.3Pc{
\aLpHAUP_1 & \aLphaup_2 \\
\!\!\medcirc\!\!\!
\ar@3{->}[r]&\!\!\meDcIRC\!\! \\
&\tImes}\quad \Qquad\bEGiN{mATrix}
\qquaD\qUad \\
\AlphAUP_1{=}2\omegAup,\;\; \aLPhAup_2{=}\omegaUp{-}{\epsiLOnUp},
\End{matrIx}$$ With $\deG(\aLphAup_1 | ;\;\;\text{with}\;\;\;
\b egin{cases }
{\ mat hpz c{ R}}_ {\;0 }=\{{\pm}2\ome g aup, \},\\
{\mathpzc{R}}_{ \;{\p m} 1 }=\{ { \m p}{\e psilonu p }+ \ o meg au p, {\m p} { \e psilo nup }-\omeg aup\},\\
{\m at hpzc{R}}_{\; { \p m}2}=\{{\m p}2 {\epsilonup} \}, \\
{\ ma thp z c{R}} _{\ ;{\pm }3}=\{ { \mp}3{ \epsilonu p} { +}\ome g aup,{\m p } 3{ \eps ilonup}{-}\omegau p \} ,
\end{cases}$$ we ob ta i na roo t s ystem of t yp e $\m a thbf{G} _ 2. $ W ith a $2$-dimensi onal Cartan sub algebr a${\ m athfra k{h}} _2 , $ t he graded L ie a lgebra $$ {\sum} _ {{\math p zc{p}}{ =}{-}3 }^3 {\m athf r ak {g }}_ {{ \ mat h pz c{p } }}, \;\; \te xt {w ith}\ ;\;{ \ math fra k{g} }_0{= }{\mathfrak{h }}_ 2{\o p lus }\lan gle{\ math pz c{R}} _{\;0} \rang le \;\text{and}\;
{\m athfrak{g }}_ {\ mat hp zc{p} } {=}\la ngl e{\ mathpzc {R}}_{\ ; {\m at h p z c{ p}}}\rangle \;\tex t{ i f $ {\mathpz c{p}}{ \ ne q} 0 ,$}$$ is a n e ffec t i ve pr olon g at ion of a n fund a me nt al grad ed Lie a lg ebr a o f the thir d kind . It isthe m a ximal one. Ind e ed, by [@MMN2 0 18 T hm . 5.3] th e effective pro l onga tion s o f $ { \text swab{ m} } {= } {\sum}_{{\mathpzc{p }} {<}0}{ \math frak{g}}_{{\m athpzc{p}} } $ are fini te d i me n sional, becaus e inthis case$ W{=}\{0\ }.$ T hen, sin ce ${\mat h f rak{g}}$ is si mpl e,i t i s maximal byP r opos it ion \[p rop -7.3\]. Th e c ros s m ar ked Dynki n diagra mis $ $\ xym atrix @ R=-.3pc{
\a lp hau p_1 & \alpha up_2 \\\! \! \ med circ\!\ ! \! \ ar@3 {- >} [r]& \!\ !\ medci rc\! \ ! \ \
&\tim es}\quad\qq u ad\b eg in {matrix }
\qquad\qua d\\
\alphau p_ 1{= }2\ome g a up,\;\;\alphaup_2{=}\omegaup{- } {\epsil onu p},
\ end{ matrix}$$ wi th $\d eg( \ alphau p_1 | ;\;\;\text{with}\;\;\;
\begin{cases}
_{\mathpzc{R}}_{\;0}=\{{\pm}2\omegaup,\},\\
{\mathpzc{R}}_{\;{\pm}1}=\{{\mp}{\epsilonup}+\omegaup,{\mp}{\epsilonup}-\omegaup\},\\
_{\mathpzc{R}}_{\;{\pm}2}=\{{\mp}2{\epsilonup}\},\\
{\mathpzc{R}}_{\;{\pm}3}=\{{\mp}3{\epsilonup}{+}\omegaup,{\mp}3{\epsilonup}{-}\omegaup\},
\end{cases}$$ we obtain_a root_system_of type_$\mathbf{G}_2.$_With a $2$-dimensional_Cartan subalgebra ${\mathfrak{h}}_2,$_the graded Lie algebra_$${\sum}_{{\mathpzc{p}}{=}{-}3}^3{\mathfrak{g}}_{{\mathpzc{p}}},\;\; \text{with}\;\;
_{\mathfrak{g}}_0{=}{\mathfrak{h}}_2{\oplus}\langle{\mathpzc{R}}_{\;0}\rangle_\;\text{and}\;
{\mathfrak{g}}_{\mathpzc{p}}{=}\langle{\mathpzc{R}}_{\;{\mathpzc{p}}}\rangle \;\text{if ${\mathpzc{p}}{\neq}0,$}$$ is an effective prolongation of an fundamental graded Lie algebra of_the_third kind._It_is_the maximal one. Indeed, by_[@MMN2018 Thm.5.3] the effective prolongations_of ${\textswab{m}}{=}{\sum}_{{\mathpzc{p}}{<}0}{\mathfrak{g}}_{{\mathpzc{p}}}$_are finite dimensional, because in this case $W{=}\{0\}.$_Then,_since ${\mathfrak{g}}$ is_simple, it is maximal by Proposition \[prop-7.3\]. The cross marked_Dynkin diagram is $$\xymatrix@R=-.3pc{
\alphaup_1_& \alphaup_2 _\\
\!\!\medcirc\!\!\!
\ar@3{->}[r]&\!\!\medcirc\!\!_\\
&\times}\quad_\qquad\begin{matrix}
\qquad\quad \\
\alphaup_1{=}2\omegaup,\;\; \alphaup_2{=}\omegaup{-}{\epsilonup},
\end{matrix}$$_with $\deg(\alphaup_1 |
Let $F$ be a real quadratic field with narrow class number equal to $1$. Let $D$, $\mathcal{O}$ and $\mathfrak d$ be the fundamental discriminant, the ring of integers and the different of $F$ respectively. Let $\textrm{N}$ and $\textrm{Tr}$ be the norm and the trace on $F$， defined by $\textrm{N}(a)=a a', \textrm{Tr}(a)=a+a'$ with $a'$ the algebraic conjugate of $a\in F$. We denote $a\gg 0$ for $a\in F$ if $a$ is totally positive, that is $a>0$ and $a'>0$. For $B \subset F$, let $B_+$ denote the subset of totally positive elements in $B$. So $\mathcal O_+$ and $\mathcal O^\times_+$ denote the set of totally positive integers and the set of totally positive units respectively.
For a $2\times 2$ matrix $\mathbf{\gamma}$ in $GL_2^+(F)$, we usually denote its entries by ${\mathbf {\gamma}}=
\begin{smallmatrix}\begin{pmatrix}
a_{\mathbf {\gamma}} &b_{\mathbf {\gamma}}\\
c_{\mathbf {\gamma}} & d_{\mathbf {\gamma}}
\end{pmatrix}
\end{smallmatrix}$ and $\gamma'=
\begin{smallmatrix}\begin{pmatrix}
a_{\mathbf {\gamma}}' &b_{\mathbf {\gamma}}'\\
c_{\mathbf {\gamma}}' & d_{\mathbf {\gamma}}'
\end{pmatrix}
\end{smallmatrix}$. The group $GL_2^+(F)$ acts on two copies of the complex upper half plane $\mathbb{H}^2$ by $\gamma z:=(\gamma z_1, \gamma' z_2)=(\frac{a_{\gamma}z_1+b_{\gamma}}{c_{\gamma}z_1+d_{\gamma}}, \frac{a_{\gamma}'z_2+b_{\gamma}'}{c_{\gamma}'
z_2+d_{\gamma}'})
$ as linear fractional transformations for all $ \gamma\in GL_2^+(F)$ and $ z=(z_1, z_2)\in \mathbb{H}^2$.
Let $\Gamma=\textrm{SL}_2(\mathcal O)$ be the modular group of $2\times 2$ matrices with determinant | Let $ F$ be a real quadratic field with narrow course numeral equal to $ 1$. Let $ D$, $ \mathcal{O}$ and $ \mathfrak d$ be the cardinal discriminant, the ring of integer and the different of $ F$ respectively. Let $ \textrm{N}$ and $ \textrm{Tr}$ be the average and the trace on $ F$ ， defined by $ \textrm{N}(a)=a a', \textrm{Tr}(a)=a+a'$ with $ a'$ the algebraic conjugate of $ a\in F$. We announce $ a\gg 0 $ for $ a\in F$ if $ a$ is wholly positive, that is $ a>0 $ and $ a'>0$. For $ B \subset F$, lease $ B_+$ denote the subset of totally positivist elements in $ B$. therefore $ \mathcal O_+$ and $ \mathcal O^\times_+$ denote the set of wholly positive integers and the set of wholly positive units respectively.
For a $ 2\times 2 $ matrix $ \mathbf{\gamma}$ in $ GL_2^+(F)$, we usually denote its entries by $ { \mathbf { \gamma}}=
\begin{smallmatrix}\begin{pmatrix }
a_{\mathbf { \gamma } } & b_{\mathbf { \gamma}}\\
c_{\mathbf { \gamma } } & d_{\mathbf { \gamma } }
\end{pmatrix }
\end{smallmatrix}$ and $ \gamma'=
\begin{smallmatrix}\begin{pmatrix }
a_{\mathbf { \gamma } }' & b_{\mathbf { \gamma}}'\\
c_{\mathbf { \gamma } }' & d_{\mathbf { \gamma } }'
\end{pmatrix }
\end{smallmatrix}$. The group $ GL_2^+(F)$ acts of the apostles on two copies of the complex upper half plane $ \mathbb{H}^2 $ by $ \gamma z:=(\gamma z_1, \gamma' z_2)=(\frac{a_{\gamma}z_1+b_{\gamma}}{c_{\gamma}z_1+d_{\gamma } }, \frac{a_{\gamma}'z_2+b_{\gamma}'}{c_{\gamma }'
z_2+d_{\gamma }' })
$ as linear fractional transformations for all $ \gamma\in GL_2^+(F)$ and $ z=(z_1, z_2)\in \mathbb{H}^2$.
lease $ \Gamma=\textrm{SL}_2(\mathcal O)$ be the modular group of $ 2\times 2 $ matrices with determinant |
Let $F$ be a real quadratic fleld with narrow class iumber squal to $1$. Let $D$, $\mathcal{O}$ and $\mathfran e$ be uke fundamental discrkminant, tje ring if iitegers and the vjfferenb of $R$ res'ertively. Let $\texjrm{N}$ and $\texdrm{Tr}$ be the nmro cnd the trace on $F$， defined by $\textrm{G}(a)=a a', \trxhrm{Tr}(a)=a+a'$ with $w'$ tht ajgebdaic conjugate of $a\in F$. We denote $z\gg 0$ fog $a\in F$ if $a$ is tptally positive, that is $a>0$ wnd $w'>0$. For $B \subset F$, lft $B_+$ denote the wubset of togally posiuire elements in $B$. So $\mathcal O_+$ and $\mathcal U^\timex_+$ denote tye seh of totally posinive integers and the set of totally positlve uiits respectively.
For a $2\tikes 2$ matrix $\mathbf{\damma}$ in $CL_2^+(Y)$, we usually denote irs entrhes ty ${\mxrhbw {\gzmka}}=
\gegin{slalmmatrix}\begjn{pmatrix}
a_{\nathbf {\gamma}} &b_{\mathbg {\dqmma}}\\
c_{\mathbf {\famma}} & d_{\iathbf {\gamma}}
\end{pmatrix}
\end{smallmatrix}$ dnd $\gamma'=
\begin{smallmatrix}\vegin{pmatrix}
a_{\mathbf {\hamma}}' &b_{\maehbf {\gamma}}'\\
c_{\mathbf {\gamma}}' & d_{\mathbf {\gamma}}'
\end{pmatrif}
\env{soalomwgeid}$. The group $GL_2^+(F)$ acts on two copies of the com[mec lpper half plane $\iathbb{H}^2$ by $\balms z:=(\gamma z_1, \gamoa' z_2)=(\frcd{a_{\famma}z_1+b_{\gamma}}{c_{\gamma}x_1+d_{\gamma}}, \frac{q_{\gamma}'z_2+b_{\gwmma}'}{v_{\gamma}'
z_2+d_{\gamma}'})
$ as linear fraxtional tranfdormations for all $ \gamma\in YL_2^+(F)$ anc $ z=(z_1, z_2)\in \mathbb{H}^2$.
Let $\Gamma=\tertrm{SL}_2(\jathcal O)$ bf the modhuar group of $2\timds 2$ mdtrices with determinant | Let $F$ be a real quadratic field class equal to Let $D$, $\mathcal{O}$ fundamental the ring of and the different $F$ respectively. Let $\textrm{N}$ and $\textrm{Tr}$ the norm and the trace on $F$， defined by $\textrm{N}(a)=a a', \textrm{Tr}(a)=a+a'$ with the algebraic conjugate of $a\in F$. We denote $a\gg 0$ for $a\in F$ $a$ totally that $a>0$ and $a'>0$. For $B \subset F$, let $B_+$ denote the subset of totally positive elements $B$. So $\mathcal O_+$ and $\mathcal O^\times_+$ denote set of totally positive and the set of totally units For a 2$ $\mathbf{\gamma}$ $GL_2^+(F)$, we usually its entries by ${\mathbf {\gamma}}= \begin{smallmatrix}\begin{pmatrix} a_{\mathbf {\gamma}} &b_{\mathbf {\gamma}}\\ c_{\mathbf {\gamma}} & d_{\mathbf {\gamma}} \end{pmatrix} \end{smallmatrix}$ $\gamma'= \begin{smallmatrix}\begin{pmatrix} &b_{\mathbf {\gamma}}'\\ {\gamma}}' d_{\mathbf \end{pmatrix} \end{smallmatrix}$. The acts on two copies of the plane $\mathbb{H}^2$ by $\gamma z:=(\gamma z_1, \gamma' z_2)=(\frac{a_{\gamma}z_1+b_{\gamma}}{c_{\gamma}z_1+d_{\gamma}}, z_2+d_{\gamma}'}) $ linear fractional transformations for all $ GL_2^+(F)$ and $ z=(z_1, z_2)\in \mathbb{H}^2$. Let $\Gamma=\textrm{SL}_2(\mathcal be the modular group of $2\times 2$ matrices with determinant |
Let $F$ be a real quadratic field With narrow Class NumBer EqUal tO $1$. Let $d$, $\mathcal{O}$ and $\maTHfraK d$ be the fundamental discRiminAnT, The rINg Of intEgers anD ThE DIffErEnT of $f$ rESpEctivEly. let $\textRm{N}$ and $\textRm{TR}$ bE the norm and tHE tRace on $F$， defIneD by $\textrm{N}(a)=a A', \teXtrm{Tr}(A)=a+A'$ wiTH $a'$ the AlgEbraiC conjuGAte of $a\In F$. We denoTe $A\Gg 0$ for $a\IN F$ if $a$ is TOTaLly pOsitive, that is $a>0$ and $A'>0$. foR $b \subset F$, let $B_+$ deNote thE sUBsET Of tOtaLly positivE eLemenTS in $B$. So $\mAThCAL o_+$ anD $\Mathcal O^\times_+$ Denote the seT Of tOtally PoSitIVe inteGers aNd THe sEt of totally PosiTive units RespecTIvely.
FoR A $2\times 2$ mAtrix $\mAthBf{\gAmma}$ IN $Gl_2^+(F)$, We uSuALly DEnOte ITs eNtries by ${\MaThBf {\gamMa}}=
\beGIN{SMallMatRix}\bEgin{pMatrix}
a_{\mathbf {\GamMa}} &b_{\mAThbF {\gammA}}\\
c_{\matHbf {\gAmMa}} & d_{\maThbf {\gaMma}}
\enD{pMatrix}
\end{smallmAtriX}$ and $\gamma'=
\BegIn{SmaLlMatriX}\Begin{pMatRix}
A_{\mathbf {\Gamma}}' &b_{\mAThbF {\gAMMA}}'\\
c_{\Mathbf {\gamma}}' & d_{\mathbf {\GaMMA}}'
\eNd{pmatriX}
\end{smALlMaTRix}$. The grOuP $GL_2^+(f)$ actS ON two cOpieS Of The complEx uppeR HaLf Plane $\maThBb{H}^2$ by $\gAmMa z:=(\GamMa z_1, \gaMMa' z_2)=(\fRac{a_{\gaMma}z_1+b_{\gamMa}}{c_{\gaMMa}z_1+d_{\gamma}}, \frac{a_{\GAmma}'z_2+b_{\gamma}'}{c_{\gAMmA}'
Z_2+D_{\gAMma}'})
$ aS liNear fractioNal tRAnsfOrmaTIoNs fOR all $ \gAmma\iN Gl_2^+(f)$ aND $ z=(z_1, z_2)\in \mathbb{H}^2$.
Let $\GamMa=\Textrm{sL}_2(\matHcal O)$ be the modUlar group oF $2\TIMes 2$ matriCes wITh DEterminant |
Let $F$ be a real quadrat ic field w ith n arr owcl assnumb er equal to $1 $ . Le t $D$, $\mathcal{O}$ a nd $\ ma t hfra k d $ bethe fun d am e n tal d is cri mi n an t, th e r ing ofintegers a ndth e differento f$F$ respec tiv ely. Let $\t ext rm{N}$ a nd$ \text rm{ Tr}$be the norm a nd the tr ac e on $F $ ， defin e d b y $\ textrm{N}(a)=a a' , \ t extrm{Tr}(a)=a +a'$ w it h $ a ' $ t healgebraicco njuga t e of $a \ in F $ . W e denote $a\gg 0$ for $a\ i n F $ if $ a$ is totall y pos it i ve, that is $a >0$and $a'>0 $. For $B \sub s et F$,let $B _+$ de note th esub se t of to tal l y p ositiveel em entsin $ B $ . So $ \ma thca l O_+ $ and $\mathc alO^\t i mes _+$ d enote the s et of total ly po si tive integers a nd t he set of to ta lly p ositi v e unit s r esp ectivel y.
For a $ 2\ t i m es 2$ matrix $\mathb f{ \ g am ma}$ in$GL_2^ + (F )$ , we usua ll y d enot e its e ntri e sby ${\ma thbf { \ ga mm a}}=
\b eg in{sma ll mat rix }\beg i n{pm atrix}
a_{\ma thbf{ \gamma}} &b_{\ m athbf {\gamma } }\ \ c _ {\ma thb f {\gamma}} & d _ {\ma thbf {\ gam m a}}
\end{ pm a tr i x}
\end{smallmatri x} $ and$\gam ma'=
\begin{ smallmatri x } \ begin{pm atri x }a_{\mathbf {\g amma} }' &b_{\ma t hbf {\ga mma}} '\\
c_{ \mathbf { \ g amma}}'& d _{\ mat hbf { \g amma}}'
\end { p matr ix }
\en d{s mallmat rix }$. Th e g ro up $GL_2^ +(F)$ ac ts o ntw o c opies of the c om ple xupp er ha l f plan e $\m athb b{ H} ^ 2$by $\ga m ma z :=(\ ga mm a z_ 1,\g amma' z_2 ) =(\ frac{a_ {\gamma}z _1+ b _{\g am ma }}{c_{\ gamma}z_1+d_{ \g amma}}, \f ra c{a _{\gam m a }'z_2+b_ {\gamma}'}{c_{\gamma}'z _2+d_{\ gam ma}'} )
$as linear fr action alt ransfo rmatio ns fo rall $ \ga m m a\ inGL _2^+(F)$ a n d $z=(z_ 1, z_2 )\in \m athbb{H}^2$.
Let$ \Ga mma=\textrm{S L}_ 2(\m a t hc alO )$ beth e mo d u lar group of $2 \times 2$ma t ri ces with d e ter mi nant |
Let $F$_be a_real quadratic field with_narrow class_number_equal to_$1$._Let $D$, $\mathcal{O}$_and $\mathfrak d$_be the fundamental discriminant,_the ring of_integers_and the different of $F$ respectively. Let $\textrm{N}$ and $\textrm{Tr}$ be the norm and_the_trace on_$F$，_defined_by $\textrm{N}(a)=a a', \textrm{Tr}(a)=a+a'$ with_$a'$ the algebraic conjugate of_$a\in F$._We denote $a\gg 0$ for $a\in F$ if_$a$_is totally positive,_that is $a>0$ and $a'>0$. For $B \subset F$,_let $B_+$ denote the subset of_totally positive elements_in_$B$._So $\mathcal O_+$ and_$\mathcal O^\times_+$ denote the set of_totally positive integers and the set_of totally positive units respectively.
For a $2\times_2$ matrix $\mathbf{\gamma}$ in $GL_2^+(F)$, we_usually denote its entries by_${\mathbf {\gamma}}=
\begin{smallmatrix}\begin{pmatrix}
_a_{\mathbf {\gamma}} &b_{\mathbf {\gamma}}\\
c_{\mathbf_{\gamma}} & d_{\mathbf_{\gamma}}
\end{pmatrix}
_\end{smallmatrix}$ and $\gamma'=
_\begin{smallmatrix}\begin{pmatrix}
a_{\mathbf {\gamma}}' &b_{\mathbf {\gamma}}'\\
c_{\mathbf_{\gamma}}' & d_{\mathbf_{\gamma}}'
\end{pmatrix}
\end{smallmatrix}$. The group_$GL_2^+(F)$_acts on two_copies_of_the complex_upper half plane_$\mathbb{H}^2$_by $\gamma_z:=(\gamma_z_1, \gamma' z_2)=(\frac{a_{\gamma}z_1+b_{\gamma}}{c_{\gamma}z_1+d_{\gamma}}, \frac{a_{\gamma}'z_2+b_{\gamma}'}{c_{\gamma}'
z_2+d_{\gamma}'})
$ as linear_fractional_transformations for all $ \gamma\in GL_2^+(F)$_and $ z=(z_1, z_2)\in_\mathbb{H}^2$.
Let_$\Gamma=\textrm{SL}_2(\mathcal O)$ be the_modular group of $2\times 2$_matrices with determinant |
arises as $$\operatorname{Sym}^2(\mathbf{20}) = \mathbf{175} \oplus \mathbf{35},$$ where $\mathbf{35} = \mathfrak{su}(6)$ is the adjoint representation.
The Mathieu group ${\mathrm M}_{21}$ is not sporadic, being isomorphic to $\mathrm{PSL}_3({{\mathds}F}_4) = {\mathrm L}_3(4)$. It embeds into ${\mathrm{PSU}}(6)$ via the 6-dimensional representation of its 6-fold cover $6{\mathrm M}_{21}$ (which GAP’s character table library only knows under the name “$6.{\mathrm L}_3(4)$,” but we will use the Mathieu name). The adjoint representation $\mathbf{35}$ remains simple when restricted to ${\mathrm M}_{21}$, and so the conformal vector $\nu$ is the unique ${\mathrm M}_{21}$-fixed spin-$2$ field. On the other hand, $\mathbf{20}$ breaks over $2{\mathrm M}_{21}$ as $\mathbf{10} \oplus \overline{\mathbf{10}}$, where $\mathbf{10}$ and $\overline{\mathbf{10}}$ are a dual pair of $10$-dimensional complex irreps of $2{\mathrm M}_{21}$, and so $\operatorname{Sym}^2(\mathbf{20})$, and hence $\mathbf{175}$, has a unique fixed point when restricted to ${\mathrm M}_{21}$, picking out the superconformal vector $\tau$.
We remark that, although ${\mathrm M}_{21} = {\mathrm L}_3(4)$ is not sporadic, it is exceptional, having very large Schur multiplier ($3 \times 4^2$) and outer automorphism group ($D_{12} = 2 \times S_3$). The outer automorphism of order $3$ permutes the three double covers. In particular, there is no (interesting) group “$2.{\mathrm M}_{21}.3$,” and so no “$6.{\mathrm M}_{21}.3$” which could embed into ${\mathrm{SU}}(6)$. (GAP can compute, in a few seconds, that $\operatorname{Aut}(6.{\mathrm M}_{21}) = {\mathrm M}_{21}{:}2^2$.)
There are three groups of shape ${\mathrm M}_{21}{:}2$, coming from the three conjugacy classes of order-$2$ element in $\operatorname{ | arises as $ $ \operatorname{Sym}^2(\mathbf{20 }) = \mathbf{175 } \oplus \mathbf{35},$$ where $ \mathbf{35 } = \mathfrak{su}(6)$ is the adjoint representation.
The Mathieu group $ { \mathrm M}_{21}$ is not sporadic, being isomorphic to $ \mathrm{PSL}_3({{\mathds}F}_4) = { \mathrm L}_3(4)$. It embeds into $ { \mathrm{PSU}}(6)$ via the 6 - dimensional theatrical performance of its 6 - fold binding $ 6{\mathrm M}_{21}$ (which GAP ’s character table library merely know under the name “ $ 6.{\mathrm L}_3(4)$, ” but we will use the Mathieu name). The adjoint representation $ \mathbf{35}$ persist dim-witted when restricted to $ { \mathrm M}_{21}$, and thus the conformal vector $ \nu$ is the unique $ { \mathrm M}_{21}$-fixed spin-$2 $ field. On the early hand, $ \mathbf{20}$ breaks over $ 2{\mathrm M}_{21}$ as $ \mathbf{10 } \oplus \overline{\mathbf{10}}$, where $ \mathbf{10}$ and $ \overline{\mathbf{10}}$ are a dual pair of $ 10$-dimensional complex irreps of $ 2{\mathrm M}_{21}$, and therefore $ \operatorname{Sym}^2(\mathbf{20})$, and hence $ \mathbf{175}$, has a unique fixed point when restrict to $ { \mathrm M}_{21}$, picking out the superconformal vector $ \tau$.
We remark that, although $ { \mathrm M}_{21 } = { \mathrm L}_3(4)$ is not sporadic, it is exceptional, receive very large Schur multiplier ($ 3 \times 4 ^ 2 $) and outer automorphism group ($ D_{12 } = 2 \times S_3 $). The outer automorphism of ordering $ 3 $ permutes the three double covers. In particular, there is no (interesting) group “ $ 2.{\mathrm M}_{21}.3 $, ” and therefore no “ $ 6.{\mathrm M}_{21}.3 $ ” which could embed into $ { \mathrm{SU}}(6)$. (GAP can compute, in a few seconds, that $ \operatorname{Aut}(6.{\mathrm M}_{21 }) = { \mathrm M}_{21}{:}2 ^ 2$.)
There are three groups of shape $ { \mathrm M}_{21}{:}2 $, coming from the three conjugacy classes of order-$2 $ element in $ \operatorname { | arlses as $$\operatorname{Sym}^2(\mxthbf{20}) = \mathbf{175} \oklys \matibf{35},$$ whede $\mathbw{35} = \mathfrak{su}(6)$ is the adjoint rwprestutation.
The Mathieu gfoup ${\mathgm M}_{21}$ is nit skoradic, being isomorphic bj $\mafmrm{PSN}_3({{\nathds}F}_4) = {\mathrk L}_3(4)$. It embads into ${\mathrk{PRU}}(6)$ via the 6-dimensional representation of its 6-flld cover $6{\mathtm M}_{21}$ (erich GAP’s character table library only knows lnder the name “$6.{\mayhrm L}_3(4)$,” but we will use the Matjieu name). The adjolnt represebtatyin $\mathbf{35}$ reoains simple when restticted to ${\mathrm M}_{21}$, and so the covformcl vector $\no$ js jhe unique ${\methrm I}_{21}$-fixed spin-$2$ npeld. On the otner hand, $\mathbn{20}$ breeks iver $2{\mathrm M}_{21}$ as $\mathuf{10} \oplus \overline{\matrbf{10}}$, where $\mcthbf{10}$ and $\overline{\matybd{10}}$ are a dgal oqir of $10$-dmmehsionap ckmplex irrsps of $2{\mathem M}_{21}$, and so $\operatotnwne{Sym}^2(\mathbf{20})$, ahd hense $\mathbf{175}$, has a unique fixed point when rtstridted to ${\mathrm M}_{21}$, pickint out the superconforlal vectow $\tau$.
We remark that, although ${\mathrm M}_{21} = {\mathrm L}_3(4)$ iv not rpoxqdic, kr ls exceptional, having very large Schur multipjjet ($3 \times 4^2$) and ouber automorphism gtokp ($Q_{12} = 2 \times S_3$). Tfe outzd zutomorphism of orfer $3$ petmutes the thret dounle covers. In particular, thwre is no (innerewting) group “$2.{\mathrm M}_{21}.3$,” and so nu “$6.{\majhrm M}_{21}.3$” which could embed into ${\matgrm{SU}}(6)$. (GAP cwn computs, in a few secondr, tmat $\operatorname{Aut}(6.{\mathrm M}_{21}) = {\iathrm M}_{21}{:}2^2$.)
Tiere cre thred grpups os shape ${\mahhrm M}_{21}{:}2$, coming from the tjree eonjucacy classfs of order-$2$ element in $\operatorieme{ | arises as $$\operatorname{Sym}^2(\mathbf{20}) = \mathbf{175} \oplus \mathbf{35},$$ = is the representation. The Mathieu sporadic, isomorphic to $\mathrm{PSL}_3({{\mathds}F}_4) {\mathrm L}_3(4)$. It into ${\mathrm{PSU}}(6)$ via the 6-dimensional representation its 6-fold cover $6{\mathrm M}_{21}$ (which GAP’s character table library only knows under name “$6.{\mathrm L}_3(4)$,” but we will use the Mathieu name). The adjoint representation remains when to M}_{21}$, and so the conformal vector $\nu$ is the unique ${\mathrm M}_{21}$-fixed spin-$2$ field. On the hand, $\mathbf{20}$ breaks over $2{\mathrm M}_{21}$ as $\mathbf{10} \overline{\mathbf{10}}$, where $\mathbf{10}$ and are a dual pair of complex of $2{\mathrm and $\operatorname{Sym}^2(\mathbf{20})$, hence $\mathbf{175}$, has unique fixed point when restricted to ${\mathrm M}_{21}$, picking out the superconformal vector $\tau$. We remark that, ${\mathrm M}_{21} L}_3(4)$ is sporadic, is having very large ($3 \times 4^2$) and outer automorphism 2 \times S_3$). The outer automorphism of order permutes the double covers. In particular, there is (interesting) group “$2.{\mathrm M}_{21}.3$,” and so no “$6.{\mathrm which could embed into ${\mathrm{SU}}(6)$. (GAP can compute, in a few seconds, that $\operatorname{Aut}(6.{\mathrm M}_{21}) M}_{21}{:}2^2$.) There are three of shape ${\mathrm coming the conjugacy of order-$2$ in $\operatorname{ | arises as $$\operatorname{Sym}^2(\maThbf{20}) = \mathbf{175} \Oplus \MatHbf{35},$$ WhEre $\mAthbF{35} = \mathfrak{su}(6)$ is tHE adjOint representation.
The MAthieU gROup ${\mAThRm M}_{21}$ is Not sporADiC, BEinG iSoMorPhIC tO $\mathRm{PsL}_3({{\mathdS}F}_4) = {\mathrm L}_3(4)$. IT emBeDs into ${\mathrm{psU}}(6)$ Via the 6-dimeNsiOnal represenTatIon of iTs 6-FolD Cover $6{\MatHrm M}_{21}$ (wHich GAp’S charaCter table LiBRary onLY knows uNDEr The nAme “$6.{\mathrm L}_3(4)$,” but we wiLL uSE the Mathieu namE). The adJoINt REPreSenTation $\mathBf{35}$ RemaiNS simple WHeN REStrICted to ${\mathrm M}_{21}$, And so the conFOrmAl vectOr $\Nu$ iS The uniQue ${\maThRM M}_{21}$-fIxed spin-$2$ fieLd. On The other hAnd, $\matHBf{20}$ breakS Over $2{\matHrm M}_{21}$ as $\MatHbf{10} \OpluS \OvErLinE{\mAThbF{10}}$, WhEre $\MAthBf{10}$ and $\oveRlInE{\mathBf{10}}$ arE A DUAl paIr oF $10$-dimEnsioNal complex irrEps Of $2{\maTHrm m}_{21}$, and sO $\operAtorNaMe{Sym}^2(\Mathbf{20})$, And heNcE $\mathbf{175}$, has a uniqUe fiXed point wHen ReStrIcTed to ${\MAthrm M}_{21}$, PicKinG out the SupercoNForMaL VECtOr $\tau$.
We remark that, aLtHOUgH ${\mathrm M}_{21} = {\Mathrm l}_3(4)$ Is NoT Sporadic, It Is eXcepTIOnal, hAvinG VeRy large SChur muLTiPlIer ($3 \timeS 4^2$) aNd outeR aUtoMorPhism GRoup ($d_{12} = 2 \times s_3$). The outeR autoMOrphism of order $3$ PErmutes the thrEE dOUBlE CoveRs. IN particular, TherE Is no (InteREsTinG) Group “$2.{\MathrM M}_{21}.3$,” ANd SO no “$6.{\mathrm M}_{21}.3$” which coulD eMbed inTo ${\matHrm{SU}}(6)$. (GAP can coMpute, in a feW SEConds, thaT $\opeRAtORname{Aut}(6.{\mathrm m}_{21}) = {\mathRm M}_{21}{:}2^2$.)
There arE Three groUps of Shape ${\matHrm M}_{21}{:}2$, cominG FRom the thRee ConJugAcy CLAsSes of order-$2$ eleMENt in $\OpEratornAme{ | arises as $$\operatorname {Sym}^2(\m athbf {20 })=\mat hbf{ 175} \oplus \m a thbf {35},$$ where $\mathbf {35}=\ math f ra k{su} (6)$ is th e adj oi nt re pr e se ntati on.
The M athieu gro up${ \mathrm M}_{ 2 1} $ is not s por adic, beingiso morphi cto$ \math rm{ PSL}_ 3({{\m a thds}F }_4) = {\ ma t hrm L} _ 3(4)$.I t e mbed s into ${\mathrm{ P SU } }(6)$ via the6-dime ns i on a l re pre sentationof its6 -fold c o ve r $ 6{\ m athrm M}_{21} $ (which GA P ’scharac te r t a ble li brary o n lyknows under the name “$6 .{\mat h rm L}_3 ( 4)$,” b ut wewil l u se t h eMa thi eu nam e ). Th e ad joint re pr es entat ion$ \ m a thbf {35 }$ r emain s simple when re stri c ted to $ {\mat hrmM} _{21} $, and so t he conformal vect or $ \nu$ is t heun iqu e${\ma t hrm M} _{2 1}$ -fixedspin-$2 $ fi el d . On the other hand, $ \m a t hb f{20}$ b reakso ve r$ 2{\mathr mM}_ {21} $ as $\ math b f{ 10} \opl us \ov e rl in e{\math bf {10}}$ ,whe re$\mat h bf{1 0}$ an d $\over line{ \ mathbf{10}}$ a r e a dual pair of $ 10 $ -dim ens ional compl ex i r reps of$ 2{ \ma t hrm M }_{21 }$ , a n d so $\operatorname {S ym}^2( \math bf{20})$, and hence $\m a t h bf{175}$ , ha s a unique fixed p ointwhen restr i cted to${\ma thrm M}_ {21}$, pi c k ing outthe su per con f o rm al vector $\t a u $.
We remark th at, alt hou gh${\ mat hr m M}_{21} = {\mat hr mL} _3 (4) $ isn ot spora di c,it is exce p tional , hav ingve ry lar ge Schu r m u l tipl ie r($3\ti me s 4^2 $) a n d o uter au tomorphis m g r oup($ D_ {12} =2 \times S_3$ ). The outer a uto morphi s m of orde r $3$ permutes the thre e double co vers. Inparticula r,thereisn o (int eresti ng) g ro up“ $ 2.{\m a t hr m M }_ {21}.3$,”a n d s o no“$ 6.{\ mathrmM}_{21}.3$” whichc oul d embed into${\ math r m {S U}} ( 6) $ . ( GA P ca n compute, in a f ew seconds ,t ha t $\operat o rna me {Aut}(6 .{\math rm M} _ {21}) = {\mathrm M}_{21}{ :} 2^2$ . )
T here are t hree gro ups of sh a pe ${ \ ma thrmM}_ {21}{: }2 $,comin g from the thre e conj ug acy cl asses o f order- $2$ element in $\operat orname { | arises_as $$\operatorname{Sym}^2(\mathbf{20})_= \mathbf{175} \oplus \mathbf{35},$$_where $\mathbf{35}_=_\mathfrak{su}(6)$ is_the_adjoint representation.
The Mathieu_group ${\mathrm M}_{21}$_is not sporadic, being_isomorphic to $\mathrm{PSL}_3({{\mathds}F}_4)_=_{\mathrm L}_3(4)$. It embeds into ${\mathrm{PSU}}(6)$ via the 6-dimensional representation of its 6-fold cover_$6{\mathrm_M}_{21}$ (which_GAP’s_character_table library only knows under_the name “$6.{\mathrm L}_3(4)$,” but_we will_use the Mathieu name). The adjoint representation $\mathbf{35}$_remains_simple when restricted_to ${\mathrm M}_{21}$, and so the conformal vector $\nu$_is the unique ${\mathrm M}_{21}$-fixed spin-$2$_field. On the_other_hand,_$\mathbf{20}$ breaks over $2{\mathrm_M}_{21}$ as $\mathbf{10} \oplus \overline{\mathbf{10}}$, where_$\mathbf{10}$ and $\overline{\mathbf{10}}$ are a dual_pair of $10$-dimensional complex irreps of $2{\mathrm_M}_{21}$, and so $\operatorname{Sym}^2(\mathbf{20})$, and hence_$\mathbf{175}$, has a unique fixed_point when_restricted to ${\mathrm M}_{21}$, picking_out the superconformal_vector $\tau$.
We_remark that, although_${\mathrm M}_{21} = {\mathrm L}_3(4)$ is_not sporadic, it_is exceptional, having very large Schur_multiplier_($3 \times 4^2$)_and_outer_automorphism group_($D_{12} = 2_\times_S_3$). The_outer_automorphism of order $3$ permutes the_three_double covers. In particular, there is no_(interesting) group “$2.{\mathrm M}_{21}.3$,”_and_so no “$6.{\mathrm M}_{21}.3$”_which could embed into ${\mathrm{SU}}(6)$._(GAP can compute, in a few_seconds, that_$\operatorname{Aut}(6.{\mathrm M}_{21})_= {\mathrm M}_{21}{:}2^2$.)
There are three groups of shape ${\mathrm M}_{21}{:}2$, coming_from the three conjugacy classes of_order-$2$ element in $\operatorname{ |
A_2}
\\
& L_2
}}$$ Then, for every $w\in\operatorname{W}^{j}_{\!Z}({Y},L_0)$, $\lambda_1\in\operatorname{W}^{i_1}({X},K_1)$ and $\lambda_2\in\operatorname{W}^{i_2}({X},K_2)$, we have $$\lambda_2{\mathchoice{\mathop{\cdot}\limits_{A_2}}{\cdot_{A_2}}{}{}}\big(\lambda_1{\mathchoice{\mathop{\cdot}\limits_{A_1}}{\cdot_{A_1}}{}{}}w\big)=\big(\lambda_2\cdot\lambda_1\big){\mathchoice{\mathop{\cdot}\limits_{A_3}}{\cdot_{A_3}}{}{}}w$$ in $\operatorname{W}^{i_1+i_2+j}_{\!Z}({Y},L_2)$.
The first part follows from Lemmas \[lemm:groupoid\] and \[lemm:invert\]. The rest is direct: $$\begin{array}{rlcr}
\lambda_2{\mathchoice{\mathop{\cdot}\limits_{A_2}}{\cdot_{A_2}}{}{}}\big(\lambda_1{\mathchoice{\mathop{\cdot}\limits_{A_1}}{\cdot_{A_1}}{}{}}w\big)
& = {{A_2}{}^{{\circlearrowleft}}}\big(\pi_{Y}^*\lambda_2\cdot{{A_1}{}^{{\circlearrowleft}}}(\pi_{Y}^*\lambda_1\cdot w)\big)
& \text{by definition}
\\
& = {{A_2}{}^{{\circlearrowleft}}}\circ {{(1\otimes A_1)}{}^{{\circlearrowleft}}}\big(\pi_{Y}^*\lambda_2\cdot\pi_{Y}^*\lambda_1\cdot w\big)
& \text{by Proposition~\ref{prop:al-comp}\,\eqref{it:al-tens}}
\\
& = {{A_3}{}^{{\circlearrowleft}}}(\pi_{Y}^*\lambda_2\cdot\pi_{Y}^*\lambda_1\cdot w)
& \text{by~\eqref{eq:assoc} and Prop.~\ref{prop:al-comp}}
\\
& = {{A_3}{}^{{\circlearrowleft | A_2 }
\\
& L_2
} } $ $ Then, for every $ w\in\operatorname{W}^{j}_{\!Z}({Y},L_0)$, $ \lambda_1\in\operatorname{W}^{i_1}({X},K_1)$ and $ \lambda_2\in\operatorname{W}^{i_2}({X},K_2)$, we have $ $ \lambda_2{\mathchoice{\mathop{\cdot}\limits_{A_2}}{\cdot_{A_2}}{}{}}\big(\lambda_1{\mathchoice{\mathop{\cdot}\limits_{A_1}}{\cdot_{A_1}}{}{}}w\big)=\big(\lambda_2\cdot\lambda_1\big){\mathchoice{\mathop{\cdot}\limits_{A_3}}{\cdot_{A_3}}{}{}}w$$ in $ \operatorname{W}^{i_1+i_2+j}_{\!Z}({Y},L_2)$.
The first part follows from Lemmas \[lemm: groupoid\ ] and \[lemm: invert\ ]. The remainder is lineal: $ $ \begin{array}{rlcr }
\lambda_2{\mathchoice{\mathop{\cdot}\limits_{A_2}}{\cdot_{A_2}}{}{}}\big(\lambda_1{\mathchoice{\mathop{\cdot}\limits_{A_1}}{\cdot_{A_1}}{}{}}w\big)
& = { { A_2}{}^{{\circlearrowleft}}}\big(\pi_{Y}^*\lambda_2\cdot{{A_1}{}^{{\circlearrowleft}}}(\pi_{Y}^*\lambda_1\cdot w)\big)
& \text{by definition }
\\
& = { { A_2}{}^{{\circlearrowleft}}}\circ { { (1\otimes A_1)}{}^{{\circlearrowleft}}}\big(\pi_{Y}^*\lambda_2\cdot\pi_{Y}^*\lambda_1\cdot w\big)
& \text{by Proposition~\ref{prop: al - comp}\,\eqref{it: al - tens } }
\\
& = { { A_3}{}^{{\circlearrowleft}}}(\pi_{Y}^*\lambda_2\cdot\pi_{Y}^*\lambda_1\cdot w)
& \text{by~\eqref{eq: assoc } and Prop.~\ref{prop: al - comp } }
\\
& = { { A_3}{}^{{\circlearrowleft | A_2}
\\
& L_2
}}$$ Then, for every $w\in\operauorname{W}^{j}_{\!Z}({Y},L_0)$, $\lambbq_1\in\opecatornaje{W}^{i_1}({X},K_1)$ avd $\lambda_2\in\operatorname{W}^{i_2}({X},K_2)$, xe hqve $$\lqmbda_2{\mathchoice{\mathop{\caot}\limits_{W_2}}{\cdot_{A_2}}{}{}}\bit(\lamuda_1{\mathchoice{\matikp{\cdot}\lliits_{Z_1}}{\gdot_{A_1}}{}{}}c\bmg)=\big(\lambda_2\cdot\kambda_1\big){\mdthchoice{\matho[{\caoc}\limits_{A_3}}{\cdot_{A_3}}{}{}}w$$ in $\operatorname{W}^{i_1+i_2+j}_{\!Z}({Y},J_2)$.
The fitsh part follows frok Lemjas \[lemm:groupoid\] and \[lemm:invert\]. The dest is direct: $$\begin{srray}{rlcr}
\lambda_2{\mathchoice{\mwthoo{\cdot}\limits_{A_2}}{\cdot_{A_2}}{}{}}\blg(\lambda_1{\matychoyxe{\mathop{\cdot}\uimits_{A_1}}{\cdou_{A_1}}{}{}}c\big)
& = {{A_2}{}^{{\circmearrowleft}}}\big(\pi_{Y}^*\lambda_2\cdot{{A_1}{}^{{\cirzlearxowleft}}}(\pi_{Y}^*\lqmvda_1\fgot w)\big)
& \teet{by dvfinition}
\\
& = {{A_2}{}^{{\clgclearrmwleft}}}\corc {{(1\otimes A_1)}{}^{{\civcleacrowoeft}}}\big(\pi_{Y}^*\lambda_2\cdot\pm_{Y}^*\lambda_1\cdot w\big)
& \tevt{by Propmsntion~\ref{prop:al-comp}\,\eqrwf{ut:al-tgns}}
\\
& = {{D_3}{}^{{\cirzoeafroslxft}}}(\li_{Y}^*\lamhda_2\rdot\pi_{Y}^*\lambsa_1\cdot w)
& \tezt{by~\eqref{eq:assoc} anc [gpp.~\ref{prop:al-ckmp}}
\\
& = {{A_3}{}^{{\siwclearrowleft | A_2} \\ & L_2 }}$$ Then, for $\lambda_1\in\operatorname{W}^{i_1}({X},K_1)$ $\lambda_2\in\operatorname{W}^{i_2}({X},K_2)$, we $$\lambda_2{\mathchoice{\mathop{\cdot}\limits_{A_2}}{\cdot_{A_2}}{}{}}\big(\lambda_1{\mathchoice{\mathop{\cdot}\limits_{A_1}}{\cdot_{A_1}}{}{}}w\big)=\big(\lambda_2\cdot\lambda_1\big){\mathchoice{\mathop{\cdot}\limits_{A_3}}{\cdot_{A_3}}{}{}}w$$ in $\operatorname{W}^{i_1+i_2+j}_{\!Z}({Y},L_2)$. Lemmas and \[lemm:invert\]. The is direct: $$\begin{array}{rlcr} & = {{A_2}{}^{{\circlearrowleft}}}\big(\pi_{Y}^*\lambda_2\cdot{{A_1}{}^{{\circlearrowleft}}}(\pi_{Y}^*\lambda_1\cdot w)\big) & \text{by \\ & = {{A_2}{}^{{\circlearrowleft}}}\circ {{(1\otimes A_1)}{}^{{\circlearrowleft}}}\big(\pi_{Y}^*\lambda_2\cdot\pi_{Y}^*\lambda_1\cdot w\big) & \text{by Proposition~\ref{prop:al-comp}\,\eqref{it:al-tens}} \\ & = w) & \text{by~\eqref{eq:assoc} and Prop.~\ref{prop:al-comp}} \\ & = {{A_3}{}^{{\circlearrowleft | A_2}
\\
& L_2
}}$$ Then, for every $w\in\operatorName{W}^{j}_{\!Z}({Y},L_0)$, $\lAmbda_1\In\oPerAtOrnaMe{W}^{i_1}({x},K_1)$ and $\lambda_2\in\oPEratOrname{W}^{i_2}({X},K_2)$, we have $$\lambda_2{\MathcHoICe{\maTHoP{\cdot}\Limits_{A_2}}{\CDoT_{a_2}}{}{}}\Big(\LaMbDa_1{\mAtHChOice{\mAthOp{\cdot}\lImits_{A_1}}{\cdot_{a_1}}{}{}}w\bIg)=\Big(\lambda_2\cdoT\LaMbda_1\big){\matHchOice{\mathop{\cdOt}\lImits_{A_3}}{\CdOt_{A_3}}{}{}}W$$ In $\opeRatOrnamE{W}^{i_1+i_2+j}_{\!Z}({y},l_2)$.
The fiRst part foLlOWs from lEmmas \[leMM:GrOupoId\] and \[lemm:invert\]. ThE ReST is direct: $$\begin{Array}{rLcR}
\LaMBDa_2{\mAthChoice{\mathOp{\Cdot}\lIMits_{A_2}}{\cdOT_{A_2}}{}{}}\BIG(\LamBDa_1{\mathchoice{\mAthop{\cdot}\liMIts_{a_1}}{\cdot_{A_1}}{}{}}W\bIg)
& = {{A_2}{}^{{\CIrcleaRrowlEfT}}}\Big(\Pi_{Y}^*\lambda_2\cdOt{{A_1}{}^{{\cIrclearroWleft}}}(\pI_{y}^*\lambda_1\CDot w)\big)
& \Text{by DefIniTion}
\\
& = {{a_2}{}^{{\CiRcLeaRrOWleFT}}}\cIrc {{(1\OTimEs A_1)}{}^{{\circlEaRrOwlefT}}}\big(\PI_{y}^*\LAmbdA_2\cdOt\pi_{y}^*\lambDa_1\cdot w\big)
& \texT{by propOSitIon~\reF{prop:Al-coMp}\,\Eqref{It:al-teNs}}
\\
& = {{A_3}{}^{{\ciRcLearrowleft}}}(\pi_{Y}^*\lAmbdA_2\cdot\pi_{Y}^*\lAmbDa_1\CdoT w)
& \Text{bY~\Eqref{eQ:asSoc} And Prop.~\Ref{prop:AL-coMp}}
\\
& = {{a_3}{}^{{\CIRcLearrowleft | A_2}
\\
& L_2
}}$$ Then, f or every $ w\in\ ope rat or name {W}^ {j}_{\!Z}({Y}, L _0)$ , $\lambda_1\in\operat ornam e{ W }^{i _ 1} ({X}, K_1)$ a n d$ \ lam bd a_ 2\i n\ o pe rator nam e{W}^{i _2}({X},K_ 2)$ ,we have $$\l a mb da_2{\math cho ice{\mathop{ \cd ot}\li mi ts_ { A_2}} {\c dot_{ A_2}}{ } {}}\bi g(\lambda _1 { \mathc h oice{\m a t ho p{\c dot}\limits_{A_1} } {\ c dot_{A_1}}{}{} }w\big )= \ bi g ( \la mbd a_2\cdot\l am bda_1 \ big){\m a th c h o ice { \mathop{\cdot }\limits_{A _ 3}} {\cdot _{ A_3 } }{}{}} w$$ i n$ \op eratorname{ W}^{ i_1+i_2+j }_{\!Z } ({Y},L_ 2 )$.
Th e firs t p art fol l ow sfro mL emm a s\[l e mm: groupoid \] a nd \[ lemm : i n v ert\ ].Therestis direct: $$ \be gin{ a rra y}{rl cr}
\ lamb da _2{\m athcho ice{\ ma thop{\cdot}\lim its_ {A_2}}{\c dot _{ A_2 }} {}{}} \ big(\l amb da_ 1{\math choice{ \ mat ho p { \ cd ot}\limits_{A_1}}{ \c d o t_ {A_1}}{} {}}w\b i g)
& = {{A_2} {} ^{{ \cir c l earro wlef t }} }\big(\p i_{Y}^ * \l am bda_2\c do t{{A_1 }{ }^{ {\c ircle a rrow left}} }(\pi_{Y }^*\l a mbda_1\cdot w) \ big)
& \text{ b yd e fi n itio n}\\
& = {{A_ 2}{} ^ {{\c ircl e ar row l eft}} }\cir c{ {( 1 \otimes A_1)}{}^{{\ ci rclear rowle ft}}}\big(\pi _{Y}^*\lam b d a _2\cdot\ pi_{ Y }^ * \lambda_1\cdot w\bi g)
& \text { by Propo sitio n~\ref{p rop:al-co m p }\,\eqre f{i t:a l-t ens } }
\ \
& = {{A_3}{ } ^ {{\c ir clearro wle ft}}}(\ pi_ {Y} ^*\ lam bd a_2\cdot\ pi_{Y}^* \l am bd a_ 1\c dot w )
& \text {b y~\ eq ref {eq:a s soc} a nd Pr op.~ \r ef { pro p:al-co m p} } \\
& = { {A_3 }{} ^{ {\cir clea r row left | A_2}
\\
& L_2
}}$$_Then, for_every $w\in\operatorname{W}^{j}_{\!Z}({Y},L_0)$, $\lambda_1\in\operatorname{W}^{i_1}({X},K_1)$ and_$\lambda_2\in\operatorname{W}^{i_2}({X},K_2)$, we_have_$$\lambda_2{\mathchoice{\mathop{\cdot}\limits_{A_2}}{\cdot_{A_2}}{}{}}\big(\lambda_1{\mathchoice{\mathop{\cdot}\limits_{A_1}}{\cdot_{A_1}}{}{}}w\big)=\big(\lambda_2\cdot\lambda_1\big){\mathchoice{\mathop{\cdot}\limits_{A_3}}{\cdot_{A_3}}{}{}}w$$ in_$\operatorname{W}^{i_1+i_2+j}_{\!Z}({Y},L_2)$.
The_first part follows_from Lemmas \[lemm:groupoid\] and \[lemm:invert\]._The rest is direct:_$$\begin{array}{rlcr}
\lambda_2{\mathchoice{\mathop{\cdot}\limits_{A_2}}{\cdot_{A_2}}{}{}}\big(\lambda_1{\mathchoice{\mathop{\cdot}\limits_{A_1}}{\cdot_{A_1}}{}{}}w\big)
& = {{A_2}{}^{{\circlearrowleft}}}\big(\pi_{Y}^*\lambda_2\cdot{{A_1}{}^{{\circlearrowleft}}}(\pi_{Y}^*\lambda_1\cdot_w)\big)
&_\text{by definition}
\\
& = {{A_2}{}^{{\circlearrowleft}}}\circ {{(1\otimes A_1)}{}^{{\circlearrowleft}}}\big(\pi_{Y}^*\lambda_2\cdot\pi_{Y}^*\lambda_1\cdot w\big)
& \text{by Proposition~\ref{prop:al-comp}\,\eqref{it:al-tens}}
\\
& = {{A_3}{}^{{\circlearrowleft}}}(\pi_{Y}^*\lambda_2\cdot\pi_{Y}^*\lambda_1\cdot w)
& \text{by~\eqref{eq:assoc} and_Prop.~\ref{prop:al-comp}}
\\
&_= {{A_3}{}^{{\circlearrowleft |
R(\m) = \ell + 1$. Besides, $\alpha < 2 a_i$ for each $1 \le i \le \ell$. Let us write $\alpha = \alpha_1 a_1 + \alpha_2 a_2 + \cdots + \alpha_{\ell}a_{\ell}$ where $\alpha_i \ge 0$. Then one of the $\{\alpha_i\}_{1 \le i \le \ell}$ is positive. Therefore, $\alpha = \alpha_ia_i$ for some $1 \le i \le \ell$, so that $\alpha_i = 1$ and $\alpha = a_i$. This makes a contradiction. Hence $R$ doesn’t have minimal multiplicity.
Consequently, we have the following.
\[4.14\] Suppose that $R_1$ is an $\AGL$ ring, but not a Gorenstein ring. Then the following assertions hold true.
1. Let $I \in {\mathcal{X}}_R$. Then either $\mu_R(I) = 2$ or $I = {\mathfrak{c}}$.
2. The set of two-generated Ulrich ideals which are generated by monomials in $t$ is $$\left\{ (t^{2m}, t^{\alpha}) \mid 0<m \in H_1,\ \alpha - m \in H_1,\ 2(\alpha - 2m) \in H\right\}.$$
\(1) Thanks to Proposition \[Ulrich\] (2), if $\mu_R(I) \ge 3$, then ${\mathfrak{c}}\subseteq I$. Since $R$ is a $2$-$\AGL$ ring and $\m \notin {\mathcal{X}}_R$, we conclude that $I = {\mathfrak{c}}$.
\(2) Let $I \in {\mathcal{X}}_R$ such that $\mu_R(I) =2$ and $I$ is generated by monomials in $t$. We write $I = (t^p, t^q)$ where $0 < p < q$ and $p, q \in H$. Notice that, for each $0 < h \in H$ with $h \ne \alpha$, we have that $t^h \in {\mathfrak{c}}$. Since $I + {\mathfrak{c}}= {\mathfrak{m}}$ by Theorem \[2. | R(\m) = \ell + 1$. Besides, $ \alpha < 2 a_i$ for each $ 1 \le i \le \ell$. Let us write $ \alpha = \alpha_1 a_1 + \alpha_2 a_2 + \cdots + \alpha_{\ell}a_{\ell}$ where $ \alpha_i \ge 0$. Then one of the $ \{\alpha_i\}_{1 \le i \le \ell}$ is positive. Therefore, $ \alpha = \alpha_ia_i$ for some $ 1 \le i \le \ell$, so that $ \alpha_i = 1 $ and $ \alpha = a_i$. This make a contradiction. therefore $ R$ does n’t have minimal multiplicity.
Consequently, we get the pursuit.
\[4.14\ ] Suppose that $ R_1 $ is an $ \AGL$ ring, but not a Gorenstein ring. Then the surveil assertions hold truthful.
1. Let $ I \in { \mathcal{X}}_R$. Then either $ \mu_R(I) = 2 $ or $ I = { \mathfrak{c}}$.
2. The set of two - beget Ulrich ideals which are generated by monomials in $ t$ is $ $ \left\ { (t^{2 m }, t^{\alpha }) \mid 0 < m \in H_1,\ \alpha - m \in H_1,\ 2(\alpha - 2 thousand) \in H\right\}.$$
\(1) Thanks to Proposition \[Ulrich\ ] (2), if $ \mu_R(I) \ge 3 $, then $ { \mathfrak{c}}\subseteq I$. Since $ R$ is a $ 2$-$\AGL$ ring and $ \m \notin { \mathcal{X}}_R$, we conclude that $ I = { \mathfrak{c}}$.
\(2) Let $ I \in { \mathcal{X}}_R$ such that $ \mu_R(I) = 2 $ and $ I$ is generated by monomials in $ t$. We write $ I = (t^p, t^q)$ where $ 0 < p < q$ and $ phosphorus, q \in H$. Notice that, for each $ 0 < h \in H$ with $ h \ne \alpha$, we have that $ t^h \in { \mathfrak{c}}$. Since $ I + { \mathfrak{c}}= { \mathfrak{m}}$ by Theorem \[2. | R(\m) = \ell + 1$. Besides, $\alpha < 2 a_l$ for each $1 \le i \le \ell$. Let ua write $\xlpha = \alpha_1 a_1 + \alpha_2 a_2 + \cdovs + \qlpha_{\tjl}a_{\ell}$ where $\alpha_k \ge 0$. Thej one of the $\{\qlpha_i\}_{1 \le m \le \ell}$ is poaltive. Vherefore, $\alpha = \alpha_ia_i$ for some $1 \le h \ue \ell$, so that $\alpha_i = 1$ and $\alpha = a_i$. This mskfs a contradicjion. Nqnce $G$ boesn’t have minimal multiplicitg.
Conseqlently, we have thr following.
\[4.14\] Suppose that $R_1$ is wn $\AGL$ ring, but noh a Gorenstgjn wung. Then the following assertions gold true.
1. Let $I \in {\mathcal{X}}_R$. Tfen enther $\mu_R(I) = 2$ ir $L = {\mathfrak{c}}$.
2. The set of two-gcmerateg Ulricn ideals which arx geberated by monomials mn $t$ is $$\left\{ (t^{2m}, t^{\alpra}) \mid 0<m \hn H_1,\ \alpha - m \in H_1,\ 2(\qlpha - 2k) \in H\rieyt\}.$$
\(1) Ghahkx fo Prooosmtion \[Ulricg\] (2), if $\mu_R(I) \te 3$, then ${\mathfrak{c}}\sibfvyeq I$. Since $D$ is a $2$-$\ADL$ ring and $\m \notin {\mathcal{X}}_R$, we concluge fhat $I = {\mathfrak{c}}$.
\(2) Let $I \in {\mathcal{X}}_R$ such thwt $\mu_R(I) =2$ wnd $I$ is generated by monomials in $t$. We write $I = (d^p, t^q)$ dhexc $0 < p < q$ and $p, q \in H$. Notice that, for each $0 < h \in R$ wotm $h \ne \alpha$, we mave that $t^h \in {\mayhvrsh{c}}$. Since $I + {\mxthfrak{c}}= {\jathfrak{m}}$ by Theorfm \[2. | R(\m) = \ell + 1$. Besides, $\alpha a_i$ each $1 i \le \ell$. \alpha_1 + \alpha_2 a_2 \cdots + \alpha_{\ell}a_{\ell}$ $\alpha_i \ge 0$. Then one of $\{\alpha_i\}_{1 \le i \le \ell}$ is positive. Therefore, $\alpha = \alpha_ia_i$ for some \le i \le \ell$, so that $\alpha_i = 1$ and $\alpha = a_i$. makes contradiction. $R$ have minimal multiplicity. Consequently, we have the following. \[4.14\] Suppose that $R_1$ is an $\AGL$ ring, not a Gorenstein ring. Then the following assertions true. 1. Let $I {\mathcal{X}}_R$. Then either $\mu_R(I) = or = {\mathfrak{c}}$. The of Ulrich ideals which generated by monomials in $t$ is $$\left\{ (t^{2m}, t^{\alpha}) \mid 0<m \in H_1,\ \alpha - m \in 2(\alpha - H\right\}.$$ \(1) to \[Ulrich\] if $\mu_R(I) \ge ${\mathfrak{c}}\subseteq I$. Since $R$ is a $\m \notin {\mathcal{X}}_R$, we conclude that $I = \(2) Let \in {\mathcal{X}}_R$ such that $\mu_R(I) =2$ $I$ is generated by monomials in $t$. We $I = (t^p, t^q)$ where $0 < p < q$ and $p, q \in H$. for each $0 < \in H$ with \ne we that \in {\mathfrak{c}}$. $I + {\mathfrak{c}}= {\mathfrak{m}}$ by Theorem \[2. | R(\m) = \ell + 1$. Besides, $\alpha < 2 a_i$ for eacH $1 \le i \le \ell$. LEt us wRitE $\alPhA = \alpHa_1 a_1 + \aLpha_2 a_2 + \cdots + \alphA_{\Ell}a_{\Ell}$ where $\alpha_i \ge 0$. Then onE of thE $\{\aLPha_i\}_{1 \LE i \Le \ell}$ Is positIVe. tHEreFoRe, $\AlpHa = \ALpHa_ia_i$ For Some $1 \le i \Le \ell$, so thaT $\alPhA_i = 1$ and $\alpha = a_i$. tHiS makes a conTraDiction. Hence $r$ doEsn’t haVe MinIMal muLtiPliciTy.
ConsEQuentlY, we have thE fOLlowinG.
\[4.14\] suppose THAt $r_1$ is aN $\AGL$ ring, but not a GoREnSTein ring. Then thE folloWiNG aSSErtIonS hold true.
1. LEt $i \in {\maTHcal{X}}_R$. THEn EITHer $\MU_R(I) = 2$ or $I = {\mathfraK{c}}$.
2. The set of tWO-geNerateD ULriCH idealS whicH aRE geNerated by moNomiAls in $t$ is $$\lEft\{ (t^{2m}, t^{\ALpha}) \mid 0<M \In H_1,\ \alphA - m \in H_1,\ 2(\aLphA - 2m) \iN H\riGHt\}.$$
\(1) thAnkS tO proPOsItiON \[UlRich\] (2), if $\mu_r(I) \Ge 3$, Then ${\mAthfRAK{C}}\SubsEteQ I$. SiNce $R$ iS a $2$-$\AGL$ ring and $\m \NotIn {\maTHcaL{X}}_R$, we ConclUde tHaT $I = {\matHfrak{c}}$.
\(2) let $I \iN {\mAthcal{X}}_R$ such thaT $\mu_R(i) =2$ and $I$ is geNerAtEd bY mOnomiALs in $t$. WE wrIte $i = (t^p, t^q)$ whEre $0 < p < q$ anD $P, q \iN H$. nOTIcE that, for each $0 < h \in H$ wiTh $H \NE \aLpha$, we haVe that $T^H \iN {\mAThfrak{c}}$. SInCe $I + {\MathFRAk{c}}= {\maThfrAK{m}}$ By TheoreM \[2. | R(\m) = \ell + 1$. Besides , $\alpha< 2 a _i$ fo reach $1\le i \le \ell $ . Le t us write $\alpha = \ alpha _1 a_1+ \ alpha _2 a_2+ \ c d ots + \ alp ha _ {\ ell}a _{\ ell}$ w here $\alp ha_ i\ge 0$. Then on e of the $ \{\ alpha_i\}_{1 \l e i \l e\el l }$ is po sitiv e. The r efore, $\alpha=\ alpha_ i a_i$ fo r so me $ 1 \le i \le \ell$ , s o that $\alpha_ i = 1$ a n d$ \ alp ha= a_i$. Th is make s a cont r ad i c t ion . Hence $R$ do esn’t havem ini mal mu lt ipl i city.
Cons eq u ent ly, we have the followin g.
\[ 4 .14\] S u ppose t hat $R _1$ is an$ \A GL $ r in g , b u tnot a G orenstei nri ng. T hent h e foll owi ng a ssert ions hold tru e.
1.Let $I \ in {\ math ca l{X}} _R$. T hen e it her $\mu_R(I) = 2$or $I = { \ma th fra k{ c}}$.
2. T heset of two -genera t edUl r i c hideals which are g en e r at ed by mo nomial s i n$ t$ is $$ \l eft \{ ( t ^ {2m}, t^{ \ al pha}) \m id 0<m \i nH_1,\ \ al pha -m\in H_ 1,\ 2 ( \alp ha - 2 m) \in H \righ t \}.$$
\(1) Th a nks to Propos i ti o n \ [ Ulri ch\ ] (2), if $ \mu_ R (I)\ge3 $, th e n ${\ mathf ra k {c } }\subseteq I$. Sinc e$R$ is a $2 $-$\AGL$ ring and $\m \ n o t in {\mat hcal { X} } _R$, we conclu de th at $I = {\ m athfrak{ c}}$.
\(2) L et $I \in { \mathcal {X} }_R $ s uch t ha t $\mu_R(I) = 2 $ and $ I$ is g ene rated b y m ono mia lsin $t$. Wewrite $I = ( t^ p, t^ q)$ w h ere $0 < p <q$ an d $p, q \inH$. N otic eth a t,for eac h $ 0 < h\i nH$ w ith $ h \ne \al p ha$ , we ha ve that $ t^h \in{\ ma thfrak{ c}}$. Since $ I+ {\mathfr ak {c} }= {\m a t hfrak{m} }$ by Theorem \[2. | R(\m) =_\ell +_1$. Besides, $\alpha <_2 a_i$_for_each $1_\le_i \le \ell$._Let us write_$\alpha = \alpha_1 a_1_+ \alpha_2 a_2_+_\cdots + \alpha_{\ell}a_{\ell}$ where $\alpha_i \ge 0$. Then one of the $\{\alpha_i\}_{1 \le i_\le_\ell}$ is_positive._Therefore,_$\alpha = \alpha_ia_i$ for some_$1 \le i \le \ell$,_so that_$\alpha_i = 1$ and $\alpha = a_i$. This_makes_a contradiction. Hence_$R$ doesn’t have minimal multiplicity.
Consequently, we have the following.
\[4.14\]_Suppose that $R_1$ is an $\AGL$_ring, but not_a_Gorenstein_ring. Then the following_assertions hold true.
1. Let $I_\in {\mathcal{X}}_R$. Then either $\mu_R(I) =_2$ or $I = {\mathfrak{c}}$.
2. The_set of two-generated Ulrich ideals which_are generated by monomials in_$t$ is_$$\left\{ (t^{2m}, t^{\alpha}) \mid 0<m_\in H_1,\ \alpha_- m_\in H_1,\ 2(\alpha_- 2m) \in H\right\}.$$
\(1) Thanks to_Proposition \[Ulrich\] (2),_if $\mu_R(I) \ge 3$, then ${\mathfrak{c}}\subseteq_I$._Since $R$ is_a_$2$-$\AGL$_ring and_$\m \notin {\mathcal{X}}_R$,_we_conclude that_$I_= {\mathfrak{c}}$.
\(2) Let $I \in {\mathcal{X}}_R$_such_that $\mu_R(I) =2$ and $I$ is generated_by monomials in $t$._We_write $I = (t^p,_t^q)$ where $0 < p_< q$ and $p, q \in_H$. Notice_that, for_each $0 < h \in H$ with $h \ne \alpha$, we_have that $t^h \in {\mathfrak{c}}$. Since_$I + {\mathfrak{c}}= {\mathfrak{m}}$_by Theorem_\[2. |
"}](fig/dEdt8H_q090_expf_logx.eps "fig:"){width="69mm"}![Same as Fig. \[fig:flux\_taylor\_1\] but using exponential resummation to the energy flux in the post-Newtonian approximation. The relative error for 11PN is less than $10^{-5}$ when $v\lessapprox 0.4$, whose region is larger than $v\lessapprox 0.33$ for the Taylor expanded energy flux in Fig. \[fig:flux\_taylor\_1\]. []{data-label="fig:flux_exponentioal_1"}](fig/dEdt8H_q0998_expf_logx.eps "fig:"){width="69mm"}
![Same as Fig. \[fig:flux\_taylor\_2\] but using exponential resummation to the energy flux in the post-Newtonian approximation. The relative error for 11PN is less than $10^{-5}$ when $v\lessapprox 0.4$, whose region is larger than $v\lessapprox 0.33$ for the Taylor expanded energy flux in Fig. \[fig:flux\_taylor\_2\]. []{data-label="fig:flux_exponentioal_2"}](fig/dEdt8H_q-001_expf_logx.eps "fig:"){width="69mm"}![Same as Fig. \[fig:flux\_taylor\_2\] but using exponential resummation to the energy flux in the post-Newtonian approximation. The relative error for 11PN is less than $10^{-5}$ when $v\lessapprox 0.4$, whose region is larger than $v\lessapprox 0.33$ for the Taylor expanded energy flux in Fig. \[fig:flux\_taylor\_2\]. []{data-label="fig:flux_exponentioal_2"}](fig/dEdt8H_q-005_expf_logx.eps "fig:"){width="69mm"}\
![Same as Fig. \[fig:flux\_taylor\_2\] but using exponential resummation to the energy flux in the post-Newtonian approximation. The relative error for 11PN is less than $10^{-5}$ when $v\lessapprox 0.4$, whose region is larger than $v\lessapprox 0.33$ for the Taylor expanded | " } ] (fig / dEdt8H_q090_expf_logx.eps " fig:"){width="69mm"}![Same as Fig. \[fig: flux\_taylor\_1\ ] but using exponential resummation to the energy flux in the post - Newtonian approximation. The proportional mistake for 11PN is less than $ 10^{-5}$ when $ v\lessapprox 0.4 $, whose region is larger than $ v\lessapprox 0.33 $ for the Taylor expand department of energy flux in Fig. \[fig: flux\_taylor\_1\ ]. [ ] { data - label="fig: flux_exponentioal_1"}](fig / dEdt8H_q0998_expf_logx.eps " fig:"){width="69 mm " }
! [ like as Fig. \[fig: flux\_taylor\_2\ ] but use exponential resummation to the energy magnetic field in the post - Newtonian approximation. The relative erroneousness for 11PN is less than $ 10^{-5}$ when $ v\lessapprox 0.4 $, whose region is larger than $ v\lessapprox 0.33 $ for the Taylor elaborate department of energy flux in Fig. \[fig: flux\_taylor\_2\ ]. [ ] { data - label="fig: flux_exponentioal_2"}](fig / dEdt8H_q-001_expf_logx.eps " fig:"){width="69mm"}![Same as Fig. \[fig: flux\_taylor\_2\ ] but use exponential resummation to the energy flux in the post - Newtonian estimate. The relative error for 11PN is less than $ 10^{-5}$ when $ v\lessapprox 0.4 $, whose region is large than $ v\lessapprox 0.33 $ for the Taylor expanded energy magnetic field in Fig. \[fig: flux\_taylor\_2\ ]. [ ] { data - label="fig: flux_exponentioal_2"}](fig / dEdt8H_q-005_expf_logx.eps " fig:"){width="69mm"}\
! [ like as Fig. \[fig: flux\_taylor\_2\ ] but using exponential resummation to the department of energy flux in the post - Newtonian approximation. The relative error for 11PN is less than $ 10^{-5}$ when $ v\lessapprox 0.4 $, whose area is larger than $ v\lessapprox 0.33 $ for the Taylor expanded | "}](fig/fEdt8H_q090_expf_logx.eps "fig:"){widuh="69mm"}![Same as Fig. \[fiy:dlux\_tablor\_1\] buf using dxponential resummation to tie ebergy flux in the post-Newtovian apprlximatiob. Tht relative error hkr 11PN is less bhan $10^{-5}$ xhen $v\lessapprow 0.4$, whose recion is larger tfau $v\lessapprox 0.33$ for the Taylor expandqd enerby flux in Fig. \[fid:fluq\_twylod\_1\]. []{data-label="fig:flux_exponentioal_1"}](fig/dSdt8H_q0998_exkf_logx.eps "fig:"){width="69km"}
![Same as Fig. \[fig:flux\_taylor\_2\] but using exponential resummatiob to rhe energy fuux in the post-Newtonizn approximation. The relative efror yor 11PN is lgsa hvan $10^{-5}$ when $v\oessaiprox 0.4$, whose vvgion iv larget than $v\lessapirox 0.33$ foe the Taylor expanded energy flux in Fid. \[fig:flux\_tdymor\_2\]. []{data-label="fig:fouz_expotenthoal_2"}](wug/dDdt8G_q-001_xxpr_logx.eos "hig:"){width="69mm"}![Szme as Fig. \[fug:flux\_taylor\_2\] but usond exponential desummwtyon to the energy flux in the post-Newtotiah approximation. The relqtive error for 11PN is less thag $10^{-5}$ when $v\lessapprox 0.4$, whose region is larger than $e\lesseporor 0.33$ for rhf Taylor expanded energy flux in Fig. \[fig:flux\_tarmot\_2\]. []{cata-label="fig:flmx_exponentioal_2"}](fig/dRdh8H_a-005_gxpf_logx.eps "fie:"){width="69mm"}\
![Szme as Fig. \[fig:flux\_twylor\_2\] bot usibg exponegtiak resummation to the energy flux in the post-Newtonian apprlximation. Tke relstive error for 11PN is less tkan $10^{-5}$ wgen $v\lessaporox 0.4$, whoad region is largdr nhan $v\lessapkfox 0.33$ for the Tayljr expandxd | "}](fig/dEdt8H_q090_expf_logx.eps "fig:"){width="69mm"}![Same as Fig. \[fig:flux\_taylor\_1\] but using to energy flux the post-Newtonian approximation. is than $10^{-5}$ when 0.4$, whose region larger than $v\lessapprox 0.33$ for the expanded energy flux in Fig. \[fig:flux\_taylor\_1\]. []{data-label="fig:flux_exponentioal_1"}](fig/dEdt8H_q0998_expf_logx.eps "fig:"){width="69mm"} ![Same as Fig. \[fig:flux\_taylor\_2\] but exponential resummation to the energy flux in the post-Newtonian approximation. The relative error 11PN less $10^{-5}$ $v\lessapprox 0.4$, whose region is larger than $v\lessapprox 0.33$ for the Taylor expanded energy flux in \[fig:flux\_taylor\_2\]. []{data-label="fig:flux_exponentioal_2"}](fig/dEdt8H_q-001_expf_logx.eps "fig:"){width="69mm"}![Same as Fig. \[fig:flux\_taylor\_2\] but using resummation to the energy in the post-Newtonian approximation. The error 11PN is than when 0.4$, whose region larger than $v\lessapprox 0.33$ for the Taylor expanded energy flux in Fig. \[fig:flux\_taylor\_2\]. []{data-label="fig:flux_exponentioal_2"}](fig/dEdt8H_q-005_expf_logx.eps "fig:"){width="69mm"}\ ![Same as \[fig:flux\_taylor\_2\] but resummation to energy in post-Newtonian approximation. The for 11PN is less than $10^{-5}$ whose region is larger than $v\lessapprox 0.33$ for Taylor expanded | "}](fig/dEdt8H_q090_expf_logx.eps "fig:"){wiDth="69mm"}![Same aS Fig. \[fIg:fLux\_TaYlor\_1\] But uSing exponentiaL ResuMmation to the energy flux In the PoST-NewTOnIan apProximaTIoN. tHe rElAtIve ErROr For 11PN Is lEss than $10^{-5}$ When $v\lessaPprOx 0.4$, Whose region iS LaRger than $v\lEssApprox 0.33$ for the tayLor expAnDed ENergy FluX in FiG. \[fig:flUX\_tayloR\_1\]. []{data-labeL="fIG:flux_eXPonentiOAL_1"}](fIg/dEDt8H_q0998_expf_logx.eps "fiG:"){WiDTh="69mm"}
![Same as Fig. \[fIg:flux\_TaYLoR\_2\] BUt uSinG exponentiAl ResumMAtion to THe ENERgy FLux in the post-NEwtonian appROxiMation. thE reLAtive eRror fOr 11pn is Less than $10^{-5}$ wheN $v\leSsapprox 0.4$, wHose reGIon is laRGer than $V\lessaPprOx 0.33$ fOr thE taYlOr eXpANdeD EnErgY FluX in Fig. \[fiG:fLuX\_taylOr\_2\]. []{daTA-LABel="fIg:fLux_eXponeNtioal_2"}](fig/dEdt8h_q-001_eXpf_lOGx.ePs "fig:"){Width="69Mm"}![SaMe As Fig. \[Fig:fluX\_taylOr\_2\] But using exponenTial ResummatiOn tO tHe eNeRgy flUX in the PosT-NeWtonian ApproxiMAtiOn. tHE ReLative error for 11PN is LeSS ThAn $10^{-5}$ when $v\lEssappROx 0.4$, WhOSe region Is LarGer tHAN $v\lesSappROx 0.33$ For the TaYlor exPAnDeD energy FlUx in FiG. \[fIg:fLux\_TayloR\_2\]. []{Data-Label="fIg:flux_exPonenTIoal_2"}](fig/dEdt8H_q-005_eXPf_logx.eps "fig:"){wIDtH="69MM"}\
![SAMe as fig. \[Fig:flux\_taylOr\_2\] buT UsinG expONeNtiAL resuMmatiOn TO tHE energy flux in the posT-NEwtoniAn appRoximation. The Relative erROR For 11PN is lEss tHAn $10^{-5}$ WHen $v\lessapprox 0.4$, Whose Region is laRGer than $v\LessaPprox 0.33$ for The Taylor EXPanded | "}](fig/dEdt8H_q090_expf_l ogx.eps "f ig:") {wi dth =" 69mm "}![ Same as Fig. \ [ fig: flux\_taylor\_1\] butusing e x pone n ti al re summati o nt o th een erg yf lu x inthe post-N ewtonian a ppr ox imation. The re lative err orfor 11PN isles s than $ 10^ { -5}$whe n $v\ lessap p rox 0. 4$, whose r e gion i s larger t ha n $v \lessapprox 0.33$ fo r the Taylor ex panded e n er g y fl uxin Fig. \[ fi g:flu x \_taylo r \_ 1 \ ] . [ ] {data-label=" fig:flux_ex p one ntioal _1 "}] ( fig/dE dt8H_ q0 9 98_ expf_logx.e ps " fig:"){wi dth="6 9 mm"}
! [ Same as Fig.\[f ig: flux \ _t ay lor \_ 2 \]b ut us i ngexponent ia lresum mati o n t o th e e nerg y flu x in the post -Ne wton i anappro ximat ion. T he re lative erro rfor 11PN is les s th an $10^{- 5}$ w hen $ v\les s approx 0. 4$, whoseregioni s l ar g e r t han $v\lessapprox0. 3 3 $for theTaylor ex pa n ded ener gy fl ux i n Fig.\[fi g :f lux\_tay lor\_2 \ ]. [ ]{data- la bel="f ig :fl ux_ expon e ntio al_2"} ](fig/dE dt8H_ q -001_expf_logx . eps "fig:"){w i dt h = "6 9 mm"} ![S ame as Fig. \[f i g:fl ux\_ t ay lor \ _2\]but u si n ge xponential resummat io n to t he en ergy flux inthe post-N e w t onian ap prox i ma t ion. The relat ive e rror for 1 1 PN is le ss th an $10^{ -5}$ when $ v\lessap pro x 0 .4$ , w h o se region is la r g er t ha n $v\le ssa pprox 0 .33 $ f orthe T aylor exp anded en er gy f lu x i n Fig . \[fig:f lu x\_ ta ylo r\_2\ ] . []{d ata-l abel =" fi g :fl ux_expo n en t i oal_ 2" }] (fig /dE dt 8H_q- 005_ e xpf _logx.e ps "fig:" ){w i dth= "6 9m m"}\
![ Same as Fig.\[ fig:flux\_ ta ylo r\_2\] b ut using exponential resummatio n to the en ergyflux in the p ost -Newto nia n appro ximati on. T he re l a tivee r ro r f or 11PN is l e s s t han $ 10 ^{-5 }$ when $v\lessapprox 0.4 $ , w hose region i s l arge r th an$ v\ l ess ap p rox 0 .33$ for the Ta ylor expan de d | "}](fig/dEdt8H_q090_expf_logx.eps "fig:"){width="69mm"}![Same_as Fig. \[fig:flux\_taylor\_1\]_but using exponential resummation_to the_energy_flux in_the_post-Newtonian approximation. The_relative error for_11PN is less than_$10^{-5}$ when $v\lessapprox_0.4$,_whose region is larger than $v\lessapprox 0.33$ for the Taylor expanded energy flux in_Fig. \[fig:flux\_taylor\_1\]._[]{data-label="fig:flux_exponentioal_1"}](fig/dEdt8H_q0998_expf_logx.eps "fig:"){width="69mm"}
![Same_as_Fig. \[fig:flux\_taylor\_2\]_but using exponential resummation to_the energy flux in the_post-Newtonian approximation._The relative error for 11PN is less than_$10^{-5}$_when $v\lessapprox 0.4$,_whose region is larger than $v\lessapprox 0.33$ for the_Taylor expanded energy flux in Fig. \[fig:flux\_taylor\_2\]._[]{data-label="fig:flux_exponentioal_2"}](fig/dEdt8H_q-001_expf_logx.eps "fig:"){width="69mm"}![Same as_Fig. \[fig:flux\_taylor\_2\]_but_using exponential resummation to_the energy flux in the post-Newtonian_approximation. The relative error for 11PN_is less than $10^{-5}$ when $v\lessapprox 0.4$,_whose region is larger than $v\lessapprox_0.33$ for the Taylor expanded_energy flux_in Fig. \[fig:flux\_taylor\_2\]. []{data-label="fig:flux_exponentioal_2"}](fig/dEdt8H_q-005_expf_logx.eps "fig:"){width="69mm"}\
![Same as_Fig. \[fig:flux\_taylor\_2\] but using_exponential resummation_to the energy_flux in the post-Newtonian approximation. The_relative error for_11PN is less than $10^{-5}$ when_$v\lessapprox_0.4$, whose region_is_larger_than $v\lessapprox_0.33$ for the_Taylor_expanded |
arg \min_{s\in m}\left\{ P_{n}\left( \gamma \left(
s\right) \right) \right\} \\
& =\arg \min_{s\in m}\left\{ \frac{1}{n}\sum_{i=1}^{n}\left( Y_{i}-s\left(
X_{i}\right) \right) ^{2}\right\},\end{aligned}$$where $P_{n}=n^{-1}\sum_{i=1}^{n}\delta _{\xi _{i}}$ is the *empirical measure* built from the data.
The performance of the least-squares estimators is tackled through their *excess loss*,$$\ell \left( s_{\ast },\widehat{s}_{m}\right) :=P\left( \gamma \left( \widehat{s}_{m}\right) -\gamma \left( s_{\ast }\right) \right) =\left\Vert \widehat{s}_{m}-s_{\ast }\right\Vert _{2}^{2}.$$We split the excess risk into a sum of two terms,$$\ell \left( s_{\ast },\widehat{s}_{m}\right) =\ell \left( s_{\ast },s_{m}\right)
+\ell \left( s_{m},\widehat{s}_{m}\right),$$where$$\ell \left( s_{\ast },s_{m}\right) :=P\left( \gamma \left( s_{m}\right)
-\gamma \left( s_{\ast }\right) \right) =\left\Vert s_{m}-s_{\ast
}\right\Vert _{2}^{2}\text{ \ \ \ and \ \ \ }\ell \left( s_{m},\widehat{s}_{m}\right) :=P\left( \gamma \left( \widehat{s}_{m}\right) -\gamma \left(
s_{m}\right) \right) \geq 0.$$The quantity $\ell \left( s_{\ast },s_{m}\right) $ is a deterministic term called the *bias* of the model $m$, while $\ell \left( s_{m},\widehat{s}_{m}\right) $ is a random variable that we call the *excess risk* of the least-squares estimator $\widehat{s}_{m}$ *on the model* $m$. Notice that by the | arg \min_{s\in m}\left\ { P_{n}\left (\gamma \left (
s\right) \right) \right\ } \\
& = \arg \min_{s\in m}\left\ { \frac{1}{n}\sum_{i=1}^{n}\left (Y_{i}-s\left (
X_{i}\right) \right) ^{2}\right\},\end{aligned}$$where $ P_{n}=n^{-1}\sum_{i=1}^{n}\delta _ { \xi _ { i}}$ is the * empirical measure * built from the data.
The operation of the least - public square estimators is tackled through their * excess loss*,$$\ell \left (s_{\ast }, \widehat{s}_{m}\right): = P\left (\gamma \left (\widehat{s}_{m}\right) -\gamma \left (s_{\ast } \right) \right) = \left\Vert \widehat{s}_{m}-s_{\ast } \right\Vert _ { 2}^{2}.$$We separate the excess hazard into a sum of two terms,$$\ell \left (s_{\ast }, \widehat{s}_{m}\right) = \ell \left (s_{\ast }, s_{m}\right)
+ \ell \left (s_{m},\widehat{s}_{m}\right),$$where$$\ell \left (s_{\ast }, s_{m}\right): = P\left (\gamma \left (s_{m}\right)
-\gamma \left (s_{\ast } \right) \right) = \left\Vert s_{m}-s_{\ast
} \right\Vert _ { 2}^{2}\text { \ \ \ and \ \ \ } \ell \left (s_{m},\widehat{s}_{m}\right): = P\left (\gamma \left (\widehat{s}_{m}\right) -\gamma \left (
s_{m}\right) \right) \geq 0.$$The quantity $ \ell \left (s_{\ast }, s_{m}\right) $ is a deterministic term call the * bias * of the model $ m$, while $ \ell \left (s_{m},\widehat{s}_{m}\right) $ is a random variable star that we visit the * excess risk * of the least - square estimator $ \widehat{s}_{m}$ * on the model * $ m$. Notice that by the | arg \min_{s\in m}\left\{ P_{n}\left( \gamoa \left(
s\right) \rntht) \rijht\} \\
& =\arf \min_{s\in m}\left\{ \frac{1}{n}\sum_{i=1}^{n}\left( Y_{i}-s\lefv(
X_{i}\rught) \eight) ^{2}\right\},\end{aligned}$$wfere $P_{n}=n^{-1}\slm_{i=1}^{n}\delta _{\xi _{m}}$ is the *empiricem measuvz* buimb frok the data.
The pgrformance ox the least-squdrds estimators is tackled through theiw *excesx poss*,$$\ell \left( s_{\wst },\eydehzn{s}_{n}\right) :=P\left( \gamma \left( \widehzt{s}_{m}\rigit) -\gamma \left( s_{\sst }\right) \right) =\left\Vert \wldehwt{s}_{m}-s_{\ast }\right\Vert _{2}^{2}.$$We split tye evxess risk ingo a sum of two terms,$$\eml \left( s_{\ast },\widehat{s}_{m}\right) =\ell \left( s_{\ast },s_{m}\rithr)
+\elp \left( s_{m},\widxhat{s}_{m}\gight),$$where$$\ell \left( s_{\avt },s_{m}\ribht) :=P\left( \gamms \lxft( w_{m}\right)
-\gamma \left( s_{\asv }\right) \right) =\left\Vett s_{m}-s_{\ast
}\rhgkt\Vert _{2}^{2}\text{ \ \ \ and \ \ \ }\eol \lext( s_{k},\widdyat{r}_{m}\rjgit) :=L\left( \hamja \left( \wisehat{s}_{m}\righr) -\gamma \left(
s_{m}\right) \rytht) \geq 0.$$The qhantitr $\qll \left( s_{\ast },s_{m}\right) $ is a deterministpc tsrm called the *bias* of rhe model $m$, while $\ell \left( s_{m},\wydehat{s}_{m}\right) $ is a random variable that we call dhe *eecdss rlsk* ud hhe least-squares estimator $\widehat{s}_{m}$ *on the mjsek* $k$. Notice that ny the | arg \min_{s\in m}\left\{ P_{n}\left( \gamma \left( s\right) \\ =\arg \min_{s\in \frac{1}{n}\sum_{i=1}^{n}\left( Y_{i}-s\left( X_{i}\right) is *empirical measure* built the data. The of the least-squares estimators is tackled their *excess loss*,$$\ell \left( s_{\ast },\widehat{s}_{m}\right) :=P\left( \gamma \left( \widehat{s}_{m}\right) -\gamma \left( s_{\ast \right) =\left\Vert \widehat{s}_{m}-s_{\ast }\right\Vert _{2}^{2}.$$We split the excess risk into a sum of terms,$$\ell s_{\ast =\ell s_{\ast },s_{m}\right) +\ell \left( s_{m},\widehat{s}_{m}\right),$$where$$\ell \left( s_{\ast },s_{m}\right) :=P\left( \gamma \left( s_{m}\right) -\gamma \left( s_{\ast }\right) =\left\Vert s_{m}-s_{\ast }\right\Vert _{2}^{2}\text{ \ \ \ and \ \ }\ell \left( :=P\left( \gamma \left( \widehat{s}_{m}\right) -\gamma s_{m}\right) \geq 0.$$The $\ell s_{\ast $ is a term called the *bias* of the model $m$, while $\ell \left( s_{m},\widehat{s}_{m}\right) $ is a random variable we call risk* of least-squares $\widehat{s}_{m}$ the model* $m$. by the | arg \min_{s\in m}\left\{ P_{n}\left( \gamma \Left(
s\right) \Right) \RigHt\} \\
& =\aRg \Min_{s\In m}\lEft\{ \frac{1}{n}\sum_{i=1}^{n}\lEFt( Y_{i}-S\left(
X_{i}\right) \right) ^{2}\right\},\End{alIgNEd}$$whERe $p_{n}=n^{-1}\suM_{i=1}^{n}\deltA _{\Xi _{I}}$ IS thE *eMpIriCaL MeAsure* BuiLt from tHe data.
The pErfOrMance of the leASt-Squares estImaTors is tackleD thRough tHeIr *eXCess lOss*,$$\Ell \leFt( s_{\ast },\WIdehat{S}_{m}\right) :=P\lEfT( \Gamma \lEFt( \widehAT{S}_{m}\RighT) -\gamma \left( s_{\ast }\rigHT) \rIGht) =\left\Vert \widEhat{s}_{m}-S_{\aST }\rIGHt\VErt _{2}^{2}.$$we split the ExCess rISk into a SUm OF TWo tERms,$$\ell \left( s_{\asT },\widehat{s}_{m}\rIGht) =\Ell \lefT( s_{\Ast },S_{M}\right)
+\Ell \leFt( S_{M},\wiDehat{s}_{m}\righT),$$wheRe$$\ell \left( S_{\ast },s_{m}\RIght) :=P\leFT( \gamma \lEft( s_{m}\rIghT)
-\gaMma \lEFt( S_{\aSt }\rIgHT) \riGHt) =\LefT\verT s_{m}-s_{\ast
}\rIgHt\vert _{2}^{2}\tExt{ \ \ \ aND \ \ \ }\ELL \lefT( s_{m},\WideHat{s}_{m}\Right) :=P\left( \gamMa \lEft( \wIDehAt{s}_{m}\rIght) -\gAmma \LeFt(
s_{m}\rIght) \riGht) \geQ 0.$$THe quantity $\ell \leFt( s_{\aSt },s_{m}\right) $ Is a DeTerMiNistiC Term caLleD thE *bias* of The modeL $M$, whIlE $\ELL \lEft( s_{m},\widehat{s}_{m}\righT) $ iS A RaNdom variAble thAT wE cALl the *excEsS riSk* of THE leasT-squAReS estimatOr $\wideHAt{S}_{m}$ *On the moDeL* $m$. NotiCe ThaT by The | arg \min_{s\in m}\left\{ P _{n}\left( \gam ma\le ft (
s\ righ t) \right) \ri g ht\} \\
& =\arg \min_{s\in m}\l ef t \{ \ f ra c{1}{ n}\sum_ { i= 1 } ^{n }\ le ft( Y _ {i }-s\l eft (
X_{i} \right) \r igh t) ^{2}\right\ } ,\ end{aligne d}$ $where $P_{n }=n ^{-1}\ su m_{ i =1}^{ n}\ delta _{\xi _{i}}$ is the * em p irical measure * bu iltfrom the data.
T h ep erformance ofthe le as t -s q u are s e stimatorsis tack l ed thro u gh t h eir *excess loss* ,$$\ell \le f t(s_{\as t},\ w idehat {s}_{ m} \ rig ht) :=P\lef t( \ gamma \le ft( \w i dehat{s } _{m}\ri ght) - \ga mma \le f t( s _{\ as t }\ r ig ht) \ri ght) =\l ef t\ Vert\wid e h a t {s}_ {m} -s_{ \ast}\right\Vert_{2 }^{2 } .$$ We sp lit t he e xc ess r isk in to asu m of two terms, $$\e ll \left( s_ {\ ast } ,\wid e hat{s} _{m }\r ight) = \ell \l e ft( s _ { \ as t },s_{m}\right)
+ \e l l \ left( s_ {m},\w i de ha t {s}_{m}\ ri ght ),$$ w h ere$$ \ell \l eft( s_{ \ast } , s_ {m }\right ):=P\le ft ( \ gam ma \l e ft(s_{m}\ right)
- \gamm a \left( s_{\as t }\right) \ri g ht ) =\ l eft\ Ver t s_{m}-s_{ \ast }\ri ght\ V er t _ { 2}^{2 }\tex t{ \\ \ and \ \ \ }\ell\l eft( s _{m}, \widehat{s}_{ m}\right): = P \left( \ gamm a \ l eft( \widehat{ s}_{m }\right) - \ gamma \l eft(s_{m}\ri ght) \rig h t ) \geq 0 .$$ The qu ant i t y$\ell \left(s _ {\as t},s_{m} \ri ght) $isa d ete rmi ni stic term calledth e*b ia s*of th e model $ m$ , w hi le$\ell \left( s_{m },\w id eh a t{s }_{m}\r i gh t ) $ i sarand omva riabl e th a t w e callthe *exce ssr isk* o fthe lea st-squares es ti mator $\wi de hat {s}_{m } $ *on the model* $m$. Notice tha t by the | arg \min_{s\in_m}\left\{ P_{n}\left(_\gamma \left(
s\right) \right) \right\}_\\
& =\arg_\min_{s\in_m}\left\{ \frac{1}{n}\sum_{i=1}^{n}\left(_Y_{i}-s\left(
X_{i}\right)_\right) ^{2}\right\},\end{aligned}$$where $P_{n}=n^{-1}\sum_{i=1}^{n}\delta__{\xi _{i}}$ is_the *empirical measure* built_from the data.
The_performance_of the least-squares estimators is tackled through their *excess loss*,$$\ell \left( s_{\ast },\widehat{s}_{m}\right) :=P\left(_\gamma_\left( \widehat{s}_{m}\right)_-\gamma_\left(_s_{\ast }\right) \right) =\left\Vert \widehat{s}_{m}-s_{\ast_}\right\Vert _{2}^{2}.$$We split the excess_risk into_a sum of two terms,$$\ell \left( s_{\ast },\widehat{s}_{m}\right)_=\ell_\left( s_{\ast },s_{m}\right)
+\ell_\left( s_{m},\widehat{s}_{m}\right),$$where$$\ell \left( s_{\ast },s_{m}\right) :=P\left( \gamma \left( s_{m}\right)
-\gamma_\left( s_{\ast }\right) \right) =\left\Vert s_{m}-s_{\ast
}\right\Vert__{2}^{2}\text{ \ \_\_and_\ \ \ }\ell_\left( s_{m},\widehat{s}_{m}\right) :=P\left( \gamma \left( \widehat{s}_{m}\right)_-\gamma \left(
s_{m}\right) \right) \geq 0.$$The quantity_$\ell \left( s_{\ast },s_{m}\right) $ is a_deterministic term called the *bias* of_the model $m$, while $\ell_\left( s_{m},\widehat{s}_{m}\right)_$ is a random variable_that we call_the *excess_risk* of the_least-squares estimator $\widehat{s}_{m}$ *on the model*_$m$. Notice that_by the |
for us to move forward. This approach for initial setups is easier, and leads to the desired asymptotic measure estimate on $\Delta$ as $L\to\infty$.
We move to the expanding property of the maps corresponding to parameters in $\Delta_{N}$. We frequently use the following notation: for $c\in C$ and $n\geq1$, $c_0=fc$ and $c_n=f^nc_0$: for $x\in S^1$ and $n\geq1$, $J(x)=|f'x|$ and $J^n(x)=J(x)J(fx)\cdots J(f^{n-1}x)$.
Let $\alpha=10^{-6}$ and $\delta = L^{-\alpha N}$. In what follows, we suppose $N$ to be a large integer for which $\delta\ll\sigma$, and the conclusion of Lemma \[initial-1\] holds. The value of $N$ will be replaced if necessary, but only a finite number of times. The letter $K$ will be used to denote generic constants which are independent of $N$ and $L$.
The next lemma, the proof of which is given in Appendix, ensures an exponential growth of derivatives for orbit segments lying outside of $C_{\delta}$.
\[outside\] There exists $L_1\geq L_0$ such that if $L\geq L_1$ and $f = f_{a}$ is such that $a \in \Delta_{N}(L)$, then the following holds:
- if $n\geq1$ and $x$, $fx,\cdots, f^{n-1}x\notin C_{\delta}$, then $J^{n}(x)\geq \delta L^{2\lambda n}$;
- if moreover $f^nx\in C_{\delta}$, then $J^{n}(x) \geq
L^{2 \lambda n}$.
[*Standing assumption for the rest of this section: $L\geq L_1$ and $a \in \Delta_{N}$.*]{}
Recovering expansion {#s2.3}
--------------------
For $f = f_a$, $c \in C$ and $n>N$ we introduce three conditions:
- $J^{j-i}(c_i)\geq L \min | for us to move forward. This approach for initial setups is easy, and go to the desired asymptotic measure estimate on $ \Delta$ as $ L\to\infty$.
We go to the expanding property of the function represent to parameters in $ \Delta_{N}$. We frequently practice the following note: for $ c\in C$ and $ n\geq1 $, $ c_0 = fc$ and $ c_n = f^nc_0 $: for $ x\in S^1 $ and $ n\geq1 $, $ J(x)=|f'x|$ and $ J^n(x)=J(x)J(fx)\cdots J(f^{n-1}x)$.
Let $ \alpha=10^{-6}$ and $ \delta = L^{-\alpha N}$. In what follows, we presuppose $ N$ to be a large integer for which $ \delta\ll\sigma$, and the conclusion of Lemma \[initial-1\ ] holds. The value of $ N$ will be replace if necessary, but only a finite number of times. The letter $ K$ will be used to denote generic constant which are independent of $ N$ and $ L$.
The next lemma, the proof of which is give in Appendix, ensures an exponential growth of derivatives for orbit section lying outside of $ C_{\delta}$.
\[outside\ ] There exists $ L_1\geq L_0 $ such that if $ L\geq L_1 $ and $ f = f_{a}$ is such that $ a \in \Delta_{N}(L)$, then the following holds:
- if $ n\geq1 $ and $ x$, $ fx,\cdots, f^{n-1}x\notin C_{\delta}$, then $ J^{n}(x)\geq \delta L^{2\lambda n}$;
- if furthermore $ f^nx\in C_{\delta}$, then $ J^{n}(x) \geq
L^{2 \lambda n}$.
[ * Standing assumption for the rest of this section: $ L\geq L_1 $ and $ a \in \Delta_{N}$. * ] { }
Recovering expansion { # s2.3 }
--------------------
For $ f = f_a$, $ c \in C$ and $ n > N$ we precede three condition:
- $ J^{j - i}(c_i)\geq L \min | fog us to move forward. Thir approach for nbitial setupa is easker, and leads to the desired awymptitic measure estimate un $\Delta$ ws $L\to\indty$.
Wt move to the expehding pvjperfn of chx maps correspokding to pasameters in $\Dentx_{N}$. We frequently use the following noeation: gog $c\in C$ and $n\ggq1$, $c_0=fb$ wnd $d_n=f^nc_0$: for $x\in S^1$ and $n\geq1$, $J(x)=|f'x|$ and $N^n(x)=J(x)J(fq)\cdots J(f^{n-1}x)$.
Let $\allha=10^{-6}$ and $\delta = L^{-\alpha N}$. In whah follows, we suppode $N$ to be q lawte integer fur which $\dtlca\ll\sigma$, ahd the conclusion of Lemma \[initkal-1\] hplds. The vqlye lx $N$ will be replwced if necessary, bud only s finite numbev of vimew. The letter $K$ will bx used to denote gengric constdncs which are independwnr of $T$ ang $L$.
Tfw ndxt lxmmz, the orokf of whicg is given un Appendix, ensures ag exponential frowth os derivatives for orbit segments lying mutaide of $C_{\delta}$.
\[outside\] Tyere exists $L_1\geq L_0$ sufh that is $L\geq L_1$ and $f = f_{a}$ is such that $a \in \Delta_{N}(L)$, then the houloclkg huodd:
- if $n\geq1$ and $x$, $fx,\cdots, f^{n-1}x\notin C_{\delta}$, theg $J^{m}(x)\neq \delta L^{2\lambdc n}$;
- if moreovet $v^nc\yn C_{\delta}$, thev $J^{n}(x) \ysq
L^{2 \lambda n}$.
[*Stanfing asfumptuon for tre rrst of this section: $L\geq L_1$ qnd $a \in \Delna_{N}$.*]{}
Rwcovering expansiou {#s2.3}
--------------------
For $f = f_a$, $c \im C$ amd $n>N$ we introduce threz condjtions:
- $J^{j-l}(c_i)\geq L \jkn | for us to move forward. This approach setups easier, and to the desired as We move to expanding property of maps corresponding to parameters in $\Delta_{N}$. frequently use the following notation: for $c\in C$ and $n\geq1$, $c_0=fc$ and $c_n=f^nc_0$: $x\in S^1$ and $n\geq1$, $J(x)=|f'x|$ and $J^n(x)=J(x)J(fx)\cdots J(f^{n-1}x)$. Let $\alpha=10^{-6}$ and $\delta = N}$. what we $N$ to be a large integer for which $\delta\ll\sigma$, and the conclusion of Lemma \[initial-1\] holds. value of $N$ will be replaced if necessary, only a finite number times. The letter $K$ will used denote generic which independent $N$ and $L$. next lemma, the proof of which is given in Appendix, ensures an exponential growth of derivatives for segments lying $C_{\delta}$. \[outside\] exists L_0$ that if $L\geq $f = f_{a}$ is such that then the following holds: - if $n\geq1$ and $fx,\cdots, f^{n-1}x\notin then $J^{n}(x)\geq \delta L^{2\lambda n}$; - moreover $f^nx\in C_{\delta}$, then $J^{n}(x) \geq L^{2 \lambda [*Standing assumption for the rest of this section: $L\geq L_1$ and $a \in \Delta_{N}$.*]{} Recovering -------------------- For $f = $c \in C$ $n>N$ introduce conditions: $J^{j-i}(c_i)\geq L | for us to move forward. This appRoach for inItial SetUps Is EasiEr, anD leads to the desIRed aSymptotic measure estimaTe on $\DElTA$ as $L\TO\iNfty$.
WE move to THe EXPanDiNg ProPeRTy Of the MapS corresPonding to pAraMeTers in $\Delta_{N}$. wE fRequently uSe tHe following nOtaTion: foR $c\In C$ ANd $n\geQ1$, $c_0=fC$ and $c_N=f^nc_0$: foR $X\in S^1$ anD $n\geq1$, $J(x)=|f'x|$ AnD $j^n(x)=J(x)J(FX)\cdots J(F^{N-1}X)$.
LEt $\alPha=10^{-6}$ and $\delta = L^{-\alpha n}$. in WHat follows, we suPpose $N$ To BE a LARge IntEger for whiCh $\Delta\LL\sigma$, aND tHE COncLUsion of Lemma \[iNitial-1\] holds. tHe vAlue of $n$ wIll BE replaCed if NeCEssAry, but only a FiniTe number oF times. tHe letteR $k$ will be Used to DenOte GeneRIc CoNstAnTS whICh Are INdePendent oF $N$ AnD $L$.
The Next LEMMA, the ProOf of Which Is given in AppeNdiX, ensURes An expOnentIal gRoWth of DerivaTives FoR orbit segments lYing Outside of $c_{\deLtA}$.
\[ouTsIde\] ThERe exisTs $L_1\Geq l_0$ such thAt if $L\geQ l_1$ anD $f = F_{A}$ IS sUch that $a \in \Delta_{N}(L)$, tHeN THe FollowinG holds:
- IF $n\GeQ1$ And $x$, $fx,\cdOtS, f^{n-1}X\notIN c_{\deltA}$, theN $j^{n}(X)\geq \deltA L^{2\lambDA n}$;
- If MoreoveR $f^Nx\in C_{\dElTa}$, tHen $j^{n}(x) \geQ
l^{2 \lamBda n}$.
[*StAnding asSumptIOn for the rest of THis section: $L\geQ l_1$ aND $A \iN \deltA_{N}$.*]{}
REcovering exPansIOn {#s2.3}
--------------------
FOr $f = f_A$, $C \iN C$ aND $n>N$ we IntroDuCE tHRee conditions:
- $J^{j-i}(c_i)\gEq l \min | for us to move forward. T his approa ch fo r i nit ia l se tups is easier, an d lea ds to the desired asym ptoti cm easu r eestim ate on$ \D e l ta$ a s$L\ to \ in fty$.
W e moveto the exp and in g property o f t he maps co rre sponding topar ameter sin$ \Delt a_{ N}$.We fre q uently use thefo l lowing notatio n : f or $ c\in C$ and $n\ge q 1$ , $c_0=fc$ and$c_n=f ^n c _0 $ : fo r $ x\in S^1$an d $n\ g eq1$, $ J (x ) = | f'x | $ and $J^n(x) =J(x)J(fx)\ c dot s J(f^ {n -1} x )$.
L et $\ al p ha= 10^{-6}$ an d $\ delta = L ^{-\al p ha N}$. In what follo ws, we sup p os e$N$ t o be alar g e i nteger f or w hich$\de l t a \ ll\s igm a$,and t he conclusion of Lem m a \ [init ial-1 \] h ol ds. T he val ue of $ N$ will be repl aced if neces sar y, bu tonlya finit e n umb er of t imes. T h e l et t e r $ K$ will be used to d e n ot e generi c cons t an ts which ar eind epen d e nt of $N$ an d $L$.
The ne x tle mma, th eproofof wh ich is g i venin App endix, e nsure s an exponentia l growth of de r iv a t iv e s fo r o rbit segmen ts l y ingouts i de of $C_{\ delta }$ .
\ [outside\] There ex is ts $L_ 1\geq L_0$ such th at if $L\g e q L_1$ and $f= f _ {a}$ is such t hat $ a \in \Del t a_{N}(L) $, th en the f ollowingh o lds:
- i f $ n\g eq1 $ an d $x$, $fx,\c d o ts,f^ {n-1}x\ not in C_{\ del ta} $,the n$J^{n}(x) \geq \de lt aL^ {2 \la mbdan }$;
- i f m or eov er $f ^ nx\inC_{\d elta }$ ,t hen $J^{n} ( x) \ geq L^{2 \l am bda n }$.[*S tanding assumpti onf or t he r est ofthis section: $L\geq L_1 $and $a \i n \Delta_{ N}$.*]{}
Recovering ex p ansion{#s 2.3}---- --------- --- ----
For $f = f _a$, $ c \in C $ a n d $n>N $ we in tr oduce thre e con ditio ns :
- $J^{ j-i}(c_i)\geq L \m i n | for_us to_move forward. This approach_for initial_setups_is easier,_and_leads to the_desired asymptotic measure_estimate on $\Delta$ as_$L\to\infty$.
We move to_the_expanding property of the maps corresponding to parameters in $\Delta_{N}$. We frequently use the_following_notation: for_$c\in_C$_and $n\geq1$, $c_0=fc$ and $c_n=f^nc_0$:_for $x\in S^1$ and $n\geq1$,_$J(x)=|f'x|$ and_$J^n(x)=J(x)J(fx)\cdots J(f^{n-1}x)$.
Let $\alpha=10^{-6}$ and $\delta = L^{-\alpha N}$._In_what follows, we_suppose $N$ to be a large integer for which_$\delta\ll\sigma$, and the conclusion of Lemma_\[initial-1\] holds. The_value_of_$N$ will be replaced_if necessary, but only a finite_number of times. The letter $K$_will be used to denote generic constants_which are independent of $N$ and_$L$.
The next lemma, the proof_of which_is given in Appendix, ensures_an exponential growth_of derivatives_for orbit segments_lying outside of $C_{\delta}$.
\[outside\] There exists_$L_1\geq L_0$ such_that if $L\geq L_1$ and $f_=_f_{a}$ is such_that_$a_\in \Delta_{N}(L)$,_then the following_holds:
-_ _if_$n\geq1$ and $x$, $fx,\cdots, f^{n-1}x\notin C_{\delta}$,_then_$J^{n}(x)\geq \delta L^{2\lambda n}$;
- if_moreover $f^nx\in C_{\delta}$, then_$J^{n}(x)_\geq
_L^{2 \lambda n}$.
[*Standing assumption for_the rest of this section: $L\geq_L_1$ and_$a \in_\Delta_{N}$.*]{}
Recovering expansion {#s2.3}
--------------------
For $f = f_a$, $c \in C$ and $n>N$_we introduce three conditions:
- _$J^{j-i}(c_i)\geq L \min |
1}_\bm({\frac{\partial}}{{\partial x}_{\br}} w_j)\\
=& q^{-m_1m_2}(j-l)(j-l-1)x^{j-l-1}_\bm w_j+a_{j+1}x^{j-l}_\bm w_{j+1}\\
& +2(d-j)q^{-m_2m_1}(j-l)x^{j-l-1}_\bm w_j\\
=& q^{-m_1m_2}(j-l)(-j-l-1+2d)x^{j-l-1}_\bm w_j+a_{j+1}x^{j-l}_\bm w_{j+1}.\end{aligned}$$
Consequently, $$\begin{aligned}
& e_{12}(-\bm)w =\sum_{j=0}^d(q^{-m_1m_2}\mu{\frac{\partial}}{{\partial x}_{\bm}}-\mathcal{E}_\bm) x^{j-l}_\bm w_j\\
=& \sum_{j=0}^d q^{-m_1m_2}\mu\big((j-l)x^{j-l-1}_\bm w_j+ x^{j-l}_\bm( {\frac{\partial}}{{\partial x}_{\bm}} w_j)\big)\\
& -\sum_{j=0}^d q^{-m_1m_2}(j-l)(-j-l-1+2d)x^{j-l-1}_\bm w_j-\sum_{j=0}^d a_{j+1}x^{j-l}_\bm w_{j+1}\\
=& \sum_{j=0}^d q^{-m_1m_2}(\mu+j+l+1-2d)(j-l)x^{j-l-1}_\bm w_j -\sum_{j=1}^d a_{j}x^{j-l-1}_\bm w_{j}\\
=& q^{-m_1m_2}(\mu+l+1-2d)(-l)x^{-l-1}_\bm w_0=0,\end{aligned}$$ in the third equality, we have used the fact that $ {\frac{\partial}}{{\partial x}_{\bm}}(w_j)=0$.
| 1}_\bm({\frac{\partial}}{{\partial x}_{\br } } w_j)\\
= & q^{-m_1m_2}(j - l)(j - l-1)x^{j - l-1}_\bm w_j+a_{j+1}x^{j - l}_\bm w_{j+1}\\
& +2(d - j)q^{-m_2m_1}(j - l)x^{j - l-1}_\bm w_j\\
= & q^{-m_1m_2}(j - l)(-j - l-1 + 2d)x^{j - l-1}_\bm w_j+a_{j+1}x^{j - l}_\bm w_{j+1}.\end{aligned}$$
Consequently, $ $ \begin{aligned }
& e_{12}(-\bm)w = \sum_{j=0}^d(q^{-m_1m_2}\mu{\frac{\partial}}{{\partial x}_{\bm}}-\mathcal{E}_\bm) x^{j - l}_\bm w_j\\
= & \sum_{j=0}^d q^{-m_1m_2}\mu\big((j - l)x^{j - l-1}_\bm w_j+ x^{j - l}_\bm ({ \frac{\partial}}{{\partial x}_{\bm } } w_j)\big)\\
& -\sum_{j=0}^d q^{-m_1m_2}(j - l)(-j - l-1 + 2d)x^{j - l-1}_\bm w_j-\sum_{j=0}^d a_{j+1}x^{j - l}_\bm w_{j+1}\\
= & \sum_{j=0}^d q^{-m_1m_2}(\mu+j+l+1 - 2d)(j - l)x^{j - l-1}_\bm w_j -\sum_{j=1}^d a_{j}x^{j - l-1}_\bm w_{j}\\
= & q^{-m_1m_2}(\mu+l+1 - 2d)(-l)x^{-l-1}_\bm w_0=0,\end{aligned}$$ in the third equality, we have used the fact that $ { \frac{\partial}}{{\partial x}_{\bm}}(w_j)=0$. | 1}_\bm({\fgac{\partial}}{{\partial x}_{\br}} w_j)\\
=& q^{-m_1m_2}(j-l)(j-l-1)x^{j-l-1}_\bm w_l+q_{j+1}x^{j-l}_\bk w_{j+1}\\
& +2(d-n)q^{-m_2m_1}(j-l)x^{j-u-1}_\bm w_j\\
=& q^{-m_1m_2}(j-l)(-j-l-1+2d)x^{j-l-1}_\bm w_j+a_{j+1}x^{j-l}_\um w_{h+1}.\end{aoigned}$$
Consequently, $$\begkn{aligned}
& e_{12}(-\bm)w =\sun_{j=0}^d(q^{-n_1n_2}\mu{\frac{\parvjal}}{{\partlcl x}_{\bj}}-\latheao{E}_\bm) x^{j-l}_\bm w_j\\
=& \xum_{j=0}^d q^{-m_1m_2}\mg\big((j-l)x^{j-l-1}_\bm w_j+ b^{j-p}_\bm( {\frac{\partial}}{{\partial x}_{\bm}} w_j)\big)\\
& -\sui_{j=0}^d q^{-k_1m_2}(u-l)(-j-l-1+2d)x^{j-l-1}_\bm w_j-\som_{j=0}^d s_{t+1}x^{j-l}_\gm w_{j+1}\\
=& \sum_{j=0}^d q^{-m_1m_2}(\mu+j+l+1-2d)(j-l)x^{j-l-1}_\bm w_j -\sum_{n=1}^d a_{j}x^{o-l-1}_\bm w_{j}\\
=& q^{-m_1m_2}(\mu+l+1-2d)(-k)x^{-l-1}_\bm w_0=0,\end{aligned}$$ in the tjird equality, we have ksed the faxt trqt $ {\frac{\partkal}}{{\partial x}_{\bm}}(w_j)=0$.
| 1}_\bm({\frac{\partial}}{{\partial x}_{\br}} w_j)\\ =& q^{-m_1m_2}(j-l)(j-l-1)x^{j-l-1}_\bm w_j+a_{j+1}x^{j-l}_\bm w_{j+1}\\ w_j\\ q^{-m_1m_2}(j-l)(-j-l-1+2d)x^{j-l-1}_\bm w_j+a_{j+1}x^{j-l}_\bm Consequently, $$\begin{aligned} & =& q^{-m_1m_2}\mu\big((j-l)x^{j-l-1}_\bm w_j+ x^{j-l}_\bm( x}_{\bm}} w_j)\big)\\ & q^{-m_1m_2}(j-l)(-j-l-1+2d)x^{j-l-1}_\bm w_j-\sum_{j=0}^d a_{j+1}x^{j-l}_\bm w_{j+1}\\ =& \sum_{j=0}^d w_j -\sum_{j=1}^d a_{j}x^{j-l-1}_\bm w_{j}\\ =& q^{-m_1m_2}(\mu+l+1-2d)(-l)x^{-l-1}_\bm w_0=0,\end{aligned}$$ in the third equality, we have the fact that $ {\frac{\partial}}{{\partial x}_{\bm}}(w_j)=0$. | 1}_\bm({\frac{\partial}}{{\partial x}_{\br}} w_j)\\
=& Q^{-m_1m_2}(j-l)(j-l-1)x^{j-l-1}_\Bm w_j+a_{J+1}x^{j-L}_\bm W_{j+1}\\
& +2(D-j)q^{-m_2M_1}(j-l)x^{J-l-1}_\bm w_j\\
=& q^{-m_1m_2}(j-l)(-j-l-1+2d)X^{J-l-1}_\bm W_j+a_{j+1}x^{j-l}_\bm w_{j+1}.\end{aligned}$$
COnseqUeNTly, $$\bEGiN{aligNed}
& e_{12}(-\bm)w =\SUm_{J=0}^D(Q^{-m_1m_2}\Mu{\FrAc{\pArTIaL}}{{\partIal X}_{\bm}}-\mathCal{E}_\bm) x^{j-l}_\bM w_j\\
=& \SuM_{j=0}^d q^{-m_1m_2}\mu\big((j-L)X^{j-L-1}_\bm w_j+ x^{j-l}_\bm( {\FraC{\partial}}{{\partIal X}_{\bm}} w_j)\bIg)\\
& -\Sum_{J=0}^D q^{-m_1m_2}(j-L)(-j-l-1+2D)x^{j-l-1}_\bM w_j-\sum_{J=0}^D a_{j+1}x^{j-l}_\Bm w_{j+1}\\
=& \sum_{j=0}^d Q^{-m_1M_2}(\Mu+j+l+1-2d)(j-L)X^{j-l-1}_\bm w_j -\SUM_{j=1}^D a_{j}x^{J-l-1}_\bm w_{j}\\
=& q^{-m_1m_2}(\mu+l+1-2d)(-l)x^{-l-1}_\bM W_0=0,\eND{aligned}$$ in the tHird eqUaLItY, WE haVe uSed the fact ThAt $ {\fraC{\Partial}}{{\PArTIAL x}_{\bM}}(W_j)=0$.
| 1}_\bm({\frac{\partial}}{{ \partial x }_{\b r}} w_ j) \\
= & q^ {-m_1m_2}(j-l) ( j-l- 1)x^{j-l-1}_\bm w_j+a_ {j+1} x^ { j-l} _ \b m w_{ j+1}\\& + 2 ( d-j )q ^{ -m_ 2m _ 1} (j-l) x^{ j-l-1}_ \bm w_j\\=&q^ {-m_1m_2}(j- l )( -j-l-1+2d) x^{ j-l-1}_\bm w _j+ a_{j+1 }x ^{j - l}_\b m w _{j+1 }.\end { aligne d}$$
Con se q uently , $$\beg i n {a lign ed}
& e_{12}(-\bm ) w= \sum_{j=0}^d(q ^{-m_1 m_ 2 }\ m u {\f rac {\partial} }{ {\par t ial x}_ { \b m } } -\m a thcal{E}_\bm) x^{j-l}_\b m w_ j\\
=& \ sum _ {j=0}^ d q^{ -m _ 1m_ 2}\mu\big(( j-l) x^{j-l-1} _\bm w _ j+ x^{ j -l}_\bm ( {\fr ac{ \pa rtia l }} {{ \pa rt i alx }_ {\b m }}w_j)\big )\ \& -\s um_{ j = 0 } ^d q^ {-m_ 1m_2} (j-l)(-j-l-1+ 2d) x^{j - l-1 }_\bm w_j- \sum _{ j=0}^ d a_{j +1}x^ {j -l}_\bm w_{j+1} \\
= & \sum_{j =0} ^d q^ {- m_1m_ 2 }(\mu+ j+l +1- 2d)(j-l )x^{j-l - 1}_ \b m w _j -\sum_{j=1}^d a_ {j } x ^{ j-l-1}_\ bm w_{ j }\ \= & q^{-m_ 1m _2} (\mu + l +1-2d )(-l ) x^ {-l-1}_\ bm w_0 = 0, \e nd{alig ne d}$$ i nthe th ird e q uali ty, we have us ed th e fact that $ { \ frac{\partial } }{ { \ pa r tial x} _{\bm}}(w_j )=0$ .
| 1}_\bm({\frac{\partial}}{{\partial x}_{\br}}_w_j)\\
=& q^{-m_1m_2}(j-l)(j-l-1)x^{j-l-1}_\bm_w_j+a_{j+1}x^{j-l}_\bm w_{j+1}\\
& +2(d-j)q^{-m_2m_1}(j-l)x^{j-l-1}_\bm w_j\\
=&_q^{-m_1m_2}(j-l)(-j-l-1+2d)x^{j-l-1}_\bm w_j+a_{j+1}x^{j-l}_\bm_w_{j+1}.\end{aligned}$$
Consequently,_$$\begin{aligned}
& e_{12}(-\bm)w_=\sum_{j=0}^d(q^{-m_1m_2}\mu{\frac{\partial}}{{\partial_x}_{\bm}}-\mathcal{E}_\bm) x^{j-l}_\bm w_j\\
=&_\sum_{j=0}^d q^{-m_1m_2}\mu\big((j-l)x^{j-l-1}_\bm w_j+_ x^{j-l}_\bm( {\frac{\partial}}{{\partial x}_{\bm}}_w_j)\big)\\
& -\sum_{j=0}^d __q^{-m_1m_2}(j-l)(-j-l-1+2d)x^{j-l-1}_\bm w_j-\sum_{j=0}^d a_{j+1}x^{j-l}_\bm w_{j+1}\\
=& \sum_{j=0}^d q^{-m_1m_2}(\mu+j+l+1-2d)(j-l)x^{j-l-1}_\bm w_j -\sum_{j=1}^d a_{j}x^{j-l-1}_\bm w_{j}\\
=& q^{-m_1m_2}(\mu+l+1-2d)(-l)x^{-l-1}_\bm w_0=0,\end{aligned}$$ in_the_third equality,_we_have_used the fact that $_{\frac{\partial}}{{\partial x}_{\bm}}(w_j)=0$.
|
_1}{N}}-\frac{p_1}{q}\right|&=\frac{y_{(1)}\sqrt{c}}{bz_{(1)}z_{(2)}}|z_{(2)}\sqrt{b}-y_{(2)}\sqrt{c}|&<\frac{4(c-b)\sqrt{c}y_{(1)}}{2b\sqrt{b}z_{(1)}z_{(2)}^2}<\frac{2c^{3/2}}{b^{3/2}}z_{(2)}^{-2}\end{aligned}$$ and similarly $$\left|\sqrt{1+\frac{a_2}{N}}-\frac{p_2}{q}\right|<\frac{2c^{3/2}}{a^{3/2}}z_{(2)}^{-2}.$$
\[prop\_odnos\_n1\_n2\] Suppose that $\{a,b,c,d_i\}$ are $D(4)$-quadruples with $a<b<c<d_1<d_2$ and $x_{(i)},y_{(i)},z_{(i)}$ are positive integers such that $ad_i+1=x_{(i)}^2$, $bd_i+1=y_{(i)}^2$ and $cd_i=z_{(i)}^2$ for $i\in\{1,2\} $.
i) If $n_1\geq 8$, then $$n_2<\frac{(n_1+1.1)(3.5205n_1+4.75675)}{0.4795n_1-3.82175}-1.1.$$ More specifically,if $n_1=8$, $n_2<2628n_1$, and if $n_1\geq9$ then $n_2<83n_1$.
ii) If $c>ab+a+b$ and $(z_0,z_1)=(t,s)$, $z_0z_1>0$ and $n_1\geq 9$ then $$n_2<\frac{(n_1+1)(2.5147n_1+5.11467)}{0.4853n_1-3.85292}-1<60n_1.$$
Put $N=abz^2$, $p_1=acy_{(1)}y_{(2)}$, $p_2=bcx_{( | _ 1}{N}}-\frac{p_1}{q}\right|&=\frac{y_{(1)}\sqrt{c}}{bz_{(1)}z_{(2)}}|z_{(2)}\sqrt{b}-y_{(2)}\sqrt{c}|&<\frac{4(c - b)\sqrt{c}y_{(1)}}{2b\sqrt{b}z_{(1)}z_{(2)}^2}<\frac{2c^{3/2}}{b^{3/2}}z_{(2)}^{-2}\end{aligned}$$ and similarly $ $ \left|\sqrt{1+\frac{a_2}{N}}-\frac{p_2}{q}\right|<\frac{2c^{3/2}}{a^{3/2}}z_{(2)}^{-2}.$$
\[prop\_odnos\_n1\_n2\ ] Suppose that $ \{a, b, c, d_i\}$ are $ D(4)$-quadruples with $ a < b < c < d_1 < d_2 $ and $ x_{(i)},y_{(i)},z_{(i)}$ are incontrovertible integer such that $ ad_i+1 = x_{(i)}^2 $, $ bd_i+1 = y_{(i)}^2 $ and $ cd_i = z_{(i)}^2 $ for $ i\in\{1,2\ } $.
i) If $ n_1\geq 8 $, then $ $ n_2<\frac{(n_1 + 1.1)(3.5205n_1 + 4.75675)}{0.4795n_1 - 3.82175}-1.1.$$ More specifically, if $ n_1=8 $, $ n_2<2628n_1 $, and if $ n_1\geq9 $ then $ n_2<83n_1$.
ii) If $ c > ab+a+b$ and $ (z_0,z_1)=(t, s)$, $ z_0z_1>0 $ and $ n_1\geq 9 $ then $ $ n_2<\frac{(n_1 + 1)(2.5147n_1 + 5.11467)}{0.4853n_1 - 3.85292}-1<60n_1.$$
Put $ N = abz^2 $, $ p_1 = acy_{(1)}y_{(2)}$, $ p_2 = bcx _ { ( | _1}{N}}-\frwc{p_1}{q}\right|&=\frac{y_{(1)}\sqrt{c}}{bz_{(1)}z_{(2)}}|z_{(2)}\sdrt{b}-y_{(2)}\sqrt{c}|&<\frac{4(c-y)\wqrt{c}y_{(1)}}{2u\sqrt{b}z_{(1)}a_{(2)}^2}<\frac{2c^{3/2}}{b^{3/2}}z_{(2)}^{-2}\dnd{aligned}$$ and similarly $$\lefv|\sqrr{1+\frac{q_2}{N}}-\frac{p_2}{q}\right|<\frac{2c^{3/2}}{a^{3/2}}z_{(2)}^{-2}.$$
\[prup\_odnos\_n1\_n2\] Suppose thau $\{a,b,c,d_i\}$ are $D(4)$-quadchples wlch $a<b<d<f_1<d_2$ aud $x_{(i)},y_{(i)},z_{(i)}$ are poxitive intagers such thad $xd_n+1=x_{(i)}^2$, $bd_i+1=y_{(i)}^2$ and $cd_i=z_{(i)}^2$ for $i\in\{1,2\} $.
i) If $n_1\gqq 8$, them $$j_2<\frac{(n_1+1.1)(3.5205n_1+4.75675)}{0.4795n_1-3.82175}-1.1.$$ More fpecpfycalmj,in $n_1=8$, $n_2<2628n_1$, and if $n_1\geq9$ then $n_2<83n_1$.
ii) If $d>ab+a+b$ aid $(z_0,z_1)=(t,s)$, $z_0z_1>0$ and $m_1\geq 9$ then $$n_2<\frac{(n_1+1)(2.5147n_1+5.11467)}{0.4853n_1-3.85292}-1<60n_1.$$
Put $N=ahz^2$, $p_1=wcy_{(1)}y_{(2)}$, $p_2=bcx_{( | _1}{N}}-\frac{p_1}{q}\right|&=\frac{y_{(1)}\sqrt{c}}{bz_{(1)}z_{(2)}}|z_{(2)}\sqrt{b}-y_{(2)}\sqrt{c}|&<\frac{4(c-b)\sqrt{c}y_{(1)}}{2b\sqrt{b}z_{(1)}z_{(2)}^2}<\frac{2c^{3/2}}{b^{3/2}}z_{(2)}^{-2}\end{aligned}$$ and similarly $$\left|\sqrt{1+\frac{a_2}{N}}-\frac{p_2}{q}\right|<\frac{2c^{3/2}}{a^{3/2}}z_{(2)}^{-2}.$$ \[prop\_odnos\_n1\_n2\] Suppose that $D(4)$-quadruples $a<b<c<d_1<d_2$ and are positive integers $cd_i=z_{(i)}^2$ $i\in\{1,2\} $. i) $n_1\geq 8$, then More specifically,if $n_1=8$, $n_2<2628n_1$, and if then $n_2<83n_1$. ii) If $c>ab+a+b$ and $(z_0,z_1)=(t,s)$, $z_0z_1>0$ and $n_1\geq 9$ then $$n_2<\frac{(n_1+1)(2.5147n_1+5.11467)}{0.4853n_1-3.85292}-1<60n_1.$$ $N=abz^2$, $p_1=acy_{(1)}y_{(2)}$, $p_2=bcx_{( | _1}{N}}-\frac{p_1}{q}\right|&=\frac{y_{(1)}\sqrt{c}}{bz_{(1)}z_{(2)}}|Z_{(2)}\sqrt{b}-y_{(2)}\sqrT{c}|&<\fraC{4(c-b)\SqrT{c}Y_{(1)}}{2b\sqRt{b}z_{(1)}Z_{(2)}^2}<\frac{2c^{3/2}}{b^{3/2}}z_{(2)}^{-2}\end{aliGNed}$$ aNd similarly $$\left|\sqrt{1+\fraC{a_2}{N}}-\frAc{P_2}{Q}\rigHT|<\fRac{2c^{3/2}}{a^{3/2}}Z_{(2)}^{-2}.$$
\[prop\_odNOs\_N1\_N2\] supPoSe ThaT $\{a,B,C,d_I\}$ are $D(4)$-QuaDruples With $a<b<c<d_1<d_2$ aNd $x_{(I)},y_{(I)},z_{(i)}$ are positiVE iNtegers sucH thAt $ad_i+1=x_{(i)}^2$, $bd_i+1=y_{(i)}^2$ And $Cd_i=z_{(i)}^2$ fOr $I\in\{1,2\} $.
I) if $n_1\geQ 8$, thEn $$n_2<\frAc{(n_1+1.1)(3.5205n_1+4.75675)}{0.4795n_1-3.82175}-1.1.$$ MORe specIfically,iF $n_1=8$, $N_2<2628N_1$, and if $N_1\Geq9$ then $N_2<83N_1$.
Ii) if $c>aB+a+b$ and $(z_0,z_1)=(t,s)$, $z_0z_1>0$ and $n_1\gEQ 9$ tHEn $$n_2<\frac{(n_1+1)(2.5147n_1+5.11467)}{0.4853n_1-3.85292}-1<60n_1.$$
Put $n=abz^2$, $p_1=aCy_{(1)}Y_{(2)}$, $P_2=bCX_{( | _1}{N}}-\frac{p_1}{q}\righ t|&=\frac{ y_{(1 )}\ sqr t{ c}}{ bz_{ (1)}z_{(2)}}|z _ {(2) }\sqrt{b}-y_{(2)}\sqrt {c}|& <\ f rac{ 4 (c -b)\s qrt{c}y _ {( 1 ) }}{ 2b \s qrt {b } z_ {(1)} z_{ (2)}^2} <\frac{2c^ {3/ 2} }{b^{3/2}}z_ { (2 )}^{-2}\en d{a ligned}$$ an d s imilar ly $$ \ left| \sq rt{1+ \frac{ a _2}{N} }-\frac{p _2 } {q}\ri g ht|<\fr a c {2 c^{3 /2}}{a^{3/2}}z_{( 2 )} ^ {-2}.$$
\[pro p\_odn os \ _n 1 \ _n2 \]Suppose th at $\{a , b,c,d_i \ }$ a r e $ D (4)$-quadrupl es with $a< b <c< d_1<d_ 2$ an d $x_{( i)},y _{ ( i)} ,z_{(i)}$ a re p ositive i nteger s such t h at $ad_ i+1=x_ {(i )}^ 2$,$ bd _i +1= y_ { (i) } ^2 $ a n d $ cd_i=z_{ (i )} ^2$ f or $ i \ i n \{1, 2\} $.
i) If $n_1\geq 8 $,then $$n _2<\f rac{( n_1+ 1. 1)(3. 5205n_ 1+4.7 56 75)}{0.4795n_1- 3.82 175}-1.1. $$Mo resp ecifi c ally,i f $ n_1 =8$, $n _2<2628 n _1$ ,a n d i f $n_1\geq9$ then$n _ 2 <8 3n_1$.
ii) If $c >a b +a+b$ an d$(z _0,z _ 1 )=(t, s)$, $z _0z_1>0$ and $ n _1 \g eq 9$ t he n $$n_ 2< \fr ac{ (n_1+ 1 )(2. 5147n_ 1+5.1146 7)}{0 . 4853n_1-3.8529 2 }-1<60n_1.$$Pu t $N = abz^ 2$, $p_1=acy_{ (1)} y _{(2 )}$, $p _2= b cx_{( | _1}{N}}-\frac{p_1}{q}\right|&=\frac{y_{(1)}\sqrt{c}}{bz_{(1)}z_{(2)}}|z_{(2)}\sqrt{b}-y_{(2)}\sqrt{c}|&<\frac{4(c-b)\sqrt{c}y_{(1)}}{2b\sqrt{b}z_{(1)}z_{(2)}^2}<\frac{2c^{3/2}}{b^{3/2}}z_{(2)}^{-2}\end{aligned}$$ and_similarly $$\left|\sqrt{1+\frac{a_2}{N}}-\frac{p_2}{q}\right|<\frac{2c^{3/2}}{a^{3/2}}z_{(2)}^{-2}.$$
\[prop\_odnos\_n1\_n2\]_Suppose that $\{a,b,c,d_i\}$ are_$D(4)$-quadruples with_$a<b<c<d_1<d_2$_and $x_{(i)},y_{(i)},z_{(i)}$_are_positive integers such_that $ad_i+1=x_{(i)}^2$, $bd_i+1=y_{(i)}^2$_and $cd_i=z_{(i)}^2$ for $i\in\{1,2\}_$.
i) If_$n_1\geq_8$, then $$n_2<\frac{(n_1+1.1)(3.5205n_1+4.75675)}{0.4795n_1-3.82175}-1.1.$$ More specifically,if $n_1=8$, $n_2<2628n_1$, and if $n_1\geq9$ then $n_2<83n_1$.
ii) If $c>ab+a+b$_and_$(z_0,z_1)=(t,s)$, $z_0z_1>0$_and_$n_1\geq_9$ then $$n_2<\frac{(n_1+1)(2.5147n_1+5.11467)}{0.4853n_1-3.85292}-1<60n_1.$$
Put $N=abz^2$, $p_1=acy_{(1)}y_{(2)}$,_$p_2=bcx_{( |
@kern{-0.2} } \macc@depth\@ne
\let\math@bgroup\@empty \let\math@egroup\macc@set@skewchar
\mathsurround\z@ \frozen@everymath{\mathgroup\macc@group\relax} \macc@set@skewchar\relax
\let\mathaccentV\macc@nested@a
\macc@nested@a\relax111{R} \endgroup
}_j} / {(\kappa \underline{R}_j)}\right)^{1/\alpha_j} \right) \right] \left[\sum_{j=1}^d \max \left\{1, \log \left( \frac{M_j-m_j}{\eta_j} \right) \right\} \right].$$
Simulations for multidimensional models {#SectionSimulationDimD}
=======================================
In this section, we complete the simulation study of Section \[SectionSimuDim1\] by dealing with multidimensional models.
Models.
-------
We propose to work with the following models.
${\mathscr{F}}= \left\{f_{(m,\sigma)}, \, (m,\sigma) \in [-5, 5]\times [1/5,5]\right\} $ where $$f_{(m,\sigma)} (x) = \frac{1}{\sqrt{2 \pi} \sigma} \exp \left(- \frac{(x-m)^2}{2 \sigma^2} \right) \quad \text{ for all $x \in {\mathbb{R}}$.}$$
${\mathscr{F}}= \left\{f_{(m,\sigma)}, \, (m,\sigma) \in [-5, 5]\times [1/5,5]\right\} $ where $$f_{(m,\sigma)} (x) = \frac{ \sigma}{\pi \left((x-m)^2 + \sigma^2\right)} \quad \text{for all $x \in {\mathbb{R}}$.}$$
${\mathscr{F}}= \left\{f_{(a,b)}, \, (a,b) \in [0.6, 10]\times [0.1,20]\right\} $ where $$f_{(a,b)} | @kern{-0.2 } } \macc@depth\@ne
\let\math@bgroup\@empty \let\math@egroup\macc@set@skewchar
\mathsurround\z@ \frozen@everymath{\mathgroup\macc@group\relax } \macc@set@skewchar\relax
\let\mathaccentV\macc@nested@a
\macc@nested@a\relax111{R } \endgroup
} _ j } / { (\kappa \underline{R}_j)}\right)^{1/\alpha_j } \right) \right ] \left[\sum_{j=1}^d \max \left\{1, \log \left ( \frac{M_j - m_j}{\eta_j } \right) \right\ } \right].$$
Simulations for multidimensional models { # SectionSimulationDimD }
= = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = =
In this section, we complete the model discipline of Section \[SectionSimuDim1\ ] by dealing with multidimensional models.
Models.
-------
We propose to exercise with the following models.
$ { \mathscr{F}}= \left\{f_{(m,\sigma) }, \, (m,\sigma) \in [ -5, 5]\times [ 1/5,5]\right\ } $ where $ $ f_{(m,\sigma) } (x) = \frac{1}{\sqrt{2 \pi } \sigma } \exp \left(- \frac{(x - m)^2}{2 \sigma^2 } \right) \quad \text { for all $ x \in { \mathbb{R}}$.}$$
$ { \mathscr{F}}= \left\{f_{(m,\sigma) }, \, (m,\sigma) \in [ -5, 5]\times [ 1/5,5]\right\ } $ where $ $ f_{(m,\sigma) } (x) = \frac { \sigma}{\pi \left((x - m)^2 + \sigma^2\right) } \quad \text{for all $ x \in { \mathbb{R}}$.}$$
$ { \mathscr{F}}= \left\{f_{(a, b) }, \, (a, b) \in [ 0.6, 10]\times [ 0.1,20]\right\ } $ where $ $ f_{(a, b) } | @kerj{-0.2} } \macc@depth\@ne
\let\matm@bgroup\@empty \let\math@egrmup\macd@set@skewzhar
\mathsurround\z@ \frozen@evxrymqth{\maukgroup\macc@group\relax} \macc@set@dkewchar\eelae
\let\mathaccentT\jacc@nesbzd@a
\jwcc@nzsved@a\relax111{R} \endnroup
}_j} / {(\kap[a \underline{R}_j)}\siehc)^{1/\alpha_j} \right) \right] \left[\sum_{j=1}^d \max \left\{1, \lpg \left( \frac{M_j-i_j}{\ets_t} \rifht) \right\} \right].$$
Simulations for mhltidimtnsional models {#SevtionSimulationDimD}
=======================================
In this sechion, we complete tje simulatiin seydy of Sectiun \[SectionSpkuDim1\] by dgaling with multidimensional moddls.
Mobels.
-------
We propisw tl work with vhe fojlowing models.
${\mathscs{F}}= \left\{g_{(m,\sigma)}, \, (m,\sigms) \ii [-5, 5]\tumes [1/5,5]\right\} $ where $$f_{(m,\smgma)} (x) = \frac{1}{\sqrt{2 \pi} \sigma} \ex[ \meft(- \frac{(x-m)^2}{2 \sigma^2} \rught) \quag \tdzt{ wor akl $x \in {\latibb{R}}$.}$$
${\mathscr{R}}= \left\{f_{(m,\signa)}, \, (m,\sigma) \in [-5, 5]\timex [1/5,5]\wpbht\} $ where $$f_{(j,\sigma)} (x) = \frac{ \sigma}{\pi \left((x-m)^2 + \sigma^2\right)} \quad \test{for all $x \in {\mathbb{R}}$.}$$
${\mqthscr{F}}= \left\{f_{(a,b)}, \, (a,b) \ij [0.6, 10]\times [0.1,20]\wight\} $ where $$f_{(a,b)} | @kern{-0.2} } \macc@depth\@ne \let\math@bgroup\@empty \let\math@egroup\macc@set@skewchar \mathsurround\z@ \frozen@everymath{\mathgroup\macc@group\relax} \macc@nested@a\relax111{R} }_j} / \underline{R}_j)}\right)^{1/\alpha_j} \right) \right] \frac{M_j-m_j}{\eta_j} \right\} \right].$$ Simulations multidimensional models {#SectionSimulationDimD} In this section, we complete the study of Section \[SectionSimuDim1\] by dealing with multidimensional models. Models. ------- We propose work with the following models. ${\mathscr{F}}= \left\{f_{(m,\sigma)}, \, (m,\sigma) \in [-5, 5]\times [1/5,5]\right\} where (x) \frac{1}{\sqrt{2 \sigma} \exp \left(- \frac{(x-m)^2}{2 \sigma^2} \right) \quad \text{ for all $x \in {\mathbb{R}}$.}$$ ${\mathscr{F}}= \left\{f_{(m,\sigma)}, \, \in [-5, 5]\times [1/5,5]\right\} $ where $$f_{(m,\sigma)} (x) \frac{ \sigma}{\pi \left((x-m)^2 + \quad \text{for all $x \in ${\mathscr{F}}= \, (a,b) [0.6, [0.1,20]\right\} where $$f_{(a,b)} | @kern{-0.2} } \macc@depth\@ne
\let\math@bgrOup\@empty \leT\math@EgrOup\MaCc@seT@skeWchar
\mathsurroUNd\z@ \fRozen@everymath{\mathgrouP\macc@GrOUp\reLAx} \Macc@sEt@skewcHAr\RELax
\LeT\mAthAcCEnTV\macC@neSted@a
\maCc@nested@a\rElaX111{R} \Endgroup
}_j} / {(\kapPA \uNderline{R}_j)}\RigHt)^{1/\alpha_j} \righT) \riGht] \lefT[\sUm_{j=1}^D \Max \leFt\{1, \lOg \lefT( \frac{M_J-M_j}{\eta_j} \Right) \righT\} \rIGht].$$
SimULations FOR mUltiDimensional models {#sEcTIonSimulationDImD}
=======================================
In tHiS SeCTIon, We cOmplete the SiMulatIOn study OF SECTIon \[sEctionSimuDim1\] By dealing wiTH muLtidimEnSioNAl modeLs.
ModElS.
-------
we pRopose to worK witH the folloWing moDEls.
${\mathSCr{F}}= \left\{F_{(m,\sigmA)}, \, (m,\sIgmA) \in [-5, 5]\tIMeS [1/5,5]\rIghT\} $ wHEre $$F_{(M,\sIgmA)} (X) = \frAc{1}{\sqrt{2 \pi} \SiGmA} \exp \lEft(- \fRAC{(X-M)^2}{2 \sigMa^2} \rIght) \Quad \tExt{ for all $x \in {\mAthBb{R}}$.}$$
${\mAThsCr{F}}= \leFt\{f_{(m,\sIgma)}, \, (M,\sIgma) \iN [-5, 5]\times [1/5,5]\Right\} $ WhEre $$f_{(m,\sigma)} (x) = \frac{ \SigmA}{\pi \left((x-m)^2 + \SigMa^2\RigHt)} \Quad \tEXt{for aLl $x \In {\mAthbb{R}}$.}$$
${\mAthscr{F}}= \LEft\{F_{(a,B)}, \, (A,B) \In [0.6, 10]\Times [0.1,20]\right\} $ where $$f_{(a,b)} | @kern{-0.2} } \macc@dept h\@ne
\l et\ma th@ bgr ou p\@e mpty \let\math@egr o up\m acc@set@skewchar
\ma thsur ro u nd\z @ \ froze n@every m at h { \ma th gr oup \m a cc @grou p\r elax} \macc@set@ ske wc har\relax
\ le t\mathacce ntV \macc@nested @a \mac c@ nes t ed@a\ rel ax111 {R} \ e ndgrou p
}_j} /{( \ kappa\ underli n e {R }_j) }\right)^{1/\alph a _j } \right) \righ t] \le ft [ \s u m _{j =1} ^d \max \ left\ { 1, \log \ l e f t(\frac{M_j-m_j }{\eta_j} \ r igh t) \r ig ht\ } \rig ht].$ $Sim ulations fo r mu ltidimens ionalm odels { # Section Simula tio nDi mD}= == == === == = === = == === = === ======== == == ===
In t h i s sect ion , we comp lete the simu lat ions tud y ofSecti on \ [S ectio nSimuD im1\] b y dealing withmult idimensio nal m ode ls .
Mo d els.
- --- ---
We pr opose t o wo rk w i th the following mod el s .
${\maths cr{F}} = \ le f t\{f_{(m ,\ sig ma)} , \, (m ,\si g ma ) \in [- 5, 5]\ t im es [1/5,5 ]\ right\ }$ w her e $$f _ {(m, \sigma )} (x) = \fr a c{1}{\sqrt{2 \ p i} \sigma} \e x p\ l ef t (- \ fra c{(x-m)^2}{ 2 \s i gma^ 2} \ r ig ht) \qua d \t ex t {f or all $x \in {\mat hb b{R}}$ .}$$
${\mathscr{F }}= \left\ { f _ {(m,\sig ma)} , \ , (m,\sigma) \i n [-5 , 5]\times [1/5,5]\ right \} $ whe re $$f_{( m , \sigma)} (x ) = \f rac { \s igma}{\pi \le f t ((x- m) ^2 + \ sig ma^2\ri ght )}\qu ad\t ext{for a ll $x \i n{\ ma th bb{ R}}$. } $$
${\m at hsc r{ F}} = \le f t\{f_{ (a,b) }, \ ,(a , b)\in [0. 6 ,1 0 ]\ti me s[0.1 ,20 ]\ right \} $ whe re $$f_ {(a,b)} | @kern{-0.2} _} _\macc@depth\@ne
\let\math@bgroup\@empty \let\math@egroup\macc@set@skewchar
_ \mathsurround\z@_\frozen@everymath{\mathgroup\macc@group\relax}_ \macc@set@skewchar\relax
__\let\mathaccentV\macc@nested@a
\macc@nested@a\relax111{R}_ \endgroup
}_j} /_{(\kappa \underline{R}_j)}\right)^{1/\alpha_j} \right) \right]_\left[\sum_{j=1}^d __\max \left\{1, \log \left( \frac{M_j-m_j}{\eta_j} \right) \right\} \right].$$
Simulations for multidimensional_models_{#SectionSimulationDimD}
=======================================
In this_section,_we_complete the simulation study of_Section \[SectionSimuDim1\] by dealing with multidimensional_models.
Models.
-------
We propose_to work with the following models.
${\mathscr{F}}= \left\{f_{(m,\sigma)}, \,_(m,\sigma)_\in [-5, 5]\times_[1/5,5]\right\} $ where $$f_{(m,\sigma)} (x) = \frac{1}{\sqrt{2 \pi}_\sigma} \exp \left(- \frac{(x-m)^2}{2 \sigma^2} \right)_ \quad _\text{_for_all $x \in {\mathbb{R}}$.}$$
${\mathscr{F}}=_\left\{f_{(m,\sigma)}, \, (m,\sigma) \in [-5, 5]\times_[1/5,5]\right\} $ where $$f_{(m,\sigma)} (x) =_\frac{ \sigma}{\pi \left((x-m)^2 + \sigma^2\right)} \quad_\text{for all $x \in {\mathbb{R}}$.}$$
${\mathscr{F}}= \left\{f_{(a,b)},_\, (a,b) \in [0.6, 10]\times_[0.1,20]\right\} $_where $$f_{(a,b)} |
63, (1999)
Kropholler, P., Martínez-Pérez, C. and Nucinkis, B., [*Cohomological finiteness conditions for elementary amenable groups*]{}, Journal für Reine und Angewandte Mathematik 637 (2009), 49–62 Kropholler, P. and Mislin, G., [*Groups acting on finite dimensional spaces with finite stabilizers*]{}, Commentarii Mathematici Helvetici, vol. 73(1) (1998), 122–136
Lewis, L. G. Jr., [*The equivariant Hurewicz map,*]{} Trans. Amer. Math. Soc., vol. 329(2) (1992), 433-472
Lewis, G., May, J. P. and McClure, J., [*Ordinary RO(G)-graded cohomology*]{} Bull. Amer. Math. Soc. 4 (1981), 208–212
Leary, I. and Nucinkis, B., [*Some groups of type VF*]{}, Invent. Math. 151 (1) (2003), 135–162
Leary, I. J. and Petrosyan, N., [*On dimensions of groups with cocompact classifying spaces for proper actions*]{}, preprint
Lück, W., [*Transformation groups and algebraic K-theory*]{}, Lecture Notes in Mathematics, vol. 1408, Springer-Berlin (1989). Lück, W., [*The type of the classifying space for a family of subgroups*]{}, J. Pure Appl. Algebra 149 (2000), 177–203 L[ü]{}ck, W., [*Survey on classifying spaces for families of subgroups*]{}, Infinite Groups: Geometric, Combinatorial and Dynamical Aspects, Springer (2005), 269–322 Lück, W. and Meintrup, D., [*On the universal space for group actions with compact isotropy*]{}, Proc. of the conference “Geometry and Topology” in Aarhus, (1998), 293–305 L[ü]{}ck, W, and Oliver, B., [*The completion theorem in K-theory for proper actions of a discrete group*]{}, Topology 40 (2001), 585–616
Lück, W. and Reich H., [*The Baum-Connes and the Farrell-Jones conjectures in K- and L-theory*]{}, Handbook of K-theory., vol. 2, Springer, Berlin (2005), | 63, (1999)
Kropholler, P., Martínez - Pérez, C. and Nucinkis, B., [ * Cohomological finiteness conditions for elementary amenable groups * ] { }, Journal für Reine und Angewandte Mathematik 637 (2009), 49–62 Kropholler, P. and Mislin, G., [ * Groups acting on finite dimensional space with finite stabilizers * ] { }, Commentarii Mathematici Helvetici, vol. 73(1) (1998), 122–136
Lewis, L. G. Jr., [ * The equivariant Hurewicz function, * ] { } Trans. Amer. Math. Soc. , vol. 329(2) (1992), 433 - 472
Lewis, G., May, J. P. and McClure, J., [ * Ordinary RO(G)-graded cohomology * ] { } Bull. Amer. Math. Soc. 4 (1981), 208–212
Leary, I. and Nucinkis, B., [ * Some groups of character VF * ] { }, Invent. Math. 151 (1) (2003), 135–162
Leary, I. J. and Petrosyan, N., [ * On dimensions of group with cocompact classifying space for proper actions * ] { }, preprint
Lück, W., [ * Transformation groups and algebraic kelvin - theory * ] { }, Lecture Notes in Mathematics, vol. 1408, Springer - Berlin (1989). Lück, W., [ * The type of the relegate space for a family of subgroup * ] { }, J. Pure Appl. Algebra 149 (2000), 177–203 L[ü]{}ck, W., [ * Survey on classifying spaces for families of subgroups * ] { }, Infinite Groups: Geometric, Combinatorial and Dynamical Aspects, Springer (2005), 269–322 Lück, W. and Meintrup, D., [ * On the cosmopolitan space for group actions with compact isotropy * ] { }, Proc. of the league “ Geometry and Topology ” in Aarhus, (1998), 293–305 L[ü]{}ck, W, and Oliver, B., [ * The completion theorem in K - theory for proper actions of a discrete group * ] { }, Topology 40 (2001), 585–616
Lück, W. and Reich H., [ * The Baum - Connes and the Farrell - Jones conjectures in K- and L - theory * ] { }, Handbook of kelvin - theory. , vol. 2, Springer, Berlin (2005), | 63, (1999)
Kgopholler, P., Martínez-Pérez, G. and Nucinkis, B., [*Cohomonogicam finitevess conditions for elementacy anenaboe groups*]{}, Journal für Rdine und Wngewandre Methematik 637 (2009), 49–62 Kro'goller, I. and Lislnn, G., [*Groups actikg on finita dimensional vpxczs with finite stabilizers*]{}, Commentaryi Mathrmwtici Helveticy, vok. 73(1) (1998), 122–136
Lswis, L. G. Jr., [*The equivariant Hurewidz map,*]{} Urans. Amer. Math. Sov., vol. 329(2) (1992), 433-472
Lewis, G., May, J. P. anf McFlure, J., [*Ordinary RL(G)-graded coyomojigy*]{} Bull. Amef. Math. Soc. 4 (1981), 208–212
Leary, I. ahd Nucinkis, B., [*Some groups of tyoe VF*]{}, Invent. Majk. 151 (1) (2003), 135–162
Neary, I. J. aid Petgosyan, N., [*On dlkensiots of gtoups with cocpmpect xlassifying spaces foc proper actions*]{}, prekrint
Lück, W., [*Txansformation groups qne alggbraiw K-tfworh*]{}, Ltctnre Notes in Mathematids, vol. 1408, Sprunger-Berlin (1989). Lück, W., [*Uhe nupe of the cmassifrigg space for a family of subgroups*]{}, J. Puge Alpl. Algebra 149 (2000), 177–203 L[ü]{}ck, W., [*Suevey on classifying skaces for samilies of subgroups*]{}, Infinite Groups: Geometric, Cmmbineturico and Eyjamical Aspects, Springer (2005), 269–322 Lück, W. and Meintrup, S., [*Pn the universal space for brlul actions with compaef jsotropy*]{}, Proc. of tje confgrence “Geometry and Topology” in Aarhus, (1998), 293–305 L[ü]{}ck, W, and Oliver, Y., [*Tye completion theoxem in K-theoxy for proprr actions of a discretz groul*]{}, Topology 40 (2001), 585–616
Lück, W. aha Reich H., [*The Bajm-Cpntes and ufe Farrell-Jones cjnjecturew in K- and U-thepry*]{}, Hagdbook of N-theovf., vol. 2, Springer, Beglin (2005), | 63, (1999) Kropholler, P., Martínez-Pérez, C. and [*Cohomological conditions for amenable groups*]{}, Journal 637 49–62 Kropholler, P. Mislin, G., [*Groups on finite dimensional spaces with finite Commentarii Mathematici Helvetici, vol. 73(1) (1998), 122–136 Lewis, L. G. Jr., [*The equivariant map,*]{} Trans. Amer. Math. Soc., vol. 329(2) (1992), 433-472 Lewis, G., May, J. and J., RO(G)-graded Bull. Amer. Math. Soc. 4 (1981), 208–212 Leary, I. and Nucinkis, B., [*Some groups of type Invent. Math. 151 (1) (2003), 135–162 Leary, I. and Petrosyan, N., [*On of groups with cocompact classifying for actions*]{}, preprint W., groups algebraic K-theory*]{}, Lecture in Mathematics, vol. 1408, Springer-Berlin (1989). Lück, W., [*The type of the classifying space for a family subgroups*]{}, J. Algebra 149 177–203 W., on classifying spaces of subgroups*]{}, Infinite Groups: Geometric, Combinatorial Springer (2005), 269–322 Lück, W. and Meintrup, D., the universal for group actions with compact isotropy*]{}, of the conference “Geometry and Topology” in Aarhus, 293–305 L[ü]{}ck, W, and Oliver, B., [*The completion theorem in K-theory for proper actions of group*]{}, Topology 40 (2001), Lück, W. and H., Baum-Connes the conjectures in and L-theory*]{}, Handbook of K-theory., vol. 2, Springer, Berlin (2005), | 63, (1999)
Kropholler, P., Martínez-Pérez, C. And NucinkiS, B., [*CohOmoLogIcAl fiNiteNess conditions FOr elEmentary amenable groups*]{}, journAl FÜr ReINe Und AnGewandtE maTHEmaTiK 637 (2009), 49–62 KRopHoLLeR, P. and misLin, G., [*GroUps acting oN fiNiTe dimensionaL SpAces with fiNitE stabilizers*]{}, comMentarIi matHEmatiCi HElvetIci, vol. 73(1) (1998), 122–136
lEwis, L. G. jr., [*The equiVaRIant HuREwicz maP,*]{} tRaNs. AmEr. Math. Soc., vol. 329(2) (1992), 433-472
Lewis, g., maY, j. P. and McClure, J., [*ORdinarY Ro(g)-gRADed CohOmology*]{} BulL. AMer. MaTH. Soc. 4 (1981), 208–212
LeaRY, I. AND nucINkis, B., [*Some grouPs of type VF*]{}, INVenT. Math. 151 (1) (2003), 135–162
LEaRy, I. j. And PetRosyaN, N., [*oN diMensions of gRoupS with cocoMpact cLAssifyiNG spaces For proPer ActIons*]{}, PRePrInt
lüCK, W., [*TRAnSfoRMatIon groupS aNd AlgebRaic k-THEOry*]{}, LEctUre NOtes iN Mathematics, vOl. 1408, SPrinGEr-BErlin (1989). lück, W., [*the tYpE of thE classIfyinG sPace for a family oF subGroups*]{}, J. PuRe APpL. AlGeBra 149 (2000), 177–203 L[ü]{}CK, W., [*SurvEy oN clAssifyiNg spaceS For FaMILIeS of subgroups*]{}, InfiniTe gROuPs: GeometRic, ComBInAtORial and DYnAmiCal ASPEcts, SPrinGEr (2005), 269–322 lück, W. and meintrUP, D., [*on The univErSal spaCe For GroUp actIOns wIth comPact isotRopy*]{}, PROc. of the confereNCe “Geometry and tOpOLOgY” In AaRhuS, (1998), 293–305 L[ü]{}ck, W, and OlIver, b., [*the cOmplETiOn tHEorem In K-thEoRY fOR proper actions of a diScRete grOup*]{}, ToPology 40 (2001), 585–616
Lück, W. anD Reich H., [*The bAUM-Connes aNd thE faRRell-Jones conjeCtureS in K- and L-thEOry*]{}, HandbOok of k-theory., vOl. 2, SpringeR, bErlin (2005), | 63, (1999)
Kropholler, P ., Martíne z-Pér ez, C. a nd N ucin kis, B., [*Coh o molo gical finiteness condi tions f o r el e me ntary amenab l eg r oup s* ]{ },Jo u rn al fü r R eine un d Angewand teMa thematik 637 (2 009), 49–6 2 K ropholler, P . a nd Mis li n,G ., [* Gro ups a ctingo n fini te dimens io n al spa c es with f in itestabilizers*]{},C om m entarii Mathem aticiHe l ve t i ci, vo l. 73(1) ( 19 98),1 22–136Le w i s , L . G. Jr., [*Th e equivaria n t H urewic zmap , *]{} T rans. A m er. Math. Soc. , vo l. 329(2) (1992 ) , 433-4 7 2
Lewi s, G., Ma y,J. P . a nd Mc Cl u re, J. , [ * Ord inary RO (G )- grade d co h o m o logy *]{ } Bu ll. A mer. Math. So c.4 (1 9 81) , 208 –212
Lea ry , I.and Nu cinki s, B., [*Some gro upsof type V F*] {} , I nv ent.M ath. 1 51(1) (2003) , 135–1 6 2
Le a r y ,I. J. and Petrosya n, N ., [*On di mensio n sof groups w it h c ocom p a ct cl assi f yi ng space s forp ro pe r actio ns *]{},pr epr int
Lüc k , W. , [*Tr ansforma tiong roups and alge b raic K-theory * ]{ } , L e ctur e N otes in Mat hema t ics, vol . 1 408 , Spri nger- Be r li n (1989). Lück, W.,[* The ty pe of the classify ing spacef o r a famil y of su b groups*]{}, J. Pure Appl. Alg e bra 149(2000 ), 177–2 03 L[ü]{} c k , W., [* Sur vey on cl a s si fying spacesf o r fa mi lies of su bgroups *]{ },Inf ini te Groups:Geometri c, C om bi nat orial and Dyna mi cal A spe cts,S pringe r (20 05), 2 69 – 322 Lück,W .a n d Me in tr up,D., [ *On t he u n ive rsal sp ace for g rou p act io ns with c ompact isotro py *]{}, Proc .ofthe co n f erence “ Geometry and Topology”i n Aarhu s,(1998 ), 2 93–305 L[ ü]{ }ck, W , a n d Oliv er, B. , [*T he co m p letio n th eor em in K-theo r y fo r pro pe r ac tions o f a discrete group * ]{} , Topology 40 (2 001) , 58 5–6 1 6Lüc k, W.a n d Reich H., [*T he Baum-Co nn e sand the Fa r rel l- Jones c onjectu res i n K- and L-theory *]{}, Han db ooko f K- theory., v ol. 2, S pringer,B erlin (2 005), | 63,_(1999)
Kropholler, P.,_Martínez-Pérez, C. and Nucinkis,_B., [*Cohomological_finiteness_conditions for_elementary_amenable groups*]{}, Journal_für Reine und_Angewandte Mathematik 637 (2009),_49–62 Kropholler, P._and_Mislin, G., [*Groups acting on finite dimensional spaces with finite stabilizers*]{}, Commentarii Mathematici Helvetici,_vol._73(1) (1998),_122–136
Lewis,_L._G. Jr., [*The equivariant Hurewicz_map,*]{} Trans. Amer. Math. Soc.,_vol. 329(2)_(1992), 433-472
Lewis, G., May, J. P. and McClure,_J.,_[*Ordinary RO(G)-graded cohomology*]{}_Bull. Amer. Math. Soc. 4 (1981), 208–212
Leary, I. and_Nucinkis, B., [*Some groups of type_VF*]{}, Invent. Math._151_(1)_(2003), 135–162
Leary, I. J._and Petrosyan, N., [*On dimensions of_groups with cocompact classifying spaces for_proper actions*]{}, preprint
Lück, W., [*Transformation groups and_algebraic K-theory*]{}, Lecture Notes in Mathematics,_vol. 1408, Springer-Berlin (1989). Lück,_W., [*The_type of the classifying space_for a family_of subgroups*]{},_J. Pure Appl._Algebra 149 (2000), 177–203 L[ü]{}ck, W.,_[*Survey on classifying_spaces for families of subgroups*]{}, Infinite_Groups:_Geometric, Combinatorial and_Dynamical_Aspects,_Springer (2005),_269–322 Lück, W._and_Meintrup, D.,_[*On_the universal space for group actions_with_compact isotropy*]{}, Proc. of the conference “Geometry_and Topology” in Aarhus,_(1998),_293–305 L[ü]{}ck, W, and_Oliver, B., [*The completion theorem_in K-theory for proper actions of_a discrete_group*]{}, Topology_40 (2001), 585–616
Lück, W. and Reich H., [*The Baum-Connes and the_Farrell-Jones conjectures in K- and L-theory*]{},_Handbook of K-theory., vol._2, Springer,_Berlin_(2005), |
dedicated to the proof of Theorem \[theoremPerturbedDelaunay\].
The ${\mathcal{C}}^2$-regularity of $\xi_t$ essentially means that $C_t(z)$ is ${\mathcal{C}}^2$ with respect to $(t,z)$. Together with the ${\mathcal{C}}^2$-regularity of $\phi_t$, it implies that $\Phi_t$ is ${\mathcal{C}}^2$ with respect to $(t,z)$. Thus, ${\mathcal{M}}(\Phi_t)$ is also ${\mathcal{C}}^2$ with respect to $t$. These regularities and the fact that there exists a solution $\Phi_t$ solving the monodromy problem are used in Section \[sectionPropertyXit\] to deduce an essential piece of information about the potential $\xi_t$ (Proposition \[propC3\]). This step then allows us to write in Section \[sectionzAP\] the holomorphic frame $\Phi_t$ in a $Mz^AP$ form given by the Fröbenius method (Proposition \[propzAP1\]), and to gauge this expression, in order to gain an order of convergence with respect to $z$ (Proposition \[propzAP2\]). During this process, the holomorphic frame will loose one order of regularity with respect to $t$, which is why Theorem \[theoremPerturbedDelaunay\] asks for a ${\mathcal{C}}^2$-regularity of the data. Section \[sectionDressedDelaunayFrames\] is devoted to the study of dressed Delaunay frames $Mz^A$ in order to ensure that the immersions $f_t$ will converge to Delaunay immersions as $t$ tends to $0$, and to estimate the growth of their unitary part around the end at $z=0$. Section \[sectionConvergenceImmersions\] proves that these immersions do converge, which is the first point of Theorem \[theoremPerturbedDelaunay\]. Before proving the embeddedness in Section \[sectionEmbeddedness\], Section \[sectionConvergenceNormales\] is devoted to the convergence of the normal maps. Finally, we compute the limit axes in Section \[sectionLimitAxis\].
A property of $\xi_t$ {#sectionPropertyXit}
---------------------
We begin by diagonalising $A_t$ in a unitary basis (Proposition \[propDiagA\]) in order simplify the computations in Proposition | dedicated to the proof of Theorem \[theoremPerturbedDelaunay\ ].
The $ { \mathcal{C}}^2$-regularity of $ \xi_t$ essentially means that $ C_t(z)$ is $ { \mathcal{C}}^2 $ with obedience to $ (t, z)$. in concert with the $ { \mathcal{C}}^2$-regularity of $ \phi_t$, it implies that $ \Phi_t$ is $ { \mathcal{C}}^2 $ with respect to $ (t, z)$. therefore, $ { \mathcal{M}}(\Phi_t)$ is also $ { \mathcal{C}}^2 $ with regard to $ t$. These regularities and the fact that there exist a solution $ \Phi_t$ resolve the monodromy problem are used in Section \[sectionPropertyXit\ ] to deduce an essential objet d'art of information about the potential $ \xi_t$ (Proposition \[propC3\ ]). This footprint then admit us to write in Section \[sectionzAP\ ] the holomorphic frame $ \Phi_t$ in a $ Mz^AP$ form given by the Fröbenius method (suggestion \[propzAP1\ ]), and to gauge this expression, in order to reach an order of convergence with respect to $ z$ (Proposition \[propzAP2\ ]). During this process, the holomorphic human body will loose one order of regularity with respect to $ t$, which is why Theorem \[theoremPerturbedDelaunay\ ] asks for a $ { \mathcal{C}}^2$-regularity of the data. incision \[sectionDressedDelaunayFrames\ ] is devoted to the study of dressed Delaunay frames $ Mz^A$ in order to ensure that the immersions $ f_t$ will converge to Delaunay immersions as $ t$ tends to $ 0 $, and to estimate the growth of their unitary part around the end at $ z=0$. Section \[sectionConvergenceImmersions\ ] proves that these immersions do converge, which is the first detail of Theorem \[theoremPerturbedDelaunay\ ]. Before proving the embeddedness in Section \[sectionEmbeddedness\ ], Section \[sectionConvergenceNormales\ ] is devote to the convergence of the normal maps. Finally, we compute the terminus ad quem axes in Section \[sectionLimitAxis\ ].
A place of $ \xi_t$ { # sectionPropertyXit }
---------------------
We begin by diagonalising $ A_t$ in a unitary footing (Proposition \[propDiagA\ ]) in order simplify the computations in Proposition | deficated to the proof of Uheorem \[theoremPettyrbedDxlaunay\].
Fhe ${\mathzal{C}}^2$-regularity of $\xi_t$ essentmallt meabs that $C_t(z)$ is ${\mathcal{Z}}^2$ with redpect to $(t,z)$. Uogether with the ${\mathcal{G}}^2$-xegulzvity mh $\phi_t$, it impligs that $\Phi_t$ is ${\mathcal{C}}^2$ whtf xespect to $(t,z)$. Thus, ${\mathcal{M}}(\Phi_t)$ is ajso ${\matncwl{C}}^2$ with respest tp $t$. Tgvst regularities and the fact that fhere eqists a solution $\Lhi_t$ solving the monodromy prohlem are used in Sfction \[sectuonPwipertyXit\] to deduce an essential pjece of information about the putentnal $\xi_t$ (Prokoaihhon \[propC3\]). Tiis stvp then allows us to frite im Section \[sectlonzA'\] thw holomorphic frame $\Pii_t$ in a $Mz^AP$ form gyven by tve Fröbenius method (Priposijion \[[ropxQP1\]), xnd tp fauge hhia expressikn, in order to gain an order og sinvergence wifh res[est to $z$ (Proposition \[propzAP2\]). During this prkcess, the holomorphic feame will loose one otder of redularity with respect to $t$, which is why Theorem \[tveorejOeruuvnedDdoaknay\] asks for a ${\mathcal{C}}^2$-regularity of the datw. Stctpon \[sectionDressebDelaunayFrames\] os drdoted to the rtudy of sressed Delaunay fgames $Me^A$ in irder to tnsurr that the immersions $f_t$ wiol converge no Dwlaunay immersions as $t$ tends to $0$, and yo estimate the growth uf tgeir unitarj part arkjnd the end at $z=0$. Sebtiot \[sectionConvergenceImmersyons\] provxs thct these immgrsions do converhe, whlwh is the first polnt oy Themrem \[theorfmPerturbedDelaunay\]. Before provmig the embeddgdnass in Sectnon \[segtionEmbeddednefs\], Section \[secjionConvexgenceVormales\] ia devotxd to the cogvergence of dje normal ma's. Finallr, we compute thd limit axes in Section \[sectionLimutAxis\].
A property on $\xi_t$ {#aectionPropertyRnt}
---------------------
Qe begin by diabonxlifijg $A_e$ in a unitarf baris (Lroporition \[kxopBixgA\]) on order simplify tha cojputations in Proppsltion | dedicated to the proof of Theorem \[theoremPerturbedDelaunay\]. of essentially means $C_t(z)$ is ${\mathcal{C}}^2$ with ${\mathcal{C}}^2$-regularity of $\phi_t$, implies that $\Phi_t$ ${\mathcal{C}}^2$ with respect to $(t,z)$. Thus, is also ${\mathcal{C}}^2$ with respect to $t$. These regularities and the fact that exists a solution $\Phi_t$ solving the monodromy problem are used in Section \[sectionPropertyXit\] deduce essential of about the potential $\xi_t$ (Proposition \[propC3\]). This step then allows us to write in Section \[sectionzAP\] holomorphic frame $\Phi_t$ in a $Mz^AP$ form given the Fröbenius method (Proposition and to gauge this expression, order gain an of with to $z$ (Proposition During this process, the holomorphic frame will loose one order of regularity with respect to $t$, which why Theorem for a of data. \[sectionDressedDelaunayFrames\] is devoted study of dressed Delaunay frames $Mz^A$ ensure that the immersions $f_t$ will converge to immersions as tends to $0$, and to estimate growth of their unitary part around the end $z=0$. Section \[sectionConvergenceImmersions\] proves that these immersions do converge, which is the first point of Before proving the embeddedness Section \[sectionEmbeddedness\], Section is to convergence the normal Finally, we compute the limit axes in Section \[sectionLimitAxis\]. A property $\xi_t$ {#sectionPropertyXit} --------------------- We begin by diagonalising $A_t$ in a (Proposition in order simplify computations in Proposition | dedicated to the proof of TheoRem \[theorempertuRbeDDeLaUnay\].
the ${\mAthcal{C}}^2$-regularITy of $\Xi_t$ essentially means thaT $C_t(z)$ iS ${\mAThcaL{c}}^2$ wIth reSpect to $(T,Z)$. TOGEthEr WiTh tHe ${\MAtHcal{C}}^2$-RegUlarity Of $\phi_t$, it imPliEs That $\Phi_t$ is ${\maTHcAl{C}}^2$ with resPecT to $(t,z)$. Thus, ${\matHcaL{M}}(\Phi_t)$ Is AlsO ${\MathcAl{C}}^2$ With rEspect TO $t$. ThesE regulariTiES and thE Fact thaT THeRe exIsts a solution $\Phi_t$ SOlVIng the monodromY problEm ARe USEd iN SeCtion \[sectiOnpropeRTyXit\] to DEdUCE An eSSential piece oF informatioN AboUt the pOtEntIAl $\xi_t$ (PRoposItIOn \[pRopC3\]). This steP theN allows us To writE In SectiON \[sectioNzAP\] thE hoLomOrphIC fRaMe $\PHi_T$ In a $mZ^Ap$ foRM giVen by the frÖbEnius MethOD (pROposItiOn \[prOpzAP1\]), And to gauge thiS exPresSIon, In ordEr to gAin aN oRder oF conveRgencE wIth respect to $z$ (PrOposItion \[propZAP2\]). duRinG tHis prOCess, thE hoLomOrphic fRame wilL LooSe ONE OrDer of regularity witH rESPeCt to $t$, whiCh is whY thEoREm \[theoreMPErtUrbeDdElaunAy\] asKS fOr a ${\mathcAl{C}}^2$-regULaRiTy of the DaTa. SectIoN \[seCtiOnDreSSedDElaunaYFrames\] iS devoTEd to the study of DRessed DelaunaY FrAMEs $mZ^A$ in OrdEr to ensure tHat tHE immErsiONs $F_t$ wILl conVerge To dElAUnay immersions as $t$ teNdS to $0$, and To estImate the growtH of their unITARy part arOund THe ENd at $z=0$. Section \[seCtionconvergencEimmersioNs\] proVes that tHese immerSIOns do conVerGe, wHicH is THE fIrst point of ThEORem \[tHeOremPerTurBedDelaUnaY\]. BeForE prOvIng the embEddednesS iN SEcTiOn \[sEctioNembeddedNeSs\], SEcTioN \[sectIOnConvErgenCeNoRmAlES\] is Devoted TO tHE ConvErGeNce oF thE nOrmal Maps. fInaLly, we coMpute the lImiT Axes In seCtion \[seCtionLimitAxiS\].
A Property of $\Xi_T$ {#seCtionPROPertyXit}
---------------------
we begin by diagonalising $A_T$ In a unitAry Basis (propOsition \[prOpDIagA\]) in OrdER simplIfy the CompuTaTioNS In ProPOSiTioN | dedicated to the proof of Theorem \ [theo rem Per tu rbed Dela unay\].
The $ { \mat hcal{C}}^2$-regularity of $ \x i _t$e ss entia lly mea n st h at$C _t (z) $i s${\ma thc al{C}}^ 2$ with re spe ct to $(t,z)$. To gether wit h t he ${\mathca l{C }}^2$- re gul a rityof$\phi _t$, i t impli es that $ \P h i_t$ i s ${\mat h c al {C}} ^2$ with respectt o$ (t,z)$. Thus,${\mat hc a l{ M } }(\ Phi _t)$ is al so ${\m a thcal{C } }^ 2 $ wit h respect to $ t$. These r e gul aritie sand the fa ct th at the re exists a sol ution $\P hi_t$s olvingt he mono dromypro ble m ar e u se d i nS ect i on \[ s ect ionPrope rt yX it\]to d e d u c e an es sent ial p iece of infor mat iona bou t the pote ntia l$\xi_ t$ (Pr oposi ti on \[propC3\]). Thi s step th enal low sus to writeinSec tion \[ section z AP\ ]t h e h olomorphic frame $ \P h i _t $ in a $ Mz^AP$ fo rm given by t heFröb e n ius m etho d ( Proposit ion \[ p ro pz AP1\]), a nd toga uge th is ex p ress ion, i n orderto ga i n an order ofc onvergence wi t hr e sp e ct t o $ z$ (Proposi tion \[pr opzA P 2\ ]). Durin g thi sp ro c ess, the holomorphi cframewillloose one ord er of regu l a r ity with res p ec t to $t$, which is w hy Theorem \[theore mPert urbedDel aunay\] a s k s for a${\ mat hca l{C } } ^2 $-regularityo f the d ata. Se cti on \[se cti onD res sed De launayFra mes\] is d ev ot ed to thes tudy ofdr ess ed De launa y frame s $Mz ^A$in o r der to ens u re t hatth eimme rsi on s $f_ t$ w i llconverg e to Dela una y imm er si ons as$t$ tends to$0 $, and toes tim ate th e growth o f their unitary part ar o und the en d at$z=0 $. Sectio n \ [secti onC o nverge nceImm ersio ns \]p r ovest h at th es e immersio n s do conv er ge,which i s the first pointo f T heorem \[theo rem Pert u r be dDe l au n ay\ ]. Bef o r e proving the e mbeddednes si nSection \[ s ect io nEmbedd edness\ ], Se c tion \[ sectionCo nvergence No rmal e s \]is devoted to theconvergen c e oft he norm almaps.Fi nal ly, w e comp u tethe l imit a xe s in S ectio n\[sectio nLimitAxis\].
A proper ty of$\xi_ t${#section Pro p ert yXit}
--- ---- ---------- --- -
We be gin by di agon a li sin g $A_t $ in a unitary ba sis ( Pr oposition \ [ p r opD iagA\ ])i n orde r si mplify the comput a tions in Propo siti o n | dedicated_to the_proof of Theorem \[theoremPerturbedDelaunay\].
The_${\mathcal{C}}^2$-regularity of_$\xi_t$_essentially means_that_$C_t(z)$ is ${\mathcal{C}}^2$_with respect to_$(t,z)$. Together with the_${\mathcal{C}}^2$-regularity of $\phi_t$,_it_implies that $\Phi_t$ is ${\mathcal{C}}^2$ with respect to $(t,z)$. Thus, ${\mathcal{M}}(\Phi_t)$ is also ${\mathcal{C}}^2$_with_respect to_$t$._These_regularities and the fact that_there exists a solution $\Phi_t$_solving the_monodromy problem are used in Section \[sectionPropertyXit\] to_deduce_an essential piece_of information about the potential $\xi_t$ (Proposition \[propC3\]). This_step then allows us to write_in Section \[sectionzAP\]_the_holomorphic_frame $\Phi_t$ in a_$Mz^AP$ form given by the Fröbenius_method (Proposition \[propzAP1\]), and to gauge_this expression, in order to gain an_order of convergence with respect to_$z$ (Proposition \[propzAP2\]). During this_process, the_holomorphic frame will loose one_order of regularity_with respect_to $t$, which_is why Theorem \[theoremPerturbedDelaunay\] asks for_a ${\mathcal{C}}^2$-regularity of_the data. Section \[sectionDressedDelaunayFrames\] is devoted_to_the study of_dressed_Delaunay_frames $Mz^A$_in order to_ensure_that the_immersions_$f_t$ will converge to Delaunay immersions_as_$t$ tends to $0$, and to estimate_the growth of their_unitary_part around the end_at $z=0$. Section \[sectionConvergenceImmersions\] proves_that these immersions do converge, which_is the_first point_of Theorem \[theoremPerturbedDelaunay\]. Before proving the embeddedness in Section \[sectionEmbeddedness\], Section_\[sectionConvergenceNormales\] is devoted to the convergence_of the normal maps._Finally, we_compute_the limit axes_in_Section \[sectionLimitAxis\].
A_property of $\xi_t$ {#sectionPropertyXit}
---------------------
We begin by diagonalising_$A_t$ in_a unitary basis (Proposition \[propDiagA\]) in_order simplify the computations_in_Proposition |
important role in manipulating the generalized Lebesgue-Sobolev spaces is played by the [*modular*]{} of the $L^{p(x)}(\Omega)$ space, which is the mapping $\rho_{p(x)}:L^{p(x)}(\Omega)\rightarrow\RR$ defined by $$\rho_{p(x)}(u)=\int_\Omega|u|^{p(x)}\;dx.$$ If $(u_n)$, $u\in L^{p(x)}(\Omega)$ and $p^+<\infty$ then the following relations holds true $$\label{L4}
|u|_{p(x)}>1\;\;\;\Rightarrow\;\;\;|u|_{p(x)}^{p^-}\leq\rho_{p(x)}(u)
\leq|u|_{p(x)}^{p^+}$$ $$\label{L5}
|u|_{p(x)}<1\;\;\;\Rightarrow\;\;\;|u|_{p(x)}^{p^+}\leq
\rho_{p(x)}(u)\leq|u|_{p(x)}^{p^-}$$ $$\label{L6}
|u_n-u|_{p(x)}\rightarrow 0\;\;\;\Leftrightarrow\;\;\;\rho_{p(x)}
(u_n-u)\rightarrow 0.$$ Spaces with $p^+ =\infty$ have been studied by Edmunds, Lang and Nekvinda [@edm].
Next, we define $W_0^{1,p(x)}(\Omega)$ as the closure of $C_0^\infty(\Omega)$ under the norm $$\|u\|=|\nabla u|_{p(x)}.$$ The space $(W_0^{1,p(x)}(\Omega),\|\cdot\|)$ is a separable and reflexive Banach space. We note that if $q\in C_+(\overline\Omega)$ and $q(x)<p^\star(x)$ for all $x\in\overline\Omega$ then the embedding $W_0^{1,p(x)}(\Omega)\hookrightarrow L^{q(x)}(\Omega)$ is compact and continuous, where $p^\star(x)=\frac{Np(x)}{N-p(x)}$ if $p(x)<N$ or $p^\star(x)=+\infty$ if $p(x)\geq N$. We refer to [@edm2; @edm3; @FZ1; @KR] for further properties | important role in manipulating the generalized Lebesgue - Sobolev spaces is toy by the [ * modular * ] { } of the $ L^{p(x)}(\Omega)$ quad, which is the mapping $ \rho_{p(x)}:L^{p(x)}(\Omega)\rightarrow\RR$ defined by $ $ \rho_{p(x)}(u)=\int_\Omega|u|^{p(x)}\;dx.$$ If $ (u_n)$, $ u\in L^{p(x)}(\Omega)$ and $ p^+<\infty$ then the following sexual intercourse holds true $ $ \label{L4 }
|u|_{p(x)}>1\;\;\;\Rightarrow\;\;\;|u|_{p(x)}^{p^-}\leq\rho_{p(x)}(u)
\leq|u|_{p(x)}^{p^+}$$ $ $ \label{L5 }
|u|_{p(x)}<1\;\;\;\Rightarrow\;\;\;|u|_{p(x)}^{p^+}\leq
\rho_{p(x)}(u)\leq|u|_{p(x)}^{p^-}$$ $ $ \label{L6 }
|u_n - u|_{p(x)}\rightarrow 0\;\;\;\Leftrightarrow\;\;\;\rho_{p(x) }
(u_n - u)\rightarrow 0.$$ Spaces with $ p^+ = \infty$ have been learn by Edmunds, Lang and Nekvinda [ @edm ].
Next, we specify $ W_0^{1,p(x)}(\Omega)$ as the closure of $ C_0^\infty(\Omega)$ under the average $ $ \|u\|=|\nabla u|_{p(x)}.$$ The quad $ (W_0^{1,p(x)}(\Omega),\|\cdot\|)$ is a separable and automatic Banach space. We note that if $ q\in C_+(\overline\Omega)$ and $ q(x)<p^\star(x)$ for all $ x\in\overline\Omega$ then the embedding $ W_0^{1,p(x)}(\Omega)\hookrightarrow L^{q(x)}(\Omega)$ is compact and continuous, where $ p^\star(x)=\frac{Np(x)}{N - p(x)}$ if $ p(x)<N$ or $ p^\star(x)=+\infty$ if $ p(x)\geq N$. We denote to [ @edm2; @edm3; @FZ1; @KR ] for further properties | imoortant role in manipulauing the generalieee Lebevgue-Sogolev spxces is played by the [*modulac*]{} of the $O^{p(x)}(\Omega)$ space, which ir the mapiing $\rho_{p(z)}:L^{p(x)}(\Inega)\rightacdow\RR$ dcyined ny $$\rhm_{'(x)}(u)=\int_\Omega|u|^{p(x)}\;dw.$$ If $(u_n)$, $u\in L^{p(x)}(\Omega)$ and $[^+<\ivfcy$ then the following relations holdf true $$\kahel{L4}
|u|_{p(x)}>1\;\;\;\Rightartow\;\;\;|u|_{p(q)}^{p^-}\jeq\rgo_{p(x)}(u)
\leq|u|_{p(x)}^{p^+}$$ $$\label{L5}
|u|_{p(x)}<1\;\;\;\Rightarrow\;\;\;|u|_{p(s)}^{p^+}\leq
\rhm_{p(x)}(u)\leq|u|_{p(x)}^{p^-}$$ $$\lsbel{L6}
|u_n-u|_{p(x)}\rightarrow 0\;\;\;\Leftrlghtwrrow\;\;\;\rho_{p(x)}
(u_n-u)\rightwrrow 0.$$ Spacga wyrh $p^+ =\infty$ hxve been suubied by Edmonds, Lang and Nekvinda [@edm].
Next, wd defnne $W_0^{1,p(x)}(\Omegq)$ qs hve closure if $C_0^\igfty(\Omega)$ under the torm $$\|u\|=|\nsbla u|_{p(x)}.$$ The siace $(X_0^{1,p(x)}(\Onega),\|\cdot\|)$ is a separabne and reflexive Bwnach spawe. We note that if $w\ib C_+(\ovgrlina\Omeeq)$ avd $s(x)<'^\stzr(x)$ fog aml $x\in\overmine\Omega$ tyen the embedding $W_0^{1,k(x)}(\Oivba)\hookrightadrow L^{z(x)}(\Jmega)$ is compact and continuous, where $p^\vtad(x)=\frac{Np(x)}{N-p(x)}$ if $p(x)<N$ or $p^\star(x)=+\infty$ if $p(x)\geq J$. We refew to [@edm2; @edm3; @FZ1; @KR] for further properties | important role in manipulating the generalized Lebesgue-Sobolev played the [*modular*]{} the $L^{p(x)}(\Omega)$ space, defined $$\rho_{p(x)}(u)=\int_\Omega|u|^{p(x)}\;dx.$$ If $(u_n)$, L^{p(x)}(\Omega)$ and $p^+<\infty$ the following relations holds true $$\label{L4} \leq|u|_{p(x)}^{p^+}$$ $$\label{L5} |u|_{p(x)}<1\;\;\;\Rightarrow\;\;\;|u|_{p(x)}^{p^+}\leq \rho_{p(x)}(u)\leq|u|_{p(x)}^{p^-}$$ $$\label{L6} |u_n-u|_{p(x)}\rightarrow 0\;\;\;\Leftrightarrow\;\;\;\rho_{p(x)} (u_n-u)\rightarrow 0.$$ Spaces with $p^+ =\infty$ been studied by Edmunds, Lang and Nekvinda [@edm]. Next, we define $W_0^{1,p(x)}(\Omega)$ as closure $C_0^\infty(\Omega)$ the $$\|u\|=|\nabla u|_{p(x)}.$$ The space $(W_0^{1,p(x)}(\Omega),\|\cdot\|)$ is a separable and reflexive Banach space. We note that if C_+(\overline\Omega)$ and $q(x)<p^\star(x)$ for all $x\in\overline\Omega$ then the $W_0^{1,p(x)}(\Omega)\hookrightarrow L^{q(x)}(\Omega)$ is compact continuous, where $p^\star(x)=\frac{Np(x)}{N-p(x)}$ if $p(x)<N$ $p^\star(x)=+\infty$ $p(x)\geq N$. refer [@edm2; @FZ1; @KR] for properties | important role in manipulatiNg the generAlizeD LeBesGuE-SobOlev Spaces is played BY the [*Modular*]{} of the $L^{p(x)}(\Omega)$ spAce, whIcH Is thE MaPping $\Rho_{p(x)}:L^{p(X)}(\omEGA)\riGhTaRroW\Rr$ DeFined By $$\rHo_{p(x)}(u)=\inT_\Omega|u|^{p(x)}\;dX.$$ If $(U_n)$, $U\in L^{p(x)}(\Omega)$ aND $p^+<\Infty$ then tHe fOllowing relaTioNs holdS tRue $$\LAbel{L4}
|U|_{p(x)}>1\;\;\;\rightArrow\;\;\;|u|_{P(X)}^{p^-}\leq\rHo_{p(x)}(u)
\leq|u|_{P(x)}^{P^+}$$ $$\Label{L5}
|U|_{P(x)}<1\;\;\;\RightARRoW\;\;\;|u|_{p(x)}^{P^+}\leq
\rho_{p(x)}(u)\leq|u|_{p(x)}^{p^-}$$ $$\LAbEL{L6}
|u_n-u|_{p(x)}\rightarRow 0\;\;\;\LefTrIGhTARroW\;\;\;\rhO_{p(x)}
(u_n-u)\righTaRrow 0.$$ SPAces witH $P^+ =\iNFTY$ haVE been studied bY Edmunds, LanG And nekvinDa [@Edm].
nExt, we dEfine $w_0^{1,p(X)}(\omeGa)$ as the closUre oF $C_0^\infty(\OmEga)$ undER the norM $$\|U\|=|\nabla u|_{P(x)}.$$ The sPacE $(W_0^{1,p(X)}(\OmeGA),\|\cDoT\|)$ is A sEParABlE anD RefLexive BaNaCh Space. we noTE THAt if $Q\in c_+(\oveRline\omega)$ and $q(x)<p^\stAr(x)$ For aLL $x\iN\overLine\OMega$ ThEn the EmbeddIng $W_0^{1,p(X)}(\OMega)\hookrightarRow L^{Q(x)}(\Omega)$ is ComPaCt aNd ContiNUous, whEre $P^\stAr(x)=\frac{np(x)}{N-p(x)}$ iF $P(x)<N$ Or $P^\STAr(X)=+\infty$ if $p(x)\geq N$. We reFeR TO [@eDm2; @edm3; @FZ1; @Kr] for fuRThEr PRopertieS | important role in manipul ating thegener ali zed L ebes gue- Sobolev spaces is p layed by the [*modular *]{}of the$ L^ {p(x) }(\Omeg a )$ s pac e, w hic hi sthe m app ing $\r ho_{p(x)}: L^{ p( x)}(\Omega)\ r ig htarrow\RR $ d efined by $$ \rh o_{p(x )} (u) = \int_ \Om ega|u |^{p(x ) }\;dx. $$ If $(u _n ) $, $u\ i n L^{p( x ) }( \Ome ga)$ and $p^+<\in f ty $ then the foll owingre l at i o nshol ds true $$ \l abel{ L 4}
|u|_ { p( x ) } >1\ ; \;\;\Rightarr ow\;\;\;|u| _ {p( x)}^{p ^- }\l e q\rho_ {p(x) }( u )
\ leq|u|_{p(x )}^{ p^+}$$ $$ \label { L5}
|u| _ {p(x)}< 1\;\;\ ;\R igh tarr o w\ ;\ ;\; |u | _{p ( x) }^{ p ^+} \leq
\rh o_ {p (x)}( u)\l e q | u |_{p (x) }^{p ^-}$$ $$\label{L6}
|u _n-u | _{p (x)}\ right arro w0\;\; \;\Lef trigh ta rrow\;\;\;\rho_ {p(x )}
(u_n-u )\r ig hta rr ow 0. $ $ Spac eswit h $p^+=\infty $ ha ve b e en studied by Edmund s, L an g and Ne kvinda [@ ed m ].
Next ,wedefi n e $W_0 ^{1, p (x )}(\Omeg a)$ as th eclosure o f $C_0 ^\ inf ty( \Omeg a )$ u nder t he norm$$\|u \ |=|\nabla u|_{ p (x)}.$$ The s p ac e $( W _0^{ 1,p (x)}(\Omega ),\| \ cdot \|)$ is as epara ble a nd re f lexive Banach space .We not e tha t if $q\in C_ +(\overlin e \ O mega)$ a nd $ q (x ) <p^\star(x)$ f or al l $x\in\ov e rline\Om ega$then the embeddin g $W_0^{1, p(x )}( \Om ega ) \ ho okrightarrowL ^ {q(x )} (\Omega )$is comp act an d c ont in uous, whe re $p^\s ta r( x) =\ fra c{Np( x )}{N-p(x )} $ i f$p( x)<N$ or $p^ \star (x)= +\ in f ty$ if $p( x )\ g e q N$ .We ref erto [@ed m2;@ edm 3; @FZ1 ; @KR] fo r f u rthe rpr opertie s | important_role in_manipulating the generalized Lebesgue-Sobolev_spaces is_played_by the_[*modular*]{}_of the $L^{p(x)}(\Omega)$_space, which is_the mapping $\rho_{p(x)}:L^{p(x)}(\Omega)\rightarrow\RR$ defined_by $$\rho_{p(x)}(u)=\int_\Omega|u|^{p(x)}\;dx.$$ If_$(u_n)$,_$u\in L^{p(x)}(\Omega)$ and $p^+<\infty$ then the following relations holds true $$\label{L4}
|u|_{p(x)}>1\;\;\;\Rightarrow\;\;\;|u|_{p(x)}^{p^-}\leq\rho_{p(x)}(u)
\leq|u|_{p(x)}^{p^+}$$ $$\label{L5}
|u|_{p(x)}<1\;\;\;\Rightarrow\;\;\;|u|_{p(x)}^{p^+}\leq
\rho_{p(x)}(u)\leq|u|_{p(x)}^{p^-}$$ $$\label{L6}
|u_n-u|_{p(x)}\rightarrow 0\;\;\;\Leftrightarrow\;\;\;\rho_{p(x)}
(u_n-u)\rightarrow_0.$$_Spaces with_$p^+_=\infty$_have been studied by Edmunds,_Lang and Nekvinda [@edm].
Next, we_define $W_0^{1,p(x)}(\Omega)$_as the closure of $C_0^\infty(\Omega)$ under the norm_$$\|u\|=|\nabla_u|_{p(x)}.$$ The space_$(W_0^{1,p(x)}(\Omega),\|\cdot\|)$ is a separable and reflexive Banach space. We_note that if $q\in C_+(\overline\Omega)$ and_$q(x)<p^\star(x)$ for all_$x\in\overline\Omega$_then_the embedding $W_0^{1,p(x)}(\Omega)\hookrightarrow L^{q(x)}(\Omega)$_is compact and continuous, where $p^\star(x)=\frac{Np(x)}{N-p(x)}$_if $p(x)<N$ or $p^\star(x)=+\infty$ if $p(x)\geq_N$. We refer to [@edm2; @edm3; @FZ1;_@KR] for further properties |
energy density of the gas of light particles. This interpretation is in fact the usual one in quantum electrodynamics [@qed], where unsuppressed loop corrections to the total energy density $\rho$ (at temperatures well below the electron mass) are attributed to self-interactions among photons that arise after integrating the electron-positron field out of the functional integral. Recently, this point of view has been investigated in detail by Braaten and Jia [@bj] for the original MY model. They show that, after integrating out the heavy $\varphi$ field to get an effective theory of the light $\chi$ field alone, the $T^6/M^2$ term in $\rho_\varphi$ is obtained as the thermal expectation value of a particular higher-dimensional operator in the hamiltonian of the effective theory. However, the coefficient of this operator can be changed to any value (including zero) by a field redefinition. This shows that no physical significance can be attached to the $T^6/M^2$ term in $\rho_\varphi$.
We therefore conclude that the QBE for WIMP relic densities should not be corrected in the manner proposed by MY. However, we should keep in mind that the theoretical situation is not as clear as we might like it to be, and so be on the lookout for possible further surprises.
I thank Anupam Singh and Scott Thomas for discussions, and Motohiko Yoshimura for correspondence. I also thank Gary Steigman for bringing ref. [@bj] to my belated attention. This work was supported in part by the National Science Foundation through grant PHY–97–22022, and by the Institute of Geophysics and Planetary Physics through grant 920.
J. Rau and B. Müller, Phys. Rep. [**272**]{}, 1 (1996), nucl-th/9505009.
Sh. Matsumoto and M. Yoshimura, Phys. Rev. D [**59**]{}, 123511 (1999), hep-ph/9811301; Phys. Rev. D [**61**]{}, 123508 (2000), hep-ph/9910393; Phys. Rev. D [**61**]{}, 123509 (2000), hep-ph/9910425.
A. Singh and M. Srednicki, Phys. Rev. D [**61**]{}, 023509 (2000), hep-ph/9908224.
A. O. Cal | energy density of the gas of light particle. This rendition is in fact the usual one in quantum electrodynamics [ @qed ], where unsuppressed loop correction to the full energy density $ \rho$ (at temperature well below the electron mass) are assign to self - interactions among photon that rebel after integrating the electron - positron playing field out of the functional integral. Recently, this point of opinion has been investigated in detail by Braaten and Jia [ @bj ] for the original MY model. They show that, after integrating out the heavy $ \varphi$ battlefield to get an effective theory of the light $ \chi$ playing field alone, the $ T^6 / M^2 $ term in $ \rho_\varphi$ is obtained as the thermal expectation value of a especial higher - dimensional operator in the hamiltonian of the effective theory. However, the coefficient of this operator can be changed to any value (include zero) by a field redefinition. This shows that no physical significance can be attached to the $ T^6 / M^2 $ term in $ \rho_\varphi$.
We therefore conclude that the QBE for WIMP relic densities should not be corrected in the manner proposed by MY. However, we should keep in mind that the theoretical situation is not as clear as we might wish it to be, and thus be on the lookout for possible further surprise.
I thank Anupam Singh and Scott Thomas for discussions, and Motohiko Yoshimura for correspondence. I also thank Gary Steigman for bringing referee. [ @bj ] to my belated attention. This workplace was supported in part by the National Science Foundation through concession PHY–97–22022, and by the Institute of Geophysics and Planetary Physics through grant 920.
J. Rau and B. Müller, Phys. Rep. [ * * 272 * * ] { }, 1 (1996), nucl - th/9505009.
Sh. Matsumoto and M. Yoshimura, Phys. Rev. D [ * * 59 * * ] { }, 123511 (1999), hep - ph/9811301; Phys. Rev. D [ * * 61 * * ] { }, 123508 (2000), hep - ph/9910393; Phys. Rev. D [ * * 61 * * ] { }, 123509 (2000), hep - ph/9910425.
A. Singh and M. Srednicki, Phys. Rev. D [ * * 61 * * ] { }, 023509 (2000), hep - ph/9908224.
A. O. Cal | enfrgy density of the gas uf light particles. Thiv intedpretatiun is in fact the usual one mn qyantun electrodynamics [@qed], dhere unslppressed look corrections to vge total enerfn denvmty $\rho$ (at tempgratures weln below the elacgrln mass) are attributed to self-interwctions along photons trat swise after integrating the electron-posjtron fpeld out of the finctional integral. Recentlj, thls point of view hws been invgatidqted in detakl by Braaueu and Jia [@bn] for the original MY model. Theh shoc that, aftet jnhggrating out the reavy $\varphi$ field tm get am effective thcory mf rhe light $\chi$ field anone, the $T^6/M^2$ term ig $\rho_\varpvi$ is obtained as tye therkal axpezratkon velus of a pacticular hifher-dimensiinal operator in tht hwniltonian of fhe efsestive theory. However, the coefficient of thjs operator can be chanted to any value (incloding zero) by a field redefinition. This shows that no physiwal smgvifngwvxe can be attached to the $T^6/M^2$ term in $\rho_\varphi$.
Wq tnegefore conclude tmat the QBE for WIKP rrjic densities shoulb nkt be corrected in the magner proposed fy MU. However, we should keep in mind that tke rheoretical situatnon is not ar clgar as we might like it to be, and so be on tje lookouf for possible fufthvr sgrprises.
I thank Anupam Sindh and Scitt Chomas fur doscussyons, and Mltohiko Yoshimura for cogresplngence. I aldo thank Gary Steigman for brinjmng ref. [@bj] to ky benated atcentiok. This work was supported in kart by tke Natkonal Sciehce Fouidation throtgh grant PHY–97–22022, wnd by the Iistitute jf Gwophtsics ava Planetary Phusics through grant 920.
J. Rau and B. Müller, Phis. Rep. [**272**]{}, 1 (1996), nucl-th/9505009.
Sk. Natsumoto and M. Yorhiiuga, 'hys. Wav. D [**59**]{}, 123511 (1999), hep-pv/9811301; Phhs. Frv. D [**61**]{}, 123508 (2000), hep-ph/9910393; Phns. Fev. C [**61**]{}, 123509 (2000), hep-ph/9910425.
A. Singh ang M. Arednicki, Phys. Rev. D [**61**]{}, 023509 (2000), hep-py/9908224.
A. O. Cal | energy density of the gas of light interpretation in fact usual one in loop to the total density $\rho$ (at well below the electron mass) are to self-interactions among photons that arise after integrating the electron-positron field out of functional integral. Recently, this point of view has been investigated in detail by and [@bj] the MY model. They show that, after integrating out the heavy $\varphi$ field to get an effective of the light $\chi$ field alone, the $T^6/M^2$ in $\rho_\varphi$ is obtained the thermal expectation value of particular operator in hamiltonian the theory. However, the of this operator can be changed to any value (including zero) by a field redefinition. This shows no physical be attached the term $\rho_\varphi$. We therefore the QBE for WIMP relic densities corrected in the manner proposed by MY. However, should keep mind that the theoretical situation is as clear as we might like it to and so be on the lookout for possible further surprises. I thank Anupam Singh and for discussions, and Motohiko for correspondence. I thank Steigman bringing [@bj] to belated attention. This work was supported in part by the National Foundation through grant PHY–97–22022, and by the Institute of Geophysics Physics grant 920. J. and B. Müller, Phys. [**272**]{}, (1996), nucl-th/9505009. Sh. Matsumoto Yoshimura, Rev. (1999), Phys. D [**61**]{}, 123508 (2000), Phys. Rev. D [**61**]{}, 123509 hep-ph/9910425. A. Singh and [**61**]{}, 023509 (2000), hep-ph/9908224. A. O. Cal | energy density of the gas of liGht particlEs. ThiS inTerPrEtatIon iS in fact the usuaL One iN quantum electrodynamicS [@qed], wHeRE unsUPpResseD loop coRReCTIonS tO tHe tOtAL eNergy DenSity $\rho$ (At temperatUreS wEll below the eLEcTron mass) arE atTributed to seLf-iNteracTiOns AMong pHotOns thAt arisE After iNtegratinG tHE electROn-positRON fIeld Out of the functionaL InTEgral. Recently, tHis poiNt OF vIEW haS beEn investigAtEd in dETail by BRAaTEN And jIa [@bj] for the oriGinal MY modeL. theY show tHaT, afTEr inteGratiNg OUt tHe heavy $\varpHi$ fiEld to get aN effecTIve theoRY of the lIght $\chI$ fiEld AlonE, ThE $T^6/m^2$ teRm IN $\rhO_\VaRphI$ Is oBtained aS tHe ThermAl exPECTAtioN vaLue oF a parTicular higher-DimEnsiONal OperaTor in The hAmIltonIan of tHe effEcTive theory. HowevEr, thE coefficiEnt Of ThiS oPeratOR can be ChaNgeD to any vAlue (incLUdiNg ZERO) bY a field redefinitioN. THIS sHows that No physICaL sIGnificanCe Can Be atTAChed tO the $t^6/m^2$ tErm in $\rho_\Varphi$.
wE tHeRefore cOnClude tHaT thE QBe for WimP reLic denSities shOuld nOT be corrected in THe manner propoSEd BY mY. hOwevEr, wE should keep In miND thaT the THeOreTIcal sItuatIoN Is NOt as clear as we might lIkE it to bE, and sO be on the lookoUt for possiBLE Further sUrprISeS.
i thank Anupam SiNgh anD Scott ThomAS for discUssioNs, and MotOhiko YoshIMUra for coRreSpoNdeNce. i ALsO thank Gary SteIGMan fOr BringinG reF. [@bj] to my BelAteD atTenTiOn. This worK was suppOrTeD iN pArt By the nAtional SCiEncE FOunDatioN ThrougH granT PHY–97–22022, AnD bY The instituTE oF gEophYsIcS and plaNeTary PHysiCS thRough grAnt 920.
J. Rau anD B. MÜLler, phYs. rep. [**272**]{}, 1 (1996), nucl-Th/9505009.
Sh. Matsumoto AnD M. YoshimurA, PHys. rev. D [**59**]{}, 123511 (1999), heP-PH/9811301; Phys. Rev. d [**61**]{}, 123508 (2000), hep-ph/9910393; Phys. Rev. D [**61**]{}, 123509 (2000), hep-ph/9910425.
A. SinGH and M. SrEdnIcki, PHys. REv. D [**61**]{}, 023509 (2000), hep-ph/9908224.
A. o. CaL | energy density of the gas of lightparti cle s.Th is i nter pretation is i n fac t the usual one in qua ntumel e ctro d yn amics [@qed] , w h e reun su ppr es s ed loop co rrectio ns to thetot al energy dens i ty $\rho$ (a t t emperatureswel l belo wthe elect ron mass ) area ttribu ted to se lf - intera c tions a m o ng pho tons that arise a f te r integrating t he ele ct r on - p osi tro n field ou tof th e functi o na l i nte g ral. Recently , this poin t of viewha s b e en inv estig at e d i n detail by Bra aten andJia [@ b j] fort he orig inal M Y m ode l. T h ey s how t h at, af ter int egrating o ut theheav y $ \ varp hi$ fie ld to get an effec tiv e th e ory of t he li ght$\ chi$fieldalone ,the $T^6/M^2$ t ermin $\rho_ \va rp hi$ i s obt a ined a s t hethermal expect a tio nv a l ue of a particular h ig h e r- dimensio nal op e ra to r in theha mil toni a n of t he e f fe ctive th eory.H ow ev er, the c oeffic ie ntofthiso pera tor ca n be cha ngedt o any value (i n cluding zero) by a f i eldred efinition.This show s th a tnop hysic al si gn i fi c ance can be attache dto the $T^6 /M^2$ term in $\rho_\va r p h i$.
Wether e fo r e conclude tha t the QBE for W I MP relic dens ities sh ould notb e correct edinthe ma n n er proposed byM Y . Ho we ver, we sh ould ke epinmin d t ha t the the oretical s it ua ti onis no t as clea raswe mi ght l i ke itto be , an dso beon thel oo k o ut f or p ossi ble f urthe r su r pri ses.
I thank An upa m Sin gh a nd Scot t Thomas fordi scussions, a ndMotohi k o Yoshimu ra for correspondence.I also t han k Gar y St eigman fo r b ringin g r e f. [@b j] tomy be la ted a ttent i o n. Th is work wass u ppo rtedin par t by th e National Science Fou ndation throu ghgran t PH Y–9 7 –2 2 022 ,a ndb y the Instituteof Geophys ic s a nd Planeta r y P hy sics th rough g rant9 20.
J. Rau andB. Müller ,Phys . Rep . [**272** ]{}, 1 ( 1996), nu c l-th/ 9 50 5009.
S h. Mat su mot o and M. Yo s him ura,Phys.Re v. D [ **59* *] {}, 1235 11 (1999), hep-ph/98113 01; Ph ys. R ev. D [**61* *]{ } , 1 23508 (20 00), hep-ph/99 103 93; Phys . R e v. D[**6 1 ** ]{} , 1235 09 ( 2 000), hep - ph /99 1 0 42 5.
A. Sing h a ndM. Sr edn i cki, P hys. Rev. D [**61**]{ } , 023509 (2000 ), h e p -ph /99 0 8224 .
A. O. Cal | energy_density of_the gas of light_particles. This_interpretation_is in_fact_the usual one_in quantum electrodynamics_[@qed], where unsuppressed loop_corrections to the_total_energy density $\rho$ (at temperatures well below the electron mass) are attributed to self-interactions_among_photons that_arise_after_integrating the electron-positron field out_of the functional integral. Recently,_this point_of view has been investigated in detail by_Braaten_and Jia [@bj]_for the original MY model. They show that, after_integrating out the heavy $\varphi$ field_to get an_effective_theory_of the light $\chi$_field alone, the $T^6/M^2$ term in_$\rho_\varphi$ is obtained as the thermal_expectation value of a particular higher-dimensional operator_in the hamiltonian of the effective_theory. However, the coefficient of_this operator_can be changed to any_value (including zero)_by a_field redefinition. This_shows that no physical significance can_be attached to_the $T^6/M^2$ term in $\rho_\varphi$.
We therefore_conclude_that the QBE_for_WIMP_relic densities_should not be_corrected_in the_manner_proposed by MY. However, we should_keep_in mind that the theoretical situation is_not as clear as_we_might like it to_be, and so be on_the lookout for possible further surprises.
I_thank Anupam_Singh and_Scott Thomas for discussions, and Motohiko Yoshimura for correspondence. I also_thank Gary Steigman for bringing ref. [@bj]_to my belated attention._This work_was_supported in part_by_the National_Science Foundation through grant PHY–97–22022, and by_the Institute_of Geophysics and Planetary Physics through_grant 920.
J. Rau and_B._Müller, Phys. Rep. [**272**]{}, 1 (1996),_nucl-th/9505009.
Sh. Matsumoto and M. Yoshimura, Phys._Rev. D [**59**]{}, 123511 (1999),_hep-ph/9811301;_Phys._Rev. D [**61**]{}, 123508 (2000),_hep-ph/9910393; Phys. Rev. D [**61**]{}, 123509_(2000), hep-ph/9910425.
A. Singh_and M. Srednicki, Phys. Rev. D [**61**]{},_023509_(2000), hep-ph/9908224.
A. O. Cal |
J. P. 1008, AJ 115, 2356 Dolphin, A. E. 1997, New Astronomy 2, 397 Dolphin, A. E. 1999, in prep. Durrell, P. R., Harris, W. E., Geisler, D., & Pudritz, R. E. 1996, AJ 112, 972 Fagotto, F., Bressan, A., Bertelli, G., & Chiosi, C. 1994, A&AS 105, 29 Feast, M. W. & Walker, A. R. 1987, 25, 345 Ferraro, F. R., Pecci, F. F., Tosi, M., & Buonanno, R. 1989, MNRAS 241, 433 Girardi, L., Bressan, A., Chiosi, C., Bertelli, G., & Nasi, E. 1996, A&AS 117, 113 Harris, W. E. 1991, Ann. Rev. Astron. Astrophys. 29, 543 Harris, W. E. & Pudritz 1994, ApJ 429, 177 Hodge, P. & Miller, B. W. 1995, ApJ 451, 176 Hodge, P., Skelton, B., & Ashizawa, J., *An Atlas of Local Group Galaxies*, Dordrecht: Kluwer Holtzman, J. A., Burrows, C. J., Casertano, S., Hester, J. J., Trauger, J. T., Watson, A. M., & Worthey, G. 1995, PASP 107, 1065 Humason, M. L., Mayall, N. U., & Sandage, A. R. 1956, AJ 61, 97 Johnson, J. A. Bolte, M.. Bond, H., Hesser, J. E., De Oliveira, C. M., Richer, H. B., Stetson, P. B., & vandenBerg, D. A. 1998, IAUS 190, 154 King, I. R. 1962, AJ 71, 64 Krist, J. 1995, Astronomical Data Analysis Software and Systems IV, 77, 349 Lee, M. G., Freedman, W. L., & Madore, B. F. 1993, | J. P. 1008, AJ 115, 2356 Dolphin, A. E. 1997, New Astronomy 2, 397 Dolphin, A. E. 1999, in prep. Durrell, P. R., Harris, W. E., Geisler, D., & Pudritz, R. E. 1996, AJ 112, 972 Fagotto, F., Bressan, A., Bertelli, G., & Chiosi, C. 1994, A&AS 105, 29 Feast, M. W. & Walker, A. R. 1987, 25, 345 Ferraro, F. R., Pecci, F. F., Tosi, M., & Buonanno, R. 1989, MNRAS 241, 433 Girardi, L., Bressan, A., Chiosi, C., Bertelli, G., & Nasi, E. 1996, A&AS 117, 113 Harris, W. E. 1991, Ann. Rev. Astron. Astrophys. 29, 543 Harris, W. E. & Pudritz 1994, ApJ 429, 177 Hodge, P. & Miller, B. W. 1995, ApJ 451, 176 Hodge, P., Skelton, B., & Ashizawa, J., * An Atlas of Local Group Galaxies *, Dordrecht: Kluwer Holtzman, J. A., Burrows, C. J., Casertano, S., Hester, J. J., Trauger, J. T., Watson, A. M., & Worthey, G. 1995, PASP 107, 1065 Humason, M. L., Mayall, N. U., & Sandage, A. R. 1956, AJ 61, 97 Johnson, J. A. Bolte, M.. Bond, H., Hesser, J. E., De Oliveira, C. M., Richer, H. B., Stetson, P. B., & vandenBerg, D. A. 1998, IAUS 190, 154 King, I. R. 1962, AJ 71, 64 Krist, J. 1995, Astronomical Data Analysis Software and Systems IV, 77, 349 Lee, M. G., Freedman, W. L., & Madore, B. F. 1993, | J. P. 1008, AJ 115, 2356 Dolphin, A. E. 1997, New Artronomy 2, 397 Dolpkun, A. E. 1999, in prsp. Durreul, P. R., Harris, W. E., Geisler, D., & Pndrirz, R. E. 1996, AJ 112, 972 Fagotto, F., Bressxn, A., Bertvlli, G., & Cyiosm, C. 1994, A&AS 105, 29 Feast, M. W. & Walker, A. R. 1987, 25, 345 Fzrcaro, F. R., Pecci, F. N., Tosi, M., & Bgonanno, R. 1989, MNRDS 241, 433 Girardi, L., Bressan, A., Chiosi, C., Bertejli, G., & Madi, E. 1996, A&AS 117, 113 Hatris, E. Q. 1991, Ahn. Rev. Astron. Astrophys. 29, 543 Harris, W. S. & Pudrptz 1994, ApJ 429, 177 Hodge, L. & Miller, B. W. 1995, ApJ 451, 176 Hodge, O., Skflton, B., & Ashizawa, U., *An Atlas if Ljxal Group Gauaxies*, Dordrecht: Kluwet Holtzman, J. A., Burrows, C. J., Casertavo, S., Kester, J. J., Ttcygeg, J. T., Watson, E. M., & Wogthey, G. 1995, PASP 107, 1065 Humasmn, M. L., Msyall, N. U., & Sandsge, A. R. 1956, AJ 61, 97 Johnson, J. A. Bolve, M.. Bond, H., Hesser, J. G., De Olivehrc, C. M., Richer, H. B., Stetsob, P. B., & vdndetBere, D. A. 1998, IZUX 190, 154 King, I. R. 1962, AJ 71, 64 Kriat, J. 1995, Astrobomical Data Analysos Woftware and Aystemf YV, 77, 349 Lee, M. G., Freedman, W. L., & Madore, B. F. 1993, | J. P. 1008, AJ 115, 2356 Dolphin, 1997, Astronomy 2, Dolphin, A. E. R., W. E., Geisler, & Pudritz, R. 1996, AJ 112, 972 Fagotto, F., A., Bertelli, G., & Chiosi, C. 1994, A&AS 105, 29 Feast, M. W. Walker, A. R. 1987, 25, 345 Ferraro, F. R., Pecci, F. F., Tosi, & R. MNRAS 433 Girardi, L., Bressan, A., Chiosi, C., Bertelli, G., & Nasi, E. 1996, A&AS 117, 113 W. E. 1991, Ann. Rev. Astron. Astrophys. 29, Harris, W. E. & 1994, ApJ 429, 177 Hodge, & B. W. ApJ 176 P., Skelton, B., Ashizawa, J., *An Atlas of Local Group Galaxies*, Dordrecht: Kluwer Holtzman, J. A., Burrows, C. J., Casertano, Hester, J. J. T., A. & G. 1995, PASP Humason, M. L., Mayall, N. U., R. 1956, AJ 61, 97 Johnson, J. A. M.. Bond, Hesser, J. E., De Oliveira, C. Richer, H. B., Stetson, P. B., & vandenBerg, A. 1998, IAUS 190, 154 King, I. R. 1962, AJ 71, 64 Krist, J. 1995, Analysis Software and Systems 77, 349 Lee, G., W. & B. F. | J. P. 1008, AJ 115, 2356 Dolphin, A. E. 1997, New Astronomy 2, 397 dolphin, A. E. 1999, iN prep. durRelL, P. r., HarRis, W. e., Geisler, D., & PudriTZ, R. E. 1996, Aj 112, 972 Fagotto, F., Bressan, A., BerteLli, G., & CHiOSi, C. 1994, A&as 105, 29 FEast, M. w. & Walker, a. r. 1987, 25, 345 FERRarO, F. r., PEccI, F. f., toSi, M., & BuOnaNno, R. 1989, MNRaS 241, 433 Girardi, L., breSsAn, A., Chiosi, C., BeRTeLli, G., & Nasi, E. 1996, A&aS 117, 113 HArris, W. E. 1991, Ann. ReV. AsTron. AsTrOphYS. 29, 543 HarrIs, W. e. & PudrItz 1994, ApJ 429, 177 hOdge, P. & MIller, B. W. 1995, Apj 451, 176 HODge, P., SkELton, B., & AsHIZaWa, J., *AN Atlas of Local GrouP gaLAxies*, Dordrecht: kluwer hoLTzMAN, J. A., burRows, C. J., CaseRtAno, S., HESter, J. J., TRAuGER, j. T., WATson, A. M., & Worthey, g. 1995, PASP 107, 1065 HumasoN, m. L., MAyall, N. u., & SAndAGe, A. R. 1956, AJ 61, 97 johnsOn, j. a. BoLte, M.. Bond, H., HeSser, j. E., De OliveIra, C. M., RICher, H. B., STEtson, P. B., & VandenberG, D. A. 1998, iAUS 190, 154 kInG, I. r. 1962, AJ 71, 64 krISt, J. 1995, aStRonOMicAl Data AnAlYsIs SofTwarE AND systEms iV, 77, 349 LeE, M. G., FrEedman, W. L., & MadorE, B. F. 1993, | J. P. 1008, AJ 115, 2356Dolphin, A . E.199 7,Ne w As tron omy 2, 397 Dol p hin, A. E. 1999, in prep.Durre ll , P.R ., Harr is, W.E ., G eis le r, D. ,& P udrit z,R. E. 1 996, AJ 11 2,97 2 Fagotto, F . ,Bressan, A .,Bertelli, G. , & Chios i, C. 1994, A& AS 10 5, 29F east,M. W. & W al k er, A. R. 1987 , 25 , 34 5 Ferraro, F. R., Pe c ci, F. F., Tos i, M., & Bu o n ann o,R. 1989, M NR AS 24 1 , 433 G i ra r d i , L . , Bressan, A. , Chiosi, C . , B ertell i, G. , & Nas i, E. 1 9 96, A&AS 117,113Harris, W . E. 1 9 91, Ann . Rev. A stron. As tro phys . 2 9, 54 3H arr i s, W. E.& Pudrit z19 94, A pJ 4 2 9 , 177Hod ge,P. &Miller, B. W. 19 95,A pJ451,176 H odge ,P., S kelton , B., & Ashizawa, J.,*AnAtlas ofLoc al Gr ou p Gal a xies*, Do rdr echt: K luwer H o ltz ma n , J. A., Burrows, C. J ., C as ertano,S., He s te r, J. J., T ra uge r, J . T., W atso n ,A. M., & Worth e y, G . 1995, P ASP 10 7, 10 65Humas o n, M . L.,Mayall,N. U. , & Sandage, A. R. 1956, AJ 6 1 ,9 7 J o hnso n,J. A. Bolte , M. . Bon d, H . ,Hes s er, J . E., D e O l iveira, C. M., Rich er , H. B ., St etson, P. B., & vandenB e r g , D. A.1998 , I A US 190, 154 Ki ng, I . R. 1962, AJ 71, 6 4 Kri st, J. 1 995, Astr o n omical D ata An aly sis S of tware and Sys t e ms I V, 77, 34 9 L ee, M.G., Fr eed man ,W. L., &Madore,B. F .19 93, | J. P._1008, AJ_115, 2356 Dolphin, A. E._1997, New_Astronomy_2, 397_Dolphin,_A. E. 1999, in_prep. Durrell, P. R.,_Harris, W. E., Geisler, D.,_& Pudritz, R. E._1996,_AJ 112, 972 Fagotto, F., Bressan, A., Bertelli, G., & Chiosi, C. 1994, A&AS_105,_29 Feast,_M. W._&_Walker, A. R. 1987, 25, 345_Ferraro, F. R., Pecci, F. F., Tosi,_M., &_Buonanno, R. 1989, MNRAS 241, 433 Girardi, L.,_Bressan,_A., Chiosi, C.,_Bertelli, G., & Nasi, E. 1996, A&AS 117, 113_Harris, W. E. 1991, Ann. Rev. Astron._Astrophys. 29, 543_Harris,_W. E._& Pudritz 1994, ApJ_429, 177 Hodge, P. & Miller,_B. W. 1995, ApJ 451, 176 Hodge,_P., Skelton, B., & Ashizawa, J., *An_Atlas of Local Group Galaxies*, Dordrecht:_Kluwer Holtzman, J. A., Burrows, C. J.,_Casertano, S.,_Hester, J. J., Trauger, J. T., Watson,_A. M., & Worthey,_G. 1995,_PASP 107, 1065_Humason, M. L., Mayall, N. U., & Sandage,_A. R. 1956, AJ_61, 97 Johnson, J. A. Bolte, M.._Bond,_H., Hesser, J. E.,_De_Oliveira,_C. M., Richer,_H. B., Stetson, P. B.,_&_vandenBerg, D. A._1998,_IAUS 190, 154 King, I. R. 1962,_AJ_71, 64 Krist, J. 1995, Astronomical Data_Analysis Software and Systems_IV,_77, 349 Lee, M. G.,_Freedman, W. L., & Madore, B. F._1993, |
therefore accurately reproducing the thermal and ionization history of the IGM is essential for improving the robustness of all Ly$\alpha$-based studies. The IGM probed by the Ly$\alpha$ forest, highly ionized at high-$z$ and progressively neutral with decreasing redshift, consists of mildly nonlinear gas density fluctuations; the gas traces the DM distribution, and it is generally assumed to be in photoionization equilibrium with the UV background produced by galaxies and quasars. In low-density regions, the IGM gas density ($\rho$) and temperature ($T$) are closely connected, via a redshift-dependent polytropic power-law relation. In simple form, the relation can be expressed as a function of redshift as: $$\log T(z) = \log T_0(z) +[\gamma(z) -1] \log \delta(z),
\label{eq_IGM_temperature_density_relation}$$ where $T_0$ is the gas mean temperature, $\delta = \rho/ \rho_{\rm c}$, $\rho_{\rm c}$ is the critical density, and $\gamma$ is a redshift-dependent parameter, which is also related to the reionization history model and spectral shape of the UV background.
To this end, Figure \[fig\_igm\_properties\_1\] compares the gas mean temperature $T_0$ as inferred from particle subsamples extracted from the BG at different redshifts ([*Grid Suite*]{}, Table \[table\_grid\_sims\_base\]) via computation of the relevant IGM quantities that enter in Equation (\[eq\_IGM\_temperature\_density\_relation\]) using standard SPH techniques, with the well-known measurement by Becker et al. (2011) of the IGM temperature at the mean density when $\gamma = 1.3$. This corresponds to a mild flattening of the $T-\rho$ relation as expected during an extended He II reionization process. In the figure, black filled dots are results from the BG, while empty blue dots with error bars refer to the measurements performed by Becker et al. (2011). Note that, for our baseline cosmology, we set $\gamma(z=3)=1.3$ and $T_0 (z=3)=15000~{\rm K}$ – see Table \[table\_baseline\_params\_sims\]. Moreover, $T_0$ is largely insensitive to $\gamma$ at $z > 4$, as highlighted by the same authors. The main point of the figure is | therefore accurately reproducing the thermal and ionization history of the IGM is essential for improving the robustness of all Ly$\alpha$-based report. The IGM probe by the Ly$\alpha$ forest, highly ionized at high-$z$ and progressively neutral with decrease redshift, consists of gently nonlinear gas density fluctuations; the gasoline traces the DM distribution, and it is generally assume to be in photoionization equilibrium with the UV background produced by galaxies and quasars. In abject - density regions, the IGM gas concentration ($ \rho$) and temperature ($ T$) are closely connected, via a red shift - dependent polytropic baron - law relative. In simple form, the relation can be express as a function of redshift as: $ $ \log T(z) = \log T_0(z) + [ \gamma(z) -1 ] \log \delta(z),
\label{eq_IGM_temperature_density_relation}$$ where $ T_0 $ is the gas mean temperature, $ \delta = \rho/ \rho_{\rm c}$, $ \rho_{\rm c}$ is the critical density, and $ \gamma$ is a redshift - pendent parameter, which is also related to the reionization history model and spectral shape of the UV background.
To this end, Figure \[fig\_igm\_properties\_1\ ] compares the gasoline mean temperature $ T_0 $ as inferred from particle subsamples extract from the BG at unlike red shift ([ * Grid Suite * ] { }, Table \[table\_grid\_sims\_base\ ]) via computation of the relevant IGM quantities that enter in Equation (\[eq\_IGM\_temperature\_density\_relation\ ]) using standard SPH techniques, with the well - known measurement by Becker et al. (2011) of the IGM temperature at the mean concentration when $ \gamma = 1.3$. This corresponds to a mild flattening of the $ T-\rho$ relation as expected during an extended He II reionization process. In the figure, black filled dots are results from the BG, while empty blue dot with error browning automatic rifle refer to the measurements performed by Becker et al. (2011). Note that, for our baseline cosmology, we jell $ \gamma(z=3)=1.3 $ and $ T_0 (z=3)=15000~{\rm K}$ – see Table \[table\_baseline\_params\_sims\ ]. Moreover, $ T_0 $ is largely insensitive to $ \gamma$ at $ z > 4 $, as highlighted by the like authors. The independent point of the figure is | thfrefore accurately reproaucing the thermal and ionizztion hirtory of the IGM is essentiap dor inproving the robustnesr of all Py$\alpha$-bqsed wtudies. Thx IGM probed bg the Oy$\alpha$ forest, highly iotized at high-$z$ avd progressively neutral with decreasyng redxhlft, consists os mikqly honlinear gas density fluctuations; the gav traces the CM distribution, and it is henegally assumed to bf in photoiinizwrion equilibfium with uhz UV backgrkund produced by galaxies and qjasarx. In low-debsuty tegions, the MGM gaf density ($\rho$) and tekperatute ($T$) are closeky ronnwcted, via a redshift-dxpendent polytropic kower-law ralction. In simple form, rhw reldtiot cav be exkrexssd as w fnnction of dedshift as: $$\log T(z) = \log T_0(z) +[\gamka(s) -1] \log \delta(z),
\mabel{ez_IDM_temperature_density_relation}$$ where $T_0$ is ths gas mean temperature, $\eelta = \rho/ \rho_{\rm c}$, $\rhl_{\rm c}$ is ehe critical density, and $\gamma$ is a redshift-depengent 'afamttcv, whkxh is also related to the reionization history ikdtl snd spectral smape of the UV baclggoigd.
To this end, Figurz \[fjg\_igm\_properties\_1\] colpares jhe gaw mean teipersture $T_0$ as inferred from paeticle subsaiples extracted frol the BG at difgereny redshifts ([*Grid Suite*]{}, Cable \[fable\_grid\_sils\_base\]) viz computation of ghe ralevant IGM quantities thae enter ii Equction (\[eq\_KGM\_tgmperattre\_density\_gelatlmn\]) using standard DPH tgchniqges, with tje well-known measurement by Becker et al. (2011) of jhe IGK tempercture st the mean dqnsity when $\gakma = 1.3$. Tkis cofresponds no a mild flattening of the $T-\rho$ sglation as ex'ected duwing an wxtendea He II reionizstion probefs. Ib the figure, black fiumed dots are rewulus from the BG, wnild eipny ulue qmts with errmr bxrs tefer to the meawjremrnts performed by Bewker et al. (2011). Note that, gov our basgline cosiology, we set $\gamma(z=3)=1.3$ and $T_0 (z=3)=15000~{\rm N}$ – sex Tablx \[tablr\_bafeline\_params\_sims\]. Moreover, $T_0$ is largely lnscnsitive to $\gwmma$ at $z > 4$, as hnghlighted by the same authors. The main 'oint of the figure is | therefore accurately reproducing the thermal and ionization the is essential improving the robustness IGM by the Ly$\alpha$ highly ionized at and progressively neutral with decreasing redshift, of mildly nonlinear gas density fluctuations; the gas traces the DM distribution, and is generally assumed to be in photoionization equilibrium with the UV background produced galaxies quasars. low-density the IGM gas density ($\rho$) and temperature ($T$) are closely connected, via a redshift-dependent polytropic power-law In simple form, the relation can be expressed a function of redshift $$\log T(z) = \log T_0(z) -1] \delta(z), \label{eq_IGM_temperature_density_relation}$$ $T_0$ the mean temperature, $\delta \rho/ \rho_{\rm c}$, $\rho_{\rm c}$ is the critical density, and $\gamma$ is a redshift-dependent parameter, which is related to history model spectral of UV background. To Figure \[fig\_igm\_properties\_1\] compares the gas mean inferred from particle subsamples extracted from the BG different redshifts Suite*]{}, Table \[table\_grid\_sims\_base\]) via computation of relevant IGM quantities that enter in Equation (\[eq\_IGM\_temperature\_density\_relation\]) standard SPH techniques, with the well-known measurement by Becker et al. (2011) of the IGM the mean density when = 1.3$. This to mild of $T-\rho$ relation expected during an extended He II reionization process. In the figure, filled dots are results from the BG, while empty blue error refer to the performed by Becker et (2011). that, for our baseline set and – Table Moreover, $T_0$ is largely to $\gamma$ at $z > as highlighted by the of the figure is | therefore accurately reprodUcing the thErmal And IonIzAtioN hisTory of the IGM is ESsenTial for improving the robUstneSs OF all lY$\aLpha$-bAsed stuDIeS. tHe IgM PrObeD bY ThE Ly$\alPha$ Forest, hIghly ionizEd aT hIgh-$z$ and progrESsIvely neutrAl wIth decreasinG reDshift, CoNsiSTs of mIldLy nonLinear GAs densIty fluctuAtIOns; the GAs traceS THe dM diStribution, and it is GEnERally assumed to Be in phOtOIoNIZatIon EquilibriuM wIth thE uV backgROuND PRodUCed by galaxies And quasars. IN Low-DensitY rEgiONs, the IgM gas DeNSitY ($\rho$) and tempEratUre ($T$) are clOsely cONnected, VIa a redsHift-dePenDenT polYTrOpIc pOwER-laW ReLatIOn. IN simple fOrM, tHe relAtioN CAN Be exPreSsed As a fuNction of redshIft As: $$\loG t(z) = \lOg T_0(z) +[\gAmma(z) -1] \Log \dElTa(z),
\laBel{eq_IgM_temPeRature_density_reLatiOn}$$ where $T_0$ iS thE gAs mEaN tempERature, $\DelTa = \rHo/ \rho_{\rm C}$, $\rho_{\rm c}$ IS thE cRITIcAl density, and $\gamma$ iS a REDsHift-depeNdent pARaMeTEr, which iS aLso RelaTED to thE reiONiZation hiStory mODeL aNd spectRaL shape Of The uV bAckgrOUnd.
TO this eNd, Figure \[Fig\_igM\_Properties\_1\] compARes the gas mean TEmPERaTUre $T_0$ As iNferred from PartICle sUbsaMPlEs eXTractEd froM tHE Bg At different redshiftS ([*GRid SuiTe*]{}, TabLe \[table\_grid\_siMs\_base\]) via cOMPUtation oF the RElEVant IGM quantitIes thAt enter in EQUation (\[eq\_iGM\_teMperaturE\_density\_rELAtion\]) usiNg sTanDarD SPh TEcHniques, with thE WEll-kNoWn measuRemEnt by BeCkeR et Al. (2011) oF thE IgM temperaTure at thE mEaN dEnSitY when $\GAmma = 1.3$. This CoRreSpOndS to a mILd flatTeninG of tHe $t-\rHO$ reLation aS ExPECted DuRiNg an ExtEnDed He iI reIOniZation pRocess. In tHe fIGure, BlAcK filled Dots are resultS fRom the BG, whIlE emPty bluE DOts with eRror bars refer to the measuREments pErfOrmed By BeCker et al. (2011). NOte That, foR ouR BaseliNe cosmOlogy, We Set $\GAMma(z=3)=1.3$ aND $t_0 (z=3)=15000~{\Rm K}$ – SeE Table \[tablE\_BAseLine\_pArAms\_sIms\]. MoreOver, $T_0$ is largely inseNSitIve to $\gamma$ at $z > 4$, As hIghlIGHtEd bY ThE SamE aUThoRS. the main point of tHe figure is | therefore accurately repr oducing th e the rma l a nd ion izat ion history of theIGM is essential for i mprov in g the ro bustn ess ofa ll L y$\ al ph a$- ba s ed stud ies . The I GM probedbyth e Ly$\alpha$ fo rest, high lyionized at h igh -$z$ a nd pr o gress ive ly ne utralw ith de creasingre d shift, consist s of mil dly nonlinear gas de n sity fluctuati ons; t he ga s tra ces the DM di st ribut i on, and it i s ge n erally assume d to be inp hot oioniz at ion equili brium w i ththe UV back grou nd produc ed byg alaxies and qua sars.Inlow -den s it yreg io n s,t he IG M ga s densit y($ \rho$ ) an d t e mper atu re ( $T$)are closely c onn ecte d , v ia aredsh ift- de pende nt pol ytrop ic power-law rela tion . In simp lefo rm, t he re l ationcan be expres sed asa fu nc t i o nof redshift as: $$ \l o g T (z) = \l og T_0 ( z) + [ \gamma(z )-1] \lo g \delt a(z) , \label{e q_IGM_ t em pe rature_ de nsity_ re lat ion }$$ w h ere$T_0$is the g as me a n temperature, $\delta = \rh o /\ r ho _ {\rm c} $, $\rho_{\ rm c } $ is the cr iti c al de nsity ,a nd $\gamma$ is a redsh if t-depe ndent parameter, w hich is al s o relatedto t h er eionization hi story model and spectral shap e of the UV backg r o und.
To th isend , F i g ur e \[fig\_igm\ _ p rope rt ies\_1\ ] c ompares th e g asmea ntemperatu re $T_0$ a sin fe rre d fro m particl esub sa mpl es ex t racted from the B Ga t d ifferen t r e d shif ts ( [*Gr idSu ite*] {},T abl e \[tab le\_grid\ _si m s\_b as e\ ]) viacomputation o fthe releva nt IG M quan t i ties tha t enter in Equation (\[ e q\_IGM\ _te mpera ture \_density \_r elatio n\] ) using stand ard S PH te c h nique s , w ith t he well-kn o w n m easur em entby Beck er et al. (2011) o f th e IGM tempera tur e at t he me a nd ens it y wh e n $\gamma = 1.3$ . This cor re s po nds to a m i ldfl attenin g of th e $T- \ rho$ re lation as expected d urin g anextended H e II rei onization proce s s. In t hefigure ,bla ck fi lled d o tsare r esults f rom th e BG, w hile emp ty blue dots with error barsrefer to the meas ure m ent s perform ed b y Becker e t a l.(2011 ).N ote t hat, fo r o u r bas elin e cosmolog y ,wes e t$\gamma(z=3 ) = 1 .3$ and$T_ 0 (z=3) =150 00~{\rm K}$ – see Table \[table\ _bas e l ine \_p a rams \_ sims\]. Moreov er, $ T _ 0$ is la rg ely insensi tive to$\ g amma$ at $z > 4$, as hig h l ig h ted by the sa me author s.Th e main p oi nt of the fig ur e is | therefore_accurately reproducing_the thermal and ionization_history of_the_IGM is_essential_for improving the_robustness of all_Ly$\alpha$-based studies. The IGM_probed by the_Ly$\alpha$_forest, highly ionized at high-$z$ and progressively neutral with decreasing redshift, consists of mildly_nonlinear_gas density_fluctuations;_the_gas traces the DM distribution,_and it is generally assumed_to be_in photoionization equilibrium with the UV background produced_by_galaxies and quasars._In low-density regions, the IGM gas density ($\rho$) and_temperature ($T$) are closely connected, via_a redshift-dependent polytropic_power-law_relation._In simple form, the_relation can be expressed as a_function of redshift as: $$\log T(z)_= \log T_0(z) +[\gamma(z) -1] \log \delta(z),_
\label{eq_IGM_temperature_density_relation}$$ where $T_0$ is the gas_mean temperature, $\delta = \rho/_\rho_{\rm c}$,_$\rho_{\rm c}$ is the critical_density, and $\gamma$_is a_redshift-dependent parameter, which_is also related to the reionization_history model and_spectral shape of the UV background.
To_this_end, Figure \[fig\_igm\_properties\_1\]_compares_the_gas mean_temperature $T_0$ as_inferred_from particle_subsamples_extracted from the BG at different_redshifts_([*Grid Suite*]{}, Table \[table\_grid\_sims\_base\]) via computation of_the relevant IGM quantities_that_enter in Equation (\[eq\_IGM\_temperature\_density\_relation\])_using standard SPH techniques, with_the well-known measurement by Becker et_al. (2011)_of the_IGM temperature at the mean density when $\gamma = 1.3$. This_corresponds to a mild flattening of_the $T-\rho$ relation as_expected during_an_extended He II_reionization_process. In_the figure, black filled dots are results_from the_BG, while empty blue dots with_error bars refer to_the_measurements performed by Becker et al._(2011). Note that, for our baseline_cosmology, we set $\gamma(z=3)=1.3$ and_$T_0_(z=3)=15000~{\rm_K}$ – see Table \[table\_baseline\_params\_sims\]._Moreover, $T_0$ is largely insensitive to_$\gamma$ at $z_> 4$, as highlighted by the same_authors._The main point of the figure_is |
is that the severity of the numerical artefacts can be enhanced by the optimization process, which tends to maximize them at the expense of real physical effects.
An issue which requires special treatment in the optimization of HHG is the requirement of prevention of permanent ionization. While ionization is an integral part of the typical HHG mechanism, the magnitude of the electronic probability liberated into the macroscopic medium should be minimized. This requirement has to be reflected in the OCT formulation.
Several studies dealing with theoretical optimization of HHG have been recently published [@RasanenHHG; @DFTHHG; @DombiGeneticHHG; @Jin2016; @SPO_HHG]. Ref. [@RasanenHHG] employs a maximization term similar to the one employed in our previous work, as well as in [@RamanASOCT]. The study aims at the extension of the cutoff frequency. However, as noted by the authors, the available band of the source includes low frequencies (as in [@OCTHG]), which increase the ponderomotive energy and thus extend the cutoff. The low frequencies dominate the spectra of the optimized pulses. Thus, the control achievement presented in this study is actually the maximization of the emission in a region which is below the cutoff frequency of the dominating frequency in the pulse. A demonstration of the extension of the cutoff without the extension of the available band to lower frequencies has not been achieved to date.
Ref. [@DFTHHG] utilizes the principles presented in [@thesis; @OCTHG] to the development of an OCT formulation in the framework of time-dependent Density-Functional-Theory (TDDFT). This theory enables an optimization of HHG in multi-electron systems. Unlike in [@thesis; @OCTHG; @RasanenHHG], the system is controlled by the variation of the profile of a *slowly varying envelope* of a fixed carrier frequency. Refs. [@DombiGeneticHHG; @Jin2016; @SPO_HHG] employ genetic algorithm optimization schemes, an approach which has already been employed in both theoretical and experimental works, as was mentioned above. The contribution of the present study lies in the thorough treatment of the numerical aspects of the problem, as well as in the theoretical treatment of the restriction of permanent ionization, a topic which has not been addressed in other studies.
The aim of this paper is to establish a comprehensive approach based on OCT to address the | is that the severity of the numerical artefacts can be enhanced by the optimization procedure, which tend to maximize them at the expense of real physical effects.
An return which requires special discussion in the optimization of HHG is the requirement of prevention of permanent ionization. While ionization is an integral part of the distinctive HHG mechanism, the order of magnitude of the electronic probability liberated into the macroscopic medium should be minimize. This requirement has to be reflect in the OCT formulation.
Several studies dealing with theoretical optimization of HHG have been recently published [ @RasanenHHG; @DFTHHG; @DombiGeneticHHG; @Jin2016; @SPO_HHG ]. Ref. [ @RasanenHHG ] use a maximization term similar to the one employed in our former work, as well as in [ @RamanASOCT ]. The cogitation aims at the propagation of the cutoff frequency. However, as noted by the authors, the available band of the source include low frequencies (as in [ @OCTHG ]), which increase the ponderomotive energy and thus extend the cutoff. The low frequencies dominate the spectra of the optimized pulses. Thus, the control achievement presented in this study is actually the maximization of the discharge in a region which is below the cutoff frequency of the dominate frequency in the pulsation. A demonstration of the extension of the cutoff without the extension of the available band to lower frequencies has not been achieved to date.
Ref. [ @DFTHHG ] use the principles presented in [ @thesis; @OCTHG ] to the development of an OCT formulation in the framework of time - subject Density - Functional - Theory (TDDFT). This theory enables an optimization of HHG in multi - electron systems. Unlike in [ @thesis; @OCTHG; @RasanenHHG ], the system is see by the variation of the visibility of a * lento vary envelope * of a fixed carrier frequency. Refs. [ @DombiGeneticHHG; @Jin2016; @SPO_HHG ] use genetic algorithm optimization schemes, an approach which has already been use in both theoretical and experimental works, as was mentioned above. The contribution of the present study lies in the exhaustive treatment of the numerical aspects of the problem, as well as in the theoretical discussion of the restriction of permanent ionization, a topic which has not been addressed in other study.
The aim of this newspaper is to establish a comprehensive approach free-base on OCT to cover the | is that the severity of tht numerical artefcxts cai be enganced bh the optimization process, wiich tendw to maximize them at ghe expende of reql piysical effects.
Ai issue which dcquirzs special treatkent in tha optimization ow KHG is the requirement of prevention of perkajent ionizatiog. Whplq iohpzction is an integral part of ths typicel HHG mechanisk, the magnitude of the eleftrojic probability liherated inti thq macroscopic medium should be minijized. This requirement has to bd refkected in jkw OFJ formulatioi.
Severwl studies dcsling fith throretical optikizetiob of HHG have been rerently published [@RaswnenHHG; @DXTKHG; @DombiGeneticHHG; @Jun2016; @SPO_HVG]. Raf. [@RarqnevHHF] xmpmoys a maeimization ferm similae to the one employtd yb our previoua work, af well as in [@RamanASOCT]. The study aims au the extension of the cutofd frequency. However, ad noted br the authors, the available band of the source inwludea loc freqjwnfies (as in [@OCTHG]), which increase the ponderomotybe ekergy and thus ewtend the cutoff. Tne lpr frequencies domincfe the spectra of thf optimyzed pulses. Thts, tne control achievement preswnted in thif study is actually the maximivation of tne emission in a region whidh is below the cutorw frequency of tfe cokinating frequency in the [ulse. A dxmonscration uf tne extqnsion of hhe cmdoff without the edtensnon ox the avaipable band to lower frequencies has not been scviened to dace.
Ref. [@DNTHHG] utilizes ehe principles presenced in [@ghesis; @OCTGG] to tie developmegt of an OCT xlrmulation ii the fraiewoek od time-ddoendent Densitu-Functioncj-Theiry (TDDFT). This thepry snables an optinizqtion of HHG in muuti-qlvctcon srvtems. Unlike in [@tfeskx; @OCTFG; @RasanenHMG], ghe xystem is controlled by fhe variation of tne profile of a *sljwly varying rnvelope* of a fixef carcier fcequenvy. Tefs. [@DombiGeneticHHG; @Jin2016; @SPO_HHG] smploy gejetlc algorithm jptinization schzmes, an approach which has already been xmployed in both theorgtical and experimentcl works, as wes meneioned abmve. The contribution of the present suudy lies in the thoroufh tredtmenh of the numerical aspects of the problem, as well as in the theoretical teeatmeit of the resfricyion mf 'ermanegt iliization, a topic ehich has not been addressed in other sdubies.
The aim of this paper is yo establish a zomprehensive approach bzsed on PCT to address the | is that the severity of the numerical be by the process, which tends expense real physical effects. issue which requires treatment in the optimization of HHG the requirement of prevention of permanent ionization. While ionization is an integral part the typical HHG mechanism, the magnitude of the electronic probability liberated into the medium be This has to be reflected in the OCT formulation. Several studies dealing with theoretical optimization of HHG been recently published [@RasanenHHG; @DFTHHG; @DombiGeneticHHG; @Jin2016; @SPO_HHG]. [@RasanenHHG] employs a maximization similar to the one employed our work, as as [@RamanASOCT]. study aims at extension of the cutoff frequency. However, as noted by the authors, the available band of the source low frequencies [@OCTHG]), which the energy thus extend the low frequencies dominate the spectra of Thus, the control achievement presented in this study actually the of the emission in a region is below the cutoff frequency of the dominating in the pulse. A demonstration of the extension of the cutoff without the extension of band to lower frequencies not been achieved date. [@DFTHHG] the presented in @OCTHG] to the development of an OCT formulation in the framework time-dependent Density-Functional-Theory (TDDFT). This theory enables an optimization of HHG systems. in [@thesis; @OCTHG; the system is controlled the of the profile of varying of frequency. [@DombiGeneticHHG; @SPO_HHG] employ genetic algorithm schemes, an approach which has been employed in both was mentioned above. The contribution of the present lies in the thorough treatment of the aspects of the problem, as well as in the theoretical treatment of restriction of a topic which has not been addressed in studies. The aim of paper is to establish a comprehensive approach based on to the | is that the severity of the numErical arteFacts Can Be eNhAnceD by tHe optimization PRoceSs, which tends to maximize Them aT tHE expENsE of reAl physiCAl EFFecTs.
an IssUe WHiCh reqUirEs speciAl treatmenT in ThE optimizatioN Of hHG is the reQuiRement of prevEntIon of pErManENt ionIzaTion. WHile ioNIzatioN is an inteGrAL part oF The typiCAL HhG meChanism, the magnituDE oF The electronic pRobabiLiTY lIBEraTed Into the macRoScopiC Medium sHOuLD BE miNImized. This reqUirement has TO be ReflecTeD in THe OCT fOrmulAtIOn.
SEveral studiEs deAling with TheoreTIcal optIMizatioN of HHG HavE beEn reCEnTlY puBlISheD [@raSanENHHg; @DFTHHG; @DOmBigenetIcHHg; @jIN2016; @sPO_HhG]. REf. [@RaSanenhHG] employs a maXimIzatIOn tErm siMilar To thE oNe empLoyed iN our pReVious work, as well As in [@ramanASOCt]. ThE sTudY aIms at THe exteNsiOn oF the cutOff freqUEncY. HOWEVeR, as noted by the authoRs, THE aVailable Band of THe SoURce incluDeS loW freQUEncieS (as iN [@oCtHG]), which IncreaSE tHe PonderoMoTive enErGy aNd tHus exTEnd tHe cutoFf. The low FrequENcies dominate tHE spectra of the OPtIMIzED pulSes. thus, the contRol aCHievEmenT PrEseNTed in This sTuDY iS Actually the maximizaTiOn of thE emisSion in a region Which is belOW THe cutoff FreqUEnCY of the dominatiNg freQuency in thE Pulse. A deMonstRation of The extensION of the cuTofF wiThoUt tHE ExTension of the aVAIlabLe Band to lOweR frequeNciEs hAs nOt bEeN achieved To date.
ReF. [@DfThHg] uTilIzes tHE principLeS prEsEntEd in [@tHEsis; @OCtHG] to The dEvElOPmeNt of an Oct fORMulaTiOn In thE frAmEwork Of tiME-dePendent density-FuNctIOnal-thEoRy (TDDFT). this theory enaBlEs an optimiZaTioN of HHG IN Multi-eleCtron systems. Unlike in [@theSIs; @OCTHG; @rasAnenHhG], thE system is ConTrolleD by THe variAtion oF the pRoFilE OF a *sloWLY vAryInG envelope* oF A FixEd carRiEr frEquency. refs. [@DombiGeneticHHg; @jin2016; @sPO_HHG] employ gEneTic aLGOrIthM OpTImiZaTIon SCHemes, an approach Which has alReADy Been employED in BoTh theorEtical aNd expERimentaL works, as wAs mentionEd AbovE. tHe cOntributioN of the prEsent studY Lies iN ThE thorOugH treatMeNt oF the nUmericAL asPects Of the pRoBlem, as Well aS iN the theoRetical treatment of the reStrictIon of PerManent ionIzaTIon, A topic whiCh haS not been adDreSseD in otHer STudieS.
The AIm Of tHIs papEr is TO establisH A cOmpREHeNsive approaCH BAseD on OCt to ADdress The | is that the severity of t he numeric al ar tef act scanbe e nhanced by the opti mization process, whic h ten ds to m a xi mizethem at th e exp en se of r e al phys ica l effec ts.
An is sue w hich require s s pecial tre atm ent in the o pti mizati on of HHG i s t he re quirem e nt ofpreventio no f perm a nent io n i za tion . While ionizatio n i s an integral p art of t h et y pic alHHG mechan is m, th e magnit u de o f th e electronic p robabilityl ibe ratedin tot he mac rosco pi c me dium should beminimized . This require m ent has to be re fle cted in t heOC T fo r mu lat i on.
Severa lst udies dea l i n g wit h t heor etica l optimizatio n o f HH G ha ve be en re cent ly publ ished[@Ras an enHHG; @DFTHHG; @Do mbiGeneti cHH G; @J in 2016; @SPO_H HG] . R ef. [@R asanenH H G]em p l o ys a maximization te rm s im ilar tothe on e e mp l oyed inou r p revi o u s wor k, a s w ell as i n [@Ra m an AS OCT]. T he study a ims at thee xten sion o f the cu tofff requency. Howe v er, as notedb yt h ea utho rs, the availa bleb andof t h esou r ce in clude sl ow frequencies (as in[@ OCTHG] ), wh ich increasethe ponder o m o tive ene rgya nd thus extend th e cut off. The l o w freque ncies dominat e the spe c t ra of th e o pti miz edp u ls es. Thus, the c ontr ol achiev eme nt pres ent edinthi sstudy isactually t he m ax imi zatio n of theem iss io n i n a r e gion w hichis b el ow the cutoff fr e q uenc yof the do mi natin g fr e que ncy inthe pulse . A demo ns tr ation o f the extensi on of the cu to ffwithou t the exte nsion of the availableb and tolow er fr eque ncies has no t been ac h ievedto dat e.
R ef . [ @ D FTHHG ] ut ili ze s the prin c i ple s pre se nted in [@t hesis; @OCTHG] tot hedevelopment o f a n OC T fo rmu l at i onin the f ramework of tim e-dependen tD en sity-Funct i ona l- Theory(TDDFT) . Thi s theory enablesan optimi za tion o f H HG in mult i-electr on system s . Unl i ke in [ @th esis;@O CTH G; @R asanen H HG] , the syste mis con troll ed by thevariation of the profil e of a *slo wly varyingenv e lop e* of a f ixed carrier f req uen cy. R efs . [@Do mbiG e ne tic H HG; @ Jin2 0 16; @SPO_ H HG ] e m p lo y genetic a l g o rit hm op tim i zation sch emes, an approach which has alre adyb e enemp l oyed i n both theoret ica la n d experi me ntal works, as wasme n tione d abov e. The contri b u ti o n of t he p res ent study li es in theth or o ugh tr eatm en t of t he num e rica l aspects of the p roble m , as w e llas in t he theo r etic al treatme nt of the r estric tion of p ermanen tioniza tio n, a topic w h ich has n ot be en addr es sedinotherstud i e s.
T he a im of this pap e r i s t oe sta blis h a c om preh ensive ap p roach ba sed on OCTto ad d r ess th e | is_that the_severity of the numerical_artefacts can_be_enhanced by_the_optimization process, which_tends to maximize_them at the expense_of real physical_effects.
An_issue which requires special treatment in the optimization of HHG is the requirement of_prevention_of permanent_ionization._While_ionization is an integral part_of the typical HHG mechanism,_the magnitude_of the electronic probability liberated into the macroscopic_medium_should be minimized._This requirement has to be reflected in the OCT_formulation.
Several studies dealing with theoretical optimization_of HHG have_been_recently_published [@RasanenHHG; @DFTHHG; @DombiGeneticHHG;_@Jin2016; @SPO_HHG]. Ref. [@RasanenHHG] employs a maximization_term similar to the one employed_in our previous work, as well as_in [@RamanASOCT]. The study aims at the_extension of the cutoff frequency._However, as_noted by the authors, the_available band of_the source_includes low frequencies_(as in [@OCTHG]), which increase the ponderomotive_energy and thus_extend the cutoff. The low frequencies_dominate_the spectra of_the_optimized_pulses. Thus,_the control achievement_presented_in this_study_is actually the maximization of the_emission_in a region which is below the_cutoff frequency of the_dominating_frequency in the pulse._A demonstration of the extension_of the cutoff without the extension_of the_available band_to lower frequencies has not been achieved to date.
Ref. [@DFTHHG] utilizes the_principles presented in [@thesis; @OCTHG] to the_development of an OCT_formulation in_the_framework of time-dependent_Density-Functional-Theory_(TDDFT). This_theory enables an optimization of HHG in_multi-electron systems._Unlike in [@thesis; @OCTHG; @RasanenHHG], the system_is controlled by the_variation_of the profile of a *slowly_varying envelope* of a fixed carrier_frequency. Refs. [@DombiGeneticHHG; @Jin2016; @SPO_HHG] employ_genetic_algorithm_optimization schemes, an approach which_has already been employed in both_theoretical and experimental_works, as was mentioned above. The contribution_of_the present study lies in the_thorough_treatment of the numerical aspects of_the_problem,_as well as in the_theoretical treatment of the restriction of_permanent ionization, a topic which has not been addressed_in other studies.
The_aim of this paper is_to_establish_a comprehensive approach based on OCT to address the |
.5 & 0.844\
\
\
& 279.601 & 7.646 & LRc02 & 4179 & 39.4494 & 0.9 & 0.859\
[**ID 105284610**]{} & 279.873 & 7.679 & LRc02 & 4421 & 37.3701 & 0.4 & 0.886\
[**ID 105367925**]{} & 279.991 & 6.731 & LRc02 & 4842 & 68.9516 & 1 & 0.901\
[**ID 105379106**]{} & 280.006 & 7.311 & LRc02 & 4401 & 67.2065 & 1 & 0.872\
\
& 291.644 & -0.053 & LRc01 & 5989 & 17.3479& 0.2 & 0.785\
[**ID 101710670**]{} & 292.671 & -0.023 & LRc01 & 6082 & 5.4984& 0.02 & 0.823\
[**ID 102692502**]{} & 100.873 & 0.001 & LRa01 & 15526 & 9.5111 & 0.02 & 0.783\
[**ID 102752622**]{} & 101.208 & -1.065 & IRa01/LRa01 & 6558 & 2.3321 & 0.0003 & 0.601\
[**ID 102770893**]{} & 101.311 & -1.217 & IRa01/LRa01 & 6140 & 4.2850 & 0.002 & 0.755\
[**ID 105503339**]{} & 280.181 & 7.651 & LRc02 & 4253 & 10.600 & 0.4 & 0.818\
[**ID 105665211**]{} & 280.436 & 5.416 & LRc02 & 8388 & 3.8148 & 0.003 & 0.700\
[**ID 105945509**]{} & 280.908 & 5.744 & LRc02 & 4725 & 2.7917 & 0.003 & 0.665\
[**ID 105957346**]{} & 280.933 & 6.441 & LRc02 & 4236 & 0.3341 & 5$\cdot10^{-5}$ & 0.612\
[**ID 105845539** ]{} & 280.705 & 7.556 & LR | .5 & 0.844\
\
\
& 279.601 & 7.646 & LRc02 & 4179 & 39.4494 & 0.9 & 0.859\
[ * * ID 105284610 * * ] { } & 279.873 & 7.679 & LRc02 & 4421 & 37.3701 & 0.4 & 0.886\
[ * * ID 105367925 * * ] { } & 279.991 & 6.731 & LRc02 & 4842 & 68.9516 & 1 & 0.901\
[ * * ID 105379106 * * ] { } & 280.006 & 7.311 & LRc02 & 4401 & 67.2065 & 1 & 0.872\
\
& 291.644 & -0.053 & LRc01 & 5989 & 17.3479 & 0.2 & 0.785\
[ * * ID 101710670 * * ] { } & 292.671 & -0.023 & LRc01 & 6082 & 5.4984 & 0.02 & 0.823\
[ * * ID 102692502 * * ] { } & 100.873 & 0.001 & LRa01 & 15526 & 9.5111 & 0.02 & 0.783\
[ * * ID 102752622 * * ] { } & 101.208 & -1.065 & IRa01 / LRa01 & 6558 & 2.3321 & 0.0003 & 0.601\
[ * * ID 102770893 * * ] { } & 101.311 & -1.217 & IRa01 / LRa01 & 6140 & 4.2850 & 0.002 & 0.755\
[ * * ID 105503339 * * ] { } & 280.181 & 7.651 & LRc02 & 4253 & 10.600 & 0.4 & 0.818\
[ * * ID 105665211 * * ] { } & 280.436 & 5.416 & LRc02 & 8388 & 3.8148 & 0.003 & 0.700\
[ * * ID 105945509 * * ] { } & 280.908 & 5.744 & LRc02 & 4725 & 2.7917 & 0.003 & 0.665\
[ * * ID 105957346 * * ] { } & 280.933 & 6.441 & LRc02 & 4236 & 0.3341 & 5$\cdot10^{-5}$ & 0.612\
[ * * ID 105845539 * * ] { } & 280.705 & 7.556 & LR | .5 & 0.844\
\
\
& 279.601 & 7.646 & LRc02 & 4179 & 39.4494 & 0.9 & 0.859\
[**ID 105284610**]{} & 279.873 & 7.679 & URc02 & 4421 & 37.3701 & 0.4 & 0.886\
[**ID 105367925**]{} & 279.991 & 6.731 & LRr02 & 4842 & 68.9516 & 1 & 0.901\
[**ID 105379106**]{} & 280.006 & 7.311 & LRc02 & 4401 & 67.2065 & 1 & 0.872\
\
& 291.644 & -0.053 & LRc01 & 5989 & 17.3479& 0.2 & 0.785\
[**IE 101710670**]{} & 292.671 & -0.023 & LRc01 & 6082 & 5.4984& 0.02 & 0.823\
[**ID 102692502**]{} & 100.873 & 0.001 & URa01 & 15526 & 9.5111 & 0.02 & 0.783\
[**ID 102752622**]{} & 101.208 & -1.065 & ICa01/LRa01 & 6558 & 2.3321 & 0.0003 & 0.601\
[**ID 102770893**]{} & 101.311 & -1.217 & IRa01/LRa01 & 6140 & 4.2850 & 0.002 & 0.755\
[**MD 105503339**]{} & 280.181 & 7.651 & LRc02 & 4253 & 10.600 & 0.4 & 0.818\
[**ID 105665211**]{} & 280.436 & 5.416 & LRc02 & 8388 & 3.8148 & 0.003 & 0.700\
[**IA 105945509**]{} & 280.908 & 5.744 & LRc02 & 4725 & 2.7917 & 0.003 & 0.665\
[**ID 105957346**]{} & 280.933 & 6.441 & LRc02 & 4236 & 0.3341 & 5$\sdot10^{-5}$ & 0.612\
[**IC 105845539** ]{} & 280.705 & 7.556 & LR | .5 & 0.844\ \ \ & 279.601 & & 4179 39.4494 & 0.9 279.873 7.679 & LRc02 4421 & 37.3701 0.4 & 0.886\ [**ID 105367925**]{} & & 6.731 & LRc02 & 4842 & 68.9516 & 1 & 0.901\ [**ID & 280.006 & 7.311 & LRc02 & 4401 & 67.2065 & 1 & \ 291.644 -0.053 LRc01 & 5989 & 17.3479& 0.2 & 0.785\ [**ID 101710670**]{} & 292.671 & -0.023 & LRc01 6082 & 5.4984& 0.02 & 0.823\ [**ID 102692502**]{} 100.873 & 0.001 & & 15526 & 9.5111 & & [**ID 102752622**]{} 101.208 -1.065 IRa01/LRa01 & 6558 2.3321 & 0.0003 & 0.601\ [**ID 102770893**]{} & 101.311 & -1.217 & IRa01/LRa01 & 6140 & 4.2850 0.002 & 105503339**]{} & & & & 4253 & 0.4 & 0.818\ [**ID 105665211**]{} & & LRc02 & 8388 & 3.8148 & 0.003 0.700\ [**ID & 280.908 & 5.744 & LRc02 4725 & 2.7917 & 0.003 & 0.665\ [**ID & 280.933 & 6.441 & LRc02 & 4236 & 0.3341 & 5$\cdot10^{-5}$ & 0.612\ [**ID & 280.705 & 7.556 LR | .5 & 0.844\
\
\
& 279.601 & 7.646 & LRc02 & 4179 & 39.4494 & 0.9 & 0.859\
[**ID 105284610**]{} & 279.873 & 7.679 & LRc02 & 4421 & 37.3701 & 0.4 & 0.886\
[**ID 105367925**]{} & 279.991 & 6.731 & LRc02 & 4842 & 68.9516 & 1 & 0.901\
[**ID 105379106**]{} & 280.006 & 7.311 & LRc02 & 4401 & 67.2065 & 1 & 0.872\
\
& 291.644 & -0.053 & LRc01 & 5989 & 17.3479& 0.2 & 0.785\
[**ID 101710670**]{} & 292.671 & -0.023 & LRC01 & 6082 & 5.4984& 0.02 & 0.823\
[**ID 102692502**]{} & 100.873 & 0.001 & LRa01 & 15526 & 9.5111 & 0.02 & 0.783\
[**ID 102752622**]{} & 101.208 & -1.065 & IRA01/LRa01 & 6558 & 2.3321 & 0.0003 & 0.601\
[**Id 102770893**]{} & 101.311 & -1.217 & IRA01/LRA01 & 6140 & 4.2850 & 0.002 & 0.755\
[**Id 105503339**]{} & 280.181 & 7.651 & LRc02 & 4253 & 10.600 & 0.4 & 0.818\
[**iD 105665211**]{} & 280.436 & 5.416 & LRC02 & 8388 & 3.8148 & 0.003 & 0.700\
[**ID 105945509**]{} & 280.908 & 5.744 & LRc02 & 4725 & 2.7917 & 0.003 & 0.665\
[**ID 105957346**]{} & 280.933 & 6.441 & LRc02 & 4236 & 0.3341 & 5$\cdoT10^{-5}$ & 0.612\
[**iD 105845539** ]{} & 280.705 & 7.556 & LR | .5 & 0.844\
\
\
& 279.601& 7.646 &LRc02 &417 9& 39 .449 4 & 0.9 & 0.85 9 \
[* *ID 105284610**]{} & 2 79.87 3& 7.6 7 9& LRc 02 & 44 2 1& 37. 37 01 &0. 4 & 0.88 6\[**ID 1 05367925** ]{} & 279.991 & 6 . 73 1 & LRc02& 4 842 & 68.951 6 & 1 & 0 .9 01\ [**ID 10 53791 06**]{ } & 280 .006 & 7. 31 1 & LRc 0 2 & 440 1 &67.2 065 & 1 & 0.872\\
& 291.644 & -0.0 53 & L Rc 0 1& 598 9 & 17.3479&0. 2 & 0 . 785\
[* * ID 1 0 171 0 670**]{} & 29 2.671 & -0. 0 23& LRc0 1& 6 0 82 & 5 .4984 &0 .02 & 0.823\
[ **ID 10269250 2**]{} & 100.8 7 3 & 0.0 01 & L Ra0 1 & 155 2 6&9.5 11 1 &0 .0 2 & 0.7 83\
[**I D10 27526 22** ] { } & 10 1.2 08 & -1.0 65 & IRa01/LR a01 & 6 5 58& 2.3 321 & 0.0 00 3 & 0 .601\[**ID 1 02770893**]{} & 101 .311 & -1 .21 7& I Ra 01/LR a 01 & 6 140 &4.2850& 0.002 & 0 .7 5 5 \
[ **ID 105503339**]{ }& 28 0.181 &7.651& L Rc 0 2 & 4253 & 10 .600 & 0.4& 0. 8 18 \
[**ID105665 2 11 ** ]{} & 2 80 .436 & 5 .41 6 & LRc0 2 & 8 388 &3.8148 & 0.00 3 & 0.700\
[**I D 105945509**] { }& 28 0 .908 &5.744 & LRc 02 & 4725 & 2 . 79 17& 0.00 3 & 0 .6 6 5\ [**ID 105957346**]{ }& 280. 933 & 6.441 & LRc0 2 & 4236 & 0 . 3341 & 5 $\cd o t1 0 ^{-5}$ & 0.612 \
[** ID 1058455 3 9** ]{}& 280 .705 & 7 .556 & LR | .5 &_0.844\
\
\
& 279.601_& 7.646 & LRc02_& 4179_&_39.4494 &_0.9_& 0.859\
[**ID 105284610**]{}_& 279.873 &_7.679 & LRc02 &_4421 & 37.3701_&_0.4 & 0.886\
[**ID 105367925**]{} & 279.991 & 6.731 & LRc02 & 4842 & 68.9516_&_1 &_0.901\
[**ID_105379106**]{}_& 280.006 & 7.311 &_LRc02 & 4401 & 67.2065_& 1_& 0.872\
\
& 291.644 & -0.053 & LRc01 &_5989_& 17.3479& 0.2_& 0.785\
[**ID 101710670**]{} & 292.671 & -0.023 & LRc01_& 6082 & 5.4984& 0.02 &_0.823\
[**ID 102692502**]{} &_100.873_&_0.001 & LRa01 &_15526 & 9.5111 & 0.02 &_0.783\
[**ID 102752622**]{} & 101.208 & -1.065_& IRa01/LRa01 & 6558 & 2.3321 &_0.0003 & 0.601\
[**ID 102770893**]{} & 101.311_& -1.217 & IRa01/LRa01 &_6140 &_4.2850 & 0.002 & 0.755\
[**ID_105503339**]{} & 280.181_& 7.651_& LRc02 &_4253 & 10.600 & 0.4 &_0.818\
[**ID 105665211**]{} &_280.436 & 5.416 & LRc02 &_8388_& 3.8148 &_0.003_&_0.700\
[**ID 105945509**]{}_& 280.908 &_5.744_& LRc02_&_4725 & 2.7917 & 0.003 &_0.665\
[**ID_105957346**]{} & 280.933 & 6.441 & LRc02_& 4236 & 0.3341_&_5$\cdot10^{-5}$ & 0.612\
[**ID 105845539**_]{} & 280.705 & 7.556_& LR |
conjecture*: any NIP valued field $(K,\mathcal{O})$ is henselian.
- The *Shelah conjecture*: any NIP field $K$ is algebraically closed, real closed, finite, or admits a non-trivial henselian valuation.
By generalizing the arguments used for dp-minimal fields (for example in Chapter 9 of [@myself]), we prove the henselianity conjecture in positive characteristic, and the Shelah conjecture for positive characteristic dp-finite fields. This yields the positive-characteristic part of the expected classification of dp-finite fields.
We also make partial progress on dp-finite fields of characteristic zero. Let $(K,+,\cdot,\ldots)$ be a sufficiently saturated dp-finite field, possibly with extra structure. Then either
- $K$ has finite Morley rank, or
- There is an ${\operatorname{Aut}}(K/A)$-invariant non-trivial valuation ring on $K$ for some small set $A$.
Unfortunately, we can only prove henselianity of this valuation ring in positive characteristic.
Following the approach used for dp-minimal fields, there are three main steps to the proof:
1. Construct a type-definable group of infinitesimals.
2. Construct a valuation ring from the infinitesimals.
3. Prove henselianity.
We discuss each of these steps, explaining the difficulties that arise when generalizing from rank 1 to rank $n$.
Constructing the infinitesimals
-------------------------------
Mimicking the case of dp-minimal fields, we would like to define the group $I_M$ of $M$-infinitesimals as $$\bigcap_{X \text{ ``big'' and $M$-definable}} \{\delta \in K ~|~ X
\cap (X + \delta) \text{ is ``big''}\}$$ for some notion of “big.” In the dp-minimal case, “big” was “infinite.” By analyzing the proof for dp-minimal fields, one can enumerate a list of desiderata for bigness:
1. Non-big sets should form an ideal.
2. Bigness should be preserved by affine transformations.
3. \[definability-condition\] Bigness should vary definably in families.
4 | conjecture *: any NIP valued field $ (K,\mathcal{O})$ is henselian.
- The * Shelah speculation *: any NIP sphere $ K$ is algebraically closed, real closed, finite, or admit a non - trivial henselian valuation.
By generalize the argument used for dp - minimal airfield (for example in Chapter 9 of [ @myself ]), we prove the henselianity guess in positive characteristic, and the Shelah conjecture for positivist characteristic dp - finite fields. This render the positive - characteristic part of the expect classification of dp - finite fields.
We also hold partial progress on dp - finite fields of characteristic zero. permit $ (K,+,\cdot,\ldots)$ be a sufficiently saturated dp - finite field, possibly with extra structure. Then either
- $ K$ has finite Morley rank, or
- There be an $ { \operatorname{Aut}}(K / A)$-invariant non - trivial valuation ring on $ K$ for some small set $ A$.
Unfortunately, we can merely prove henselianity of this valuation ring in positive characteristic.
Following the approach used for dp - minimal fields, there are three chief steps to the proof:
1. reconstruct a character - definable group of infinitesimal.
2. Construct a valuation ring from the infinitesimals.
3. Prove henselianity.
We discuss each of these steps, explaining the difficulty that arise when generalizing from rank 1 to rank $ n$.
Constructing the infinitesimals
-------------------------------
mimic the case of dp - minimal fields, we would like to define the group $ I_M$ of $ M$-infinitesimals as $ $ \bigcap_{X \text { '` big" and $ M$-definable } } \{\delta \in K ~|~ X
\cap (X + \delta) \text { is '` big''}\}$$ for some notion of “ large. ” In the dp - minimal case, “ big ” was “ infinite. ” By analyzing the proof for dp - minimal sphere, one can enumerate a list of desiderata for largeness:
1. Non - big set should form an ideal.
2. Bigness should be preserved by affine transformation.
3. \[definability - condition\ ] Bigness should vary definably in family.
4 | cojjecture*: any NIP valued nield $(K,\mathcal{O})$ nw hensxlian.
- The *Sheuah conjecture*: any NIP field $K$ is aogebraically closed, rexl closed, finite, ir avmits a non-triviem henselian vzpuatnoi.
By generalizinn the argumants used for gp-oiuimal fields (for example in Chapter 9 of [@mysrlv]), we prove the henxqliahptn conjecture in positive charactsristic, and the Shelsh conjecture for positive chagacteristic dp-finihe fields. Tyis ruelds the poritive-characteristic pzrt of the expected classificatkon oy dp-finite diwldd.
Fe also makx partpal progress on dp-finhte fiekds of charactcristmc zwro. Let $(K,+,\cdot,\ldots)$ be a sufficiently sajurated dp-xiuite field, possibly wuty extta stsuctjee. Gheh xitger
- $N$ hes finite Mkrley rank, ir
- There is an ${\optraeirname{Aut}}(K/A)$-inbariane gon-trivial valuation ring on $K$ for some smzll set $A$.
Unfortunately, qe can only prove hendelianity of this valuation ring in positive characteristiw.
Follkdiny the xppgoach used for dp-minimal fields, there are thrqs kapn steps to the pvoof:
1. Construct a yyoe-cgfinable group of inyjnjtesimals.
2. Construft a vajuatiin ring fwom yhe infinitesimals.
3. Prove hwnselianity.
Wv diwcuss each of thesz steps, explcining the cifficulties that arise wheh generalizlng from dxnk 1 to rank $n$.
Covstgucthng the infinitesimals
-------------------------------
Mimisking the casz of dp-mknimsl fiejds, we woupd like to define the grlup $I_L$ mf $M$-infinihesimals as $$\bigcap_{X \text{ ``big'' anv $M$-definable}} \{\centa \in K ~|~ X
\cap (X + \delta) \texe{ is ``big''}\}$$ for spme notnon of “big.” In thv dp-minimel case, “big” ras “infinite.” Ti analyzing tie proof sor ep-mibimal fkdlds, one can emumerate c lisr of desiderata fov bigvsss:
1. Non-big setw syould form an iceau.
2. Fihnxss srmuld be presarvea bh affive transfornxtioms.
3. \[definability-condhtioh\] Bigness should vsrn definaboy in faiilies.
4 | conjecture*: any NIP valued field $(K,\mathcal{O})$ is The conjecture*: any field $K$ is or a non-trivial henselian By generalizing the used for dp-minimal fields (for example Chapter 9 of [@myself]), we prove the henselianity conjecture in positive characteristic, and Shelah conjecture for positive characteristic dp-finite fields. This yields the positive-characteristic part of expected of fields. also make partial progress on dp-finite fields of characteristic zero. Let $(K,+,\cdot,\ldots)$ be a sufficiently saturated field, possibly with extra structure. Then either - has finite Morley rank, - There is an ${\operatorname{Aut}}(K/A)$-invariant valuation on $K$ some set Unfortunately, we can prove henselianity of this valuation ring in positive characteristic. Following the approach used for dp-minimal fields, there three main the proof: Construct type-definable of infinitesimals. 2. valuation ring from the infinitesimals. 3. discuss each of these steps, explaining the difficulties arise when from rank 1 to rank $n$. the infinitesimals ------------------------------- Mimicking the case of dp-minimal we would like to define the group $I_M$ of $M$-infinitesimals as $$\bigcap_{X \text{ ``big'' and \in K ~|~ X (X + \delta) is for notion “big.” In dp-minimal case, “big” was “infinite.” By analyzing the proof for dp-minimal one can enumerate a list of desiderata for bigness: 1. should an ideal. 2. should be preserved by transformations. \[definability-condition\] Bigness should vary families. | conjecture*: any NIP valued fieLd $(K,\mathcal{o})$ is heNseLiaN.
- THe *ShElah Conjecture*: any Nip fieLd $K$ is algebraically closEd, reaL cLOsed, FInIte, or Admits a NOn-TRIviAl HeNseLiAN vAluatIon.
by generAlizing the ArgUmEnts used for dP-MiNimal fieldS (foR example in ChAptEr 9 of [@mySeLf]), wE Prove The HenseLianitY ConjecTure in posItIVe charACteristIC, AnD the shelah conjecture fOR pOSitive characteRistic Dp-FInITE fiEldS. This yieldS tHe posITive-chaRAcTERIstIC part of the expEcted classiFIcaTion of Dp-FinITe fielDs.
We aLsO MakE partial proGresS on dp-finiTe fielDS of charACteristIc zero. let $(k,+,\cdOt,\ldOTs)$ Be A suFfICieNTlY saTUraTed dp-finItE fIeld, pOssiBLY WIth eXtrA strUcturE. Then either
- $K$ hAs fInitE morLey raNk, or
- THere Is An ${\opeRatornAme{AuT}}(K/a)$-invariant non-trIviaL valuatioN riNg On $K$ FoR some SMall seT $A$.
UNfoRtunateLy, we can ONly PrOVE HeNselianity of this vaLuATIoN ring in pOsitivE ChArACteristiC.
FOllOwinG THe appRoacH UsEd for dp-mInimal FIeLdS, there aRe Three mAiN stEps To the PRoof:
1. constrUct a type-DefinABle group of infiNItesimals.
2. ConsTRuCT A vALuatIon Ring from the InfiNItesImalS.
3. prOve HEnselIanitY.
WE DiSCuss each of these stepS, eXplainIng thE difficulties That arise wHEN GeneraliZing FRoM Rank 1 to rank $n$.
ConStrucTing the infINitesimaLs
-------------------------------
MimIcking thE case of dp-MINimal fieLds, We wOulD liKE To Define the grouP $i_m$ of $M$-InFinitesImaLs as $$\bigCap_{x \teXt{ ``bIg'' aNd $m$-definablE}} \{\delta \in k ~|~ X
\CaP (X + \DeLta) \Text{ iS ``Big''}\}$$ for soMe NotIoN of “Big.” In THe dp-miNimal Case, “BiG” wAS “inFinite.” BY AnALYzinG tHe ProoF foR dP-miniMal fIEldS, one can Enumerate A liST of dEsIdErata foR bigness:
1. Non-biG sEts should fOrM an Ideal.
2. BIGNess shouLd be preserved by affine trANsformaTioNs.
3. \[defInabIlity-condItiOn\] BignEss SHould vAry defInablY iN faMILies.
4 | conjecture*: any NIP valu ed field $ (K,\m ath cal {O })$is h enselian.
- The*Shelah conjecture*: a ny NI Pf ield $K $ isalgebra i ca l l y c lo se d,re a lclose d,finite, or admits ano n-trivial he n se lian valua tio n.
By gener ali zing t he ar g ument s u sed f or dp- m inimal fields ( fo r examp l e in Ch a p te r 9of [@myself]), we pr o ve the henseli anityco n je c t ure in positivech aract e ristic, an d t heS helah conject ure for pos i tiv e char ac ter i stic d p-fin it e fi elds. Thisyiel ds the po sitive - charact e risticpart o f t heexpe c te dcla ss i fic a ti ono f d p-finite f ie lds.
Wea l s o mak e p arti al pr ogress on dp- fin itef iel ds of char acte ri sticzero.Let $ (K ,+,\cdot,\ldots )$ b e a suffi cie nt lysa turat e d dp-f ini tefield,possibl y wi th e x tr a structure. Thenei t h er
- $K $ hasf in it e Morleyra nk, or- Th erei san ${\op erator n am e{ Aut}}(K /A )$-inv ar ian t n on-tr i vial valua tion rin g on$ K$ for some sm a ll set $A$.
U nf o r tu n atel y,we can only pro v e he nsel i an ity of th is va lu a ti o n ring in positivech aracte risti c.
Following the appro a c h used fo r dp - mi n imal fields, t hereare threem ain step s tothe proo f:
1. C o n struct a ty pe- def ina b l egroup of infi n i tesi ma ls.
2. C onstruc t a va lua tio nring from the inf in it es im als .
3. Prove h en sel ia nit y.
W e discu ss ea ch o fth e sesteps,e xp l a inin gth e di ffi cu lties tha t ar ise whe n general izi n g fr om r ank 1 t o rank $n$.
Co nstructing t heinfini t e simals
- ----------------------- - ------
Mi micki ng t he case o f d p-mini mal fields , we w ouldli ket o defi n e t hegr oup $I_M$o f $M $-inf in ites imals a s $$\bigcap_{X \te x t{``big'' and $ M$- defi n a bl e}} \{ \ del ta \in K ~|~ X
\cap ( X + \delta )\ te xt{ is ``b i g'' }\ }$$ for some n otion of “big .” In the dp-minim al cas e , “b ig” was “i nfinite. ” By anal y zingt he proo f f or dp- mi nim al fi elds,o necan e numera te a lis t ofde siderata for bigness:
1. Non- big se ts sh oul d form an id e al.
2. Big ness should be pr ese rvedbya ffine tra n sf orm a tions .
3 . \[defin a bi lit y - co ndition\] B i g n ess shou ldv ary de fina bly in families.4 | conjecture*:_any NIP_valued field $(K,\mathcal{O})$ is_henselian.
- __The *Shelah_conjecture*:_any NIP field_$K$ is algebraically_closed, real closed, finite,_or admits a_non-trivial_henselian valuation.
By generalizing the arguments used for dp-minimal fields (for example in Chapter 9_of_[@myself]), we_prove_the_henselianity conjecture in positive characteristic,_and the Shelah conjecture for_positive characteristic_dp-finite fields. This yields the positive-characteristic part of_the_expected classification of_dp-finite fields.
We also make partial progress on dp-finite fields_of characteristic zero. Let $(K,+,\cdot,\ldots)$ be_a sufficiently saturated_dp-finite_field,_possibly with extra structure._Then either
- $K$ has_finite Morley rank, or
- _There is an ${\operatorname{Aut}}(K/A)$-invariant non-trivial valuation ring_on $K$ for some small set_$A$.
Unfortunately, we can only prove_henselianity of_this valuation ring in positive_characteristic.
Following the approach_used for_dp-minimal fields, there_are three main steps to the_proof:
1. Construct_a type-definable group of infinitesimals.
2. _Construct_a valuation ring_from_the_infinitesimals.
3. _Prove henselianity.
We discuss_each_of these_steps,_explaining the difficulties that arise when_generalizing_from rank 1 to rank $n$.
Constructing the_infinitesimals
-------------------------------
Mimicking the case of_dp-minimal_fields, we would like_to define the group $I_M$_of $M$-infinitesimals as $$\bigcap_{X \text{ ``big''_and $M$-definable}}_\{\delta \in_K ~|~ X
\cap (X + \delta) \text{ is ``big''}\}$$_for some notion of “big.” In_the dp-minimal case, “big”_was “infinite.”_By_analyzing the proof_for_dp-minimal fields,_one can enumerate a list of desiderata_for bigness:
1._ Non-big sets should form an_ideal.
2. Bigness should_be_preserved by affine transformations.
3. \[definability-condition\]_Bigness should vary definably in families.
4 |
boson space $$T^{B}(M1) = \sqrt{\frac{3}{4\pi}}(g_{\pi}\hat{L}_{\pi} +
g_{\nu}\hat{L}_{\nu})\mu_{N},$$ with $g_l$, $g_s$ being the fermion orbital and spin $g$-factors and $g_{\pi}$, $g_{\nu}$ the respective proton and neutron boson $g$ factors. The quantities $\hat{L}_{\pi}$ and $\hat{L}_{\nu}$ are the corresponding orbital angular momentum operators of the proton and neutron boson system.
In IBM-2 studies, concentrating on deformed nuclei, one is using the SU(3) reduction of the U(6) group structure and, for those nuclei, it was shown that the lowest-lying states of the family of mixed-symmetry character were characterized by the $F_{max} -1, J^{\pi} = 1^{+}$ quantum numbers [@Iachello:1981; @Iachello:1984]. These findings corroborate results obtained from a totally different starting point, viz. the TRM. The IBM-2 $J^{\pi} = 1^{+}$ states are also called scissors states although there is no immediate reference in the algebraic formulation to specific coordinate forms and thus also not of shapes and shape dynamics. Using a coherent-state formalism, @Dieperink:1983 was able to show the correspondence explicitly and, moreover, found indeed that only the valence nucleons are contributing to the strength of the scissors mode thus leading in a natural way to a much lower $B(M1)$ strength compared to the early TRM calculations.
In studying the $M1$ excitation properties within the IBM-2, because of the specific difference in magnetization properties for proton and neutron bosons, it was clear that $M1$ transitions could appear naturally now, in contrast to the former IBM-1. Using mapping from fermion magnetic properties onto boson ones, it was possible to also determine the analogous boson $g_{\pi}$ and $g_{\nu}$ factors [@Sambataro:1981; @Sambataro:1984; @Allaart:1988]. This item has been a topic of much discussion because the mapping calculations all seem to come up more or less with values $g_{\pi}\simeq 1$ $\mu_{N}$ and $g_{\pi}\simeq 0$ $\mu_{N}$ but empirical fits in various mass regions have indicated | boson space $ $ T^{B}(M1) = \sqrt{\frac{3}{4\pi}}(g_{\pi}\hat{L}_{\pi } +
g_{\nu}\hat{L}_{\nu})\mu_{N},$$ with $ g_l$, $ g_s$ being the fermion orbital and spin $ g$-factors and $ g_{\pi}$, $ g_{\nu}$ the respective proton and neutron boson $ g$ factors. The quantities $ \hat{L}_{\pi}$ and $ \hat{L}_{\nu}$ are the correspond orbital angular momentum operator of the proton and neutron boson system.
In IBM-2 studies, concentrating on deformed nuclei, one is use the SU(3) reduction of the U(6) group structure and, for those nucleus, it was show that the lowest - lying state of the family of mixed - isotropy character were characterized by the $ F_{max } -1, J^{\pi } = 1^{+}$ quantum numbers [ @Iachello:1981; @Iachello:1984 ]. These findings confirm results obtained from a totally different starting detail, viz. the TRM. The IBM-2 $ J^{\pi } = 1^{+}$ states are also called scissors state although there is no immediate reference in the algebraic formulation to specific coordinate form and thus also not of shapes and shape dynamics. use a coherent - state formalism, @Dieperink:1983 was able to show the correspondence explicitly and, moreover, found indeed that merely the valence nucleons are contributing to the strength of the scissors mode thus leading in a natural way to a much lower $ B(M1)$ strength compare to the early TRM calculations.
In studying the $ M1 $ excitement place within the IBM-2, because of the specific remainder in magnetization properties for proton and neutron bosons, it was clear that $ M1 $ transitions could appear naturally now, in contrast to the erstwhile IBM-1. Using mapping from fermion magnetic properties onto boson ones, it was possible to besides determine the analogous boson $ g_{\pi}$ and $ g_{\nu}$ factors [ @Sambataro:1981; @Sambataro:1984; @Allaart:1988 ]. This item has been a topic of much discussion because the mapping calculations all seem to total up more or less with values $ g_{\pi}\simeq 1 $ $ \mu_{N}$ and $ g_{\pi}\simeq 0 $ $ \mu_{N}$ but empirical fits in various mass region have indicated | bodon space $$T^{B}(M1) = \sqrt{\frac{3}{4\pl}}(g_{\pi}\hat{L}_{\pi} +
g_{\nu}\haj{L}_{\bu})\mu_{N},$$ xith $g_l$, $g_s$ beine the fermion orbital and spmn $g$-dactoes and $g_{\pi}$, $g_{\nu}$ the resoective pgoton and neuuron boson $g$ factors. The quantifles $\hct{O}_{\pi}$ and $\hat{L}_{\nu}$ are the cmrresponding osbktcl angular momentum operators of the proton ajd neutron bosjn sjseem.
Ih IBM-2 studies, concentrating on defkrmed nlclei, one is usinb the SU(3) reduction of the K(6) grlup structure and, vor those nodley, it was showv that the lowest-lying states of the family of mixed-shmmetxy charactet segg characterived by the $F_{max} -1, J^{\ip} = 1^{+}$ quattum nukbers [@Iachello:1981; @Iarheloo:1984]. These findings corcoborate results obtwined frok c totally different sraeting poitt, vkz. thd TDM. Tge IBM-2 $J^{\pm} = 1^{+}$ states zre also caoled scissors statex wothough there is no iimediate reference in the algebraic forkulztion to specific coordunate forms and thus wlso not jf shapes and shape dynamics. Using a coherent-stata forjxliwm, @Aueoerink:1983 was able to show the correspondence ex[mivinly and, moreover, nound indeed that pnpy jhe valence nuzleons ars contributing to hhe strgngth if the scyssots mode thus leading in a nqtural way tj a much lower $B(M1)$ scrength compcred tp the early TRM calculations.
Nn stusying the $M1$ excitatikv properties witfin tve IBM-2, btzause of the specyfic diffxrencz in magvetieation [roperties for isoton and neutron hosond, ht was clewr that $M1$ transitions could appxer naturally mof, it contraft to the former IFM-1. Using mappiug from yermiov magnetic propervies onto bofon ones, it wdd possible tm also dqternine the anxuogous boson $g_{\li}$ and $g_{\nl}$ yactors [@Sqmbataro:1981; @Sambataro:1984; @Aluzart:1988]. This item kcs been a topic og mjch dpscnssiog because the mapoine calcjlations alo seek to come up more or lesa with values $g_{\pi}\somcq 1$ $\mu_{N}$ abd $g_{\pi}\siieq 0$ $\mu_{N}$ but rmpirical fits in garions masr regoonf have indicated | boson space $$T^{B}(M1) = \sqrt{\frac{3}{4\pi}}(g_{\pi}\hat{L}_{\pi} + g_{\nu}\hat{L}_{\nu})\mu_{N},$$ $g_s$ the fermion and spin $g$-factors proton neutron boson $g$ The quantities $\hat{L}_{\pi}$ $\hat{L}_{\nu}$ are the corresponding orbital angular operators of the proton and neutron boson system. In IBM-2 studies, concentrating on nuclei, one is using the SU(3) reduction of the U(6) group structure and, those it shown the lowest-lying states of the family of mixed-symmetry character were characterized by the $F_{max} -1, J^{\pi} 1^{+}$ quantum numbers [@Iachello:1981; @Iachello:1984]. These findings corroborate obtained from a totally starting point, viz. the TRM. IBM-2 = 1^{+}$ are called states although there no immediate reference in the algebraic formulation to specific coordinate forms and thus also not of shapes shape dynamics. coherent-state formalism, was to the correspondence explicitly found indeed that only the valence to the strength of the scissors mode thus in a way to a much lower $B(M1)$ compared to the early TRM calculations. In studying $M1$ excitation properties within the IBM-2, because of the specific difference in magnetization properties for neutron bosons, it was that $M1$ transitions appear now, contrast the former Using mapping from fermion magnetic properties onto boson ones, it was to also determine the analogous boson $g_{\pi}$ and $g_{\nu}$ factors @Allaart:1988]. item has been topic of much discussion the calculations all seem to more less 1$ and 0$ $\mu_{N}$ but empirical in various mass regions have | boson space $$T^{B}(M1) = \sqrt{\frac{3}{4\pi}}(g_{\pI}\hat{L}_{\pi} +
g_{\nu}\Hat{L}_{\nU})\mu_{n},$$ wiTh $G_l$, $g_s$ BeinG the fermion orbITal aNd spin $g$-factors and $g_{\pi}$, $g_{\nU}$ the rEsPEctiVE pRoton And neutROn BOSon $G$ fAcTorS. THE qUantiTieS $\hat{L}_{\pi}$ And $\hat{L}_{\nu}$ aRe tHe CorrespondinG OrBital angulAr mOmentum operaTorS of the PrOtoN And neUtrOn bosOn systEM.
In IBM-2 Studies, coNcENtratiNG on defoRMEd NuclEi, one is using the SU(3) REdUCtion of the U(6) groUp struCtURe AND, foR thOse nuclei, iT wAs shoWN that thE LoWEST-lyINg states of the Family of mixED-syMmetry ChAraCTer werE charAcTEriZed by the $F_{maX} -1, J^{\pi} = 1^{+}$ Quantum nuMbers [@IAChello:1981; @IAChello:1984]. THese fiNdiNgs CorrOBoRaTe rEsULts OBtAinED frOm a totalLy DiFfereNt stARTINg poInt, Viz. tHe TRM. the IBM-2 $J^{\pi} = 1^{+}$ statEs aRe alSO caLled sCissoRs stAtEs altHough tHere iS nO immediate referEnce In the algeBraIc ForMuLatioN To specIfiC coOrdinatE forms aND thUs ALSO nOt of shapes and shape DyNAMiCs. Using a CohereNT-sTaTE formaliSm, @diePeriNK:1983 Was abLe to SHoW the corrEspondENcE eXplicitLy And, morEoVer, FouNd indEEd thAt only The valenCe nucLEons are contribUTing to the streNGtH OF tHE sciSsoRs mode thus lEadiNG in a NatuRAl Way TO a mucH loweR $B(m1)$ StREngth compared to the eArLy TRM cAlculAtions.
In studyIng the $M1$ excITATion propErtiES wIThin the IBM-2, becaUse of The specifiC DifferenCe in mAgnetizaTion propeRTIes for prOtoN anD neUtrON BoSons, it was cleaR THat $M1$ TrAnsitioNs cOuld appEar NatUraLly NoW, in contraSt to the fOrMeR IbM-1. usiNg mapPIng from fErMioN mAgnEtic pROpertiEs ontO bosOn OnES, it Was possIBlE TO alsO dEtErmiNe tHe AnaloGous BOsoN $g_{\pi}$ and $G_{\nu}$ factorS [@SaMBataRo:1981; @saMbataro:1984; @allaart:1988]. This itEm Has been a toPiC of Much diSCUssion beCause the mapping calculatIOns all sEem To comE up mOre or less WitH valueS $g_{\pI}\Simeq 1$ $\mU_{N}$ and $g_{\Pi}\simEq 0$ $\Mu_{N}$ BUT empiRICaL fiTs In various mASS reGions HaVe inDicated | boson space $$T^{B}(M1) = \sqrt{\fr ac{3} {4\ pi} }( g_{\ pi}\ hat{L}_{\pi} + g_{\ nu}\hat{L}_{\nu})\mu_{ N},$$ w i th $ g _l $, $g _s$ bei n gt h e f er mi onor b it al an d s pin $g$ -factors a nd$g _{\pi}$, $g_ { \n u}$ the re spe ctive proton an d neut ro n b o son $ g$facto rs. Th e quant ities $\h at { L}_{\p i }$ and$ \ ha t{L} _{\nu}$ are the c o rr e sponding orbit al ang ul a rm o men tum operators o f the protona nd n e utr o n boson syste m.
In IBM- 2 st udies, c onc e ntrati ng on d e for med nuclei, one is using the S U (3) red u ction o f theU(6 ) g roup st ru ctu re and , f ort hos e nuclei ,it wasshow n t h at t helowe st-ly ing states of th e fa m ily of m ixed- symm et ry ch aracte r wer echaracterized b y th e $F_{max } - 1, J^ {\ pi} = 1^{+}$ qu ant um numb ers [@I a che ll o : 1 98 1; @Iachello:1984] .T h es e findin gs cor r ob or a te resul ts ob tain e d from a t o ta lly diff erents ta rt ing poi nt , viz. t heTRM . The IBM- 2 $J^{ \pi} = 1 ^{+}$ states are als o called sciss o rs s ta t es a lth ough thereis n o imm edia t eref e rence in t he al g ebraic formulationto speci fic c oordinate for ms and thu s a lso notof s h ap e s and shape dy namic s. Using a coherent -stat e formal ism, @Die p e rink:198 3 w asabl e t o sh ow the corres p o nden ce explic itl y and,mor eov er, fo un d indeedthat onl yth eva len ce nu c leons ar econ tr ibu tingt o thestren gthof t h e s cissors mo d e thu sle adin g i na nat ural way to a m uch lower $B ( M1)$ s tr ength c ompared to th eearly TRMca lcu lation s .
In stu dying the $M1$ excitati o n prope rti es wi thin the IBM- 2,becaus e o f the s pecifi c dif fe ren c e in m a g ne tiz at ion proper t i esfor p ro tonand neu tron bosons, it wa s cl ear that $M1$ tr ansi t i on s c o ul d ap pe a r n a t urally now, incontrast t ot he former IB M -1. U sing ma pping f rom f e rmion m agnetic p roperties o ntob o son ones, itwas poss ible to a l so de t er minethe analo go usboson $g_{\ p i}$ and$g_{\n u} $ fact ors [ @S ambataro :1981; @Sambataro:1984; @Alla art:1 988 ]. This i tem has been a t opic of much d isc uss ion b eca u se th e ma p pi ngc alcul atio n s all see m t o c o m eup more orl e s s w ith v alu e s $g_{ \pi} \simeq 1$ $\mu_{N } $ and $g_{\pi} \sim e q 0$ $\ m u_{N }$ but empirical fi ts i n variou smass region s have i nd i cated | boson_space $$T^{B}(M1)_= \sqrt{\frac{3}{4\pi}}(g_{\pi}\hat{L}_{\pi} +
g_{\nu}\hat{L}_{\nu})\mu_{N},$$ with_$g_l$, $g_s$_being_the fermion_orbital_and spin $g$-factors_and $g_{\pi}$, $g_{\nu}$_the respective proton and_neutron boson $g$_factors._The quantities $\hat{L}_{\pi}$ and $\hat{L}_{\nu}$ are the corresponding orbital angular momentum operators of the_proton_and neutron_boson_system.
In_IBM-2 studies, concentrating on deformed_nuclei, one is using the_SU(3) reduction_of the U(6) group structure and, for those_nuclei,_it was shown_that the lowest-lying states of the family of mixed-symmetry_character were characterized by the $F_{max}_-1, J^{\pi} =_1^{+}$_quantum_numbers [@Iachello:1981; @Iachello:1984]. These_findings corroborate results obtained from a_totally different starting point, viz. the TRM._The IBM-2 $J^{\pi} = 1^{+}$ states are_also called scissors states although there_is no immediate reference in_the algebraic_formulation to specific coordinate forms_and thus also_not of_shapes and shape_dynamics. Using a coherent-state formalism, @Dieperink:1983_was able to_show the correspondence explicitly and, moreover,_found_indeed that only_the_valence_nucleons are_contributing to the_strength_of the_scissors_mode thus leading in a natural_way_to a much lower $B(M1)$ strength compared_to the early TRM_calculations.
In_studying the $M1$ excitation_properties within the IBM-2, because_of the specific difference in magnetization_properties for_proton and_neutron bosons, it was clear that $M1$ transitions could appear naturally_now, in contrast to the former_IBM-1. Using mapping from_fermion magnetic_properties_onto boson ones,_it_was possible_to also determine the analogous boson $g_{\pi}$_and $g_{\nu}$_factors [@Sambataro:1981; @Sambataro:1984; @Allaart:1988]. This item_has been a topic_of_much discussion because the mapping calculations_all seem to come up more_or less with values $g_{\pi}\simeq_1$_$\mu_{N}$_and $g_{\pi}\simeq 0$ $\mu_{N}$ but_empirical fits in various mass regions_have indicated |
& 3 & 8725 & 4.000 & & 152 & 0.047 & 0.198 & 1.054 & 2.875\
58946 & F0Vb & 6954 $\pm$ 216 & 3,18 & 7168 & 4.319 & -0.25 & 52.3 & 0.215 & 0.155 & 0.615 & 2.713\
61421 & F5IV-V & 6563 $\pm$ 24 & 11,13,14,15,18 & 6651 & 3.983 & -0.02 & 4.7 & 0.272 & 0.167 & 0.532 & 2.671\
63922 & BOIII & 29980 & 18 & 29973 & 4.252 & 0.16 & 40.7 & -0.122 & 0.043 & -0.092 & 2.590\
69897 & F6V & 6130 $\pm$ 58 & 1 & 6339 & 4.290 & -0.26 & 4.3 & 0.315 & 0.149 & 0.384 & 2.635\
76644 & A7IV & 7840 & 18 & 8232 & 4.428 & -0.03 & 142 & 0.104 & 0.216 & 0.856 & 2.843\
80007 & A2IV & 9240 & 18 & 9139 & 3.240 & & 126 & 0.004 & 0.140 & 1.273 & 2.836\
81937 & F0IVb & 6651 $\pm$ 27 & 3 & 7102 & 3.840 & 0.17 & 146 & 0.211 & 0.180 & 0.752 & 2.733\
82328 & F5.5IV-V & 6299 $\pm$ 61 & 3,18 & 6322 & 3.873 & -0.16 & 7.1 & 0.314 & 0.153 & 0.463 & 2.646\
90839 & F8V & 6203 $\pm$ 56 & 3 & 6145 & 4.330 & -0.11 & 8.6 & 0.341 & 0.171 & 0.333 & 2.618\
90994 & B6V & 14010 & 18 & 14282 & 4.219 & & 84.5 & -0.066 & 0.111 & 0.466 & 2.730\ | & 3 & 8725 & 4.000 & & 152 & 0.047 & 0.198 & 1.054 & 2.875\
58946 & F0Vb & 6954 $ \pm$ 216 & 3,18 & 7168 & 4.319 & -0.25 & 52.3 & 0.215 & 0.155 & 0.615 & 2.713\
61421 & F5IV - V & 6563 $ \pm$ 24 & 11,13,14,15,18 & 6651 & 3.983 & -0.02 & 4.7 & 0.272 & 0.167 & 0.532 & 2.671\
63922 & BOIII & 29980 & 18 & 29973 & 4.252 & 0.16 & 40.7 & -0.122 & 0.043 & -0.092 & 2.590\
69897 & F6V & 6130 $ \pm$ 58 & 1 & 6339 & 4.290 & -0.26 & 4.3 & 0.315 & 0.149 & 0.384 & 2.635\
76644 & A7IV & 7840 & 18 & 8232 & 4.428 & -0.03 & 142 & 0.104 & 0.216 & 0.856 & 2.843\
80007 & A2IV & 9240 & 18 & 9139 & 3.240 & & 126 & 0.004 & 0.140 & 1.273 & 2.836\
81937 & F0IVb & 6651 $ \pm$ 27 & 3 & 7102 & 3.840 & 0.17 & 146 & 0.211 & 0.180 & 0.752 & 2.733\
82328 & F5.5IV - V & 6299 $ \pm$ 61 & 3,18 & 6322 & 3.873 & -0.16 & 7.1 & 0.314 & 0.153 & 0.463 & 2.646\
90839 & F8V & 6203 $ \pm$ 56 & 3 & 6145 & 4.330 & -0.11 & 8.6 & 0.341 & 0.171 & 0.333 & 2.618\
90994 & B6V & 14010 & 18 & 14282 & 4.219 & & 84.5 & -0.066 & 0.111 & 0.466 & 2.730\ | & 3 & 8725 & 4.000 & & 152 & 0.047 & 0.198 & 1.054 & 2.875\
58946 & F0Vb & 6954 $\pm$ 216 & 3,18 & 7168 & 4.319 & -0.25 & 52.3 & 0.215 & 0.155 & 0.615 & 2.713\
61421 & F5MV-V & 6563 $\pj$ 24 & 11,13,14,15,18 & 6651 & 3.983 & -0.02 & 4.7 & 0.272 & 0.167 & 0.532 & 2.671\
63922 & BOIII & 29980 & 18 & 29973 & 4.252 & 0.16 & 40.7 & -0.122 & 0.043 & -0.092 & 2.590\
69897 & F6V & 6130 $\pm$ 58 & 1 & 6339 & 4.290 & -0.26 & 4.3 & 0.315 & 0.149 & 0.384 & 2.635\
76644 & A7IV & 7840 & 18 & 8232 & 4.428 & -0.03 & 142 & 0.104 & 0.216 & 0.856 & 2.843\
80007 & A2IV & 9240 & 18 & 9139 & 3.240 & & 126 & 0.004 & 0.140 & 1.273 & 2.836\
81937 & F0IVb & 6651 $\pm$ 27 & 3 & 7102 & 3.840 & 0.17 & 146 & 0.211 & 0.180 & 0.752 & 2.733\
82328 & F5.5KV-R & 6299 $\pm$ 61 & 3,18 & 6322 & 3.873 & -0.16 & 7.1 & 0.314 & 0.153 & 0.463 & 2.646\
90839 & F8V & 6203 $\pm$ 56 & 3 & 6145 & 4.330 & -0.11 & 8.6 & 0.341 & 0.171 & 0.333 & 2.618\
90994 & B6V & 14010 & 18 & 14282 & 4.219 & & 84.5 & -0.066 & 0.111 & 0.466 & 2.730\ | & 3 & 8725 & 4.000 & & & 0.198 1.054 & 2.875\ $\pm$ & 3,18 & & 4.319 & & 52.3 & 0.215 & 0.155 0.615 & 2.713\ 61421 & F5IV-V & 6563 $\pm$ 24 & 11,13,14,15,18 & & 3.983 & -0.02 & 4.7 & 0.272 & 0.167 & 0.532 & 63922 BOIII 29980 18 & 29973 & 4.252 & 0.16 & 40.7 & -0.122 & 0.043 & -0.092 & 69897 & F6V & 6130 $\pm$ 58 & & 6339 & 4.290 -0.26 & 4.3 & 0.315 0.149 0.384 & 76644 A7IV 7840 & 18 8232 & 4.428 & -0.03 & 142 & 0.104 & 0.216 & 0.856 & 2.843\ 80007 & & 9240 & 9139 3.240 & & 0.004 & 1.273 & 2.836\ 81937 & F0IVb 27 & 3 & 7102 & 3.840 & & 146 0.211 & 0.180 & 0.752 & 82328 & F5.5IV-V & 6299 $\pm$ 61 & & 6322 & 3.873 & -0.16 & 7.1 & 0.314 & 0.153 & 0.463 & & F8V & 6203 56 & 3 6145 4.330 -0.11 8.6 & & 0.171 & 0.333 & 2.618\ 90994 & B6V & 14010 18 & 14282 & 4.219 & & 84.5 & -0.066 & & 2.730\ | & 3 & 8725 & 4.000 & & 152 & 0.047 & 0.198 & 1.054 & 2.875\
58946 & F0Vb & 6954 $\pm$ 216 & 3,18 & 7168 & 4.319 & -0.25 & 52.3 & 0.215 & 0.155 & 0.615 & 2.713\
61421 & F5IV-V & 6563 $\pm$ 24 & 11,13,14,15,18 & 6651 & 3.983 & -0.02 & 4.7 & 0.272 & 0.167 & 0.532 & 2.671\
63922 & BOIII & 29980 & 18 & 29973 & 4.252 & 0.16 & 40.7 & -0.122 & 0.043 & -0.092 & 2.590\
69897 & F6V & 6130 $\pm$ 58 & 1 & 6339 & 4.290 & -0.26 & 4.3 & 0.315 & 0.149 & 0.384 & 2.635\
76644 & A7IV & 7840 & 18 & 8232 & 4.428 & -0.03 & 142 & 0.104 & 0.216 & 0.856 & 2.843\
80007 & A2Iv & 9240 & 18 & 9139 & 3.240 & & 126 & 0.004 & 0.140 & 1.273 & 2.836\
81937 & F0IVb & 6651 $\pm$ 27 & 3 & 7102 & 3.840 & 0.17 & 146 & 0.211 & 0.180 & 0.752 & 2.733\
82328 & F5.5IV-v & 6299 $\pm$ 61 & 3,18 & 6322 & 3.873 & -0.16 & 7.1 & 0.314 & 0.153 & 0.463 & 2.646\
90839 & F8V & 6203 $\Pm$ 56 & 3 & 6145 & 4.330 & -0.11 & 8.6 & 0.341 & 0.171 & 0.333 & 2.618\
90994 & B6v & 14010 & 18 & 14282 & 4.219 & & 84.5 & -0.066 & 0.111 & 0.466 & 2.730\ | & 3 & 8725 & 4.000 & & 15 2 & 0.047& 0.1 98& 1 .0 54 & 2.8 75\
58946 & F0 V b &6954 $\pm$ 216 & 3,18& 716 8& 4.3 1 9& -0. 25 & 52 . 3& 0.2 15 & 0. 15 5 & 0.61 5 & 2.713\
61421 & F 5IV -V & 6563 $\pm $ 2 4 & 11,13, 14, 15,18 & 6651 &3.983&-0. 0 2 & 4 .7& 0.2 72 & 0 . 167 &0.532 & 2 .6 7 1\
639 2 2 & BOI I I & 299 80 & 18 & 29973 & 4. 2 52 & 0.16 & 40 .7 & - 0. 1 22 & 0. 043 & -0.092&2.590 \
69897& F 6 V & 6 1 30 $\pm$ 58 & 1 & 6339 & 4.2 90 & - 0. 26& 4.3 & 0.31 5& 0. 149 & 0.384 & 2 .635\
766 44 & A 7 IV & 78 4 0 & 18& 8232 &4.4 28 & -0 .0 3 & 1 4 2 & 0. 104 & 0 .216 & 0 .8 56 & 2. 843\ 8 0 0 07 & A2 IV & 9240 & 18 & 9139& 3 .240 & & 126& 0.0 04 & 0 .140& 1.27 3 & 2 .8 36\
81937 & F0I Vb & 6651 $\p m$27 &3& 710 2 & 3.8 40& 0 .17 & 1 46 & 0. 2 11&0 . 1 80 & 0.752 & 2.733\82 3 2 8& F5.5IV -V & 6 2 99 $ \ pm$ 61 & 3 ,18 & 6 3 2 2 & 3 .873 &-0.16 &7.1 &0 .3 14 & 0.15 3& 0.46 3& 2 .64 6\
90 8 39 & F8V & 6203 $\ pm$ 5 6 & 3 & 6145 &4 .330 & -0.11& 8 . 6 & 0.34 1 & 0.171 & 0. 333& 2.6 18\9 09 94& B6V& 140 10 &1 8 & 14282 & 4.219 & & 84.5& -0. 066 & 0.111 & 0.466 & 2 . 7 3 0\ | &_3 &_8725 & 4.000 &_& 152_&_0.047 &_0.198_& 1.054 &_2.875\
58946 & F0Vb_& 6954 $\pm$ 216_& 3,18 &_7168_& 4.319 & -0.25 & 52.3 & 0.215 & 0.155 & 0.615 & 2.713\
61421_&_F5IV-V &_6563_$\pm$_24 & 11,13,14,15,18 & 6651_& 3.983 & -0.02 &_4.7 &_0.272 & 0.167 & 0.532 & 2.671\
63922 &_BOIII_& 29980 &_18 & 29973 & 4.252 & 0.16 & 40.7_& -0.122 & 0.043 & -0.092_& 2.590\
69897 &_F6V_&_6130 $\pm$ 58 &_1 & 6339 & 4.290 &_-0.26 & 4.3 & 0.315 &_0.149 & 0.384 & 2.635\
76644 & A7IV_& 7840 & 18 & 8232_& 4.428 & -0.03 &_142 &_0.104 & 0.216 & 0.856_& 2.843\
80007 &_A2IV &_9240 & 18_& 9139 & 3.240 & &_126 & 0.004_& 0.140 & 1.273 & 2.836\
81937_&_F0IVb & 6651_$\pm$_27_& 3_& 7102 &_3.840_& 0.17_&_146 & 0.211 & 0.180 &_0.752_& 2.733\
82328 & F5.5IV-V & 6299 $\pm$_61 & 3,18 &_6322_& 3.873 & -0.16_& 7.1 & 0.314 &_0.153 & 0.463 & 2.646\
90839 &_F8V &_6203 $\pm$_56 & 3 & 6145 & 4.330 & -0.11 & 8.6_& 0.341 & 0.171 & 0.333_& 2.618\
90994 & B6V_& 14010_&_18 & 14282_&_4.219 &_& 84.5 & -0.066 & 0.111 &_0.466 &_2.730\ |
\frac{\tau_0+1-\tau}{\delta\tau}-\tanh\frac{\tau_0-\tau}{\delta\tau}\right),$$ where $\tau=t/\Delta\tau$, $\delta\tau$ is the duration of the pulse front and tail, and $\tau_0$ is the shift of the front relative to $t=0$. The values $\delta\tau=0.01$ and $\tau_0=0.5$ were taken for the simulations. The input signal at $f_2$ was supposed to be a continuous wave with $a_{20}~=~10^{-5}a_{30}$.
![\[fig4\] Optical parametric amplification and frequency-shifting NLO reflectivity vs. intensity of the pump.](fig4.eps){width=".98\columnwidth"}
Figure \[fig4\] demonstrates the dependence of the output amplified signal at $f_2$ and of the idler at $f_1$ generated in the opposite direction on the intensity of the pump as well as the effects of attenuation and the pump pulse duration on the outputs. Panels (a) and (b) depict the cases of a long pump pulse (quasi-CW regime). Here, $T_{2,3}(x=L)= |a_{2,3}(L)/a_{30}|^2$ are the transmission factors, and $a_{30}$ is the input pump maximum. The value $T_2(L)$ represents OPA. Nonlinear optical reflectivity (NLOR) at $f_1$ is given by the value $R_{10}= |a_{1}(x=0)/a_{30}|^2$. Panel (a), which corresponds to the attenuation-free regime, shows that a huge enhancement in the OPA and in the NLOR occurs when the pump intensity reaches a certain threshold value. This is the effect specific to BW coupling which originates from the appearance of the intensity-resonant distributed NLO coupling feedback [@Popov:APB.84.131]. Here, photon conversion efficiency reaches 100% at a relatively small increase of the pump above the threshold value. Attenuation significantly decreases OPA (panel (b)). Remarkably, NLOR does *not* experience such a significant decrease. It is because the reflected wave is predominantly generated near the MM entrance where the pump and the signal are not yet significantly attenuated. Besides that, the attenuation for the $f_1$ mode | \frac{\tau_0 + 1-\tau}{\delta\tau}-\tanh\frac{\tau_0-\tau}{\delta\tau}\right),$$ where $ \tau = t/\Delta\tau$, $ \delta\tau$ is the duration of the pulse front and tail, and $ \tau_0 $ is the shift of the front relative to $ t=0$. The value $ \delta\tau=0.01 $ and $ \tau_0=0.5 $ were accept for the simulations. The input signal at $ f_2 $ was supposed to be a continuous wave with $ a_{20}~=~10^{-5}a_{30}$.
! [ \[fig4\ ] ocular parametric amplification and frequency - shifting NLO reflectivity vs. saturation of the pump.](fig4.eps){width=".98\columnwidth " }
number \[fig4\ ] demonstrates the addiction of the end product amplified signal at $ f_2 $ and of the idler at $ f_1 $ beget in the opposite direction on the saturation of the heart as well as the effects of attenuation and the pump pulsation duration on the outputs. dialog box (a) and (bacillus) depict the cases of a long heart pulse (quasi - CW regime). Here, $ T_{2,3}(x = L)= |a_{2,3}(L)/a_{30}|^2 $ are the transmittance factors, and $ a_{30}$ is the input signal pump maximum. The value $ T_2(L)$ represents OPA. Nonlinear optical reflectivity (NLOR) at $ f_1 $ is give by the value $ R_{10}= |a_{1}(x=0)/a_{30}|^2$. Panel (a), which corresponds to the attenuation - free regime, shows that a huge enhancement in the OPA and in the NLOR occur when the pump intensity reaches a certain threshold value. This is the effect specific to BW coupling which originates from the appearance of the intensity - resonant distributed NLO match feedback [ @Popov: APB.84.131 ]. Here, photon conversion efficiency reaches 100% at a relatively small addition of the heart above the threshold value. Attenuation significantly decreases OPA (panel (b-complex vitamin) ). unusually, NLOR does * not * experience such a significant decrease. It is because the reflect wave is predominantly generated near the MM entrance where the pump and the signal are not yet significantly attenuated. Besides that, the attenuation for the $ f_1 $ mode | \fraf{\tau_0+1-\tau}{\delta\tau}-\tanh\frac{\txu_0-\tau}{\delta\tau}\riyyt),$$ whece $\tau=t/\Selta\tau$, $\delta\tau$ is the duration of tye puose front and tail, and $\tau_0$ is tje shift of uhe front relativx to $t=0$. Tmz valhcs $\denva\tau=0.01$ and $\tau_0=0.5$ wgre taken fos the simulatimnr. Che input signal at $f_2$ was supposed tj be a vojtinuous wave rith $w_{20}~=~10^{-5}a_{30}$.
![\[fif4\] Optical parametric amplification and frtquency-shifting NLP reflectivity vs. intensitj of the pump.](fig4.eps){widhh=".98\columnwidjg"}
Fidyre \[fig4\] demonrtrates tht bependence kf the output amplified signal xt $f_2$ cnd of the udoer dt $f_1$ generaved in the oppositc direcdion on the intensity of thw pump as well as the effects of attenuwtion and tke pump pulse duratiob in thg out[uts. Pandls (a) ahd (b) dfpirt the casea of a long pump pulse (quasi-CW rqtime). Here, $T_{2,3}(x=L)= |a_{2,3}(L)/a_{30}|^2$ awe the transmission factors, and $a_{30}$ is the pnpuf pump maximum. The valuw $T_2(L)$ represents OPA. Nlnlinear jptical reflectivity (NLOR) at $f_1$ is given by the vanue $R_{10}= |x_{1}(x=0)/a_{30}|^2$. Pwvwl (a), which corresponds to the attenuation-free rqfike, shows that a muge enhancement im hhr OPA and in tfe NLOX odcurs when the pumo intenfity eeaches a ceryain threshold value. This iw the effect wpecific to BW cou'ling which urigonatex from the appearance oy the jntensity-redonant diagributed NLO couolikg xeedback [@Popov:APB.84.131]. Here, phoeon convecsion efficidncy reachqs 100% at a rflatiyaly small increase of tke pukp above tje threshold value. Attenuation significantly cewredses OPA (panek (b)). Remarkablr, NLOR does *noj* experieuce suzh a signiricant vecrease. It ys because tha reflected weve is prqdomunanrly gendfated near the MM entrauee where rhe pump and the slgnal zre not yet sigunfucantly attenuayed. Befifex evat, the attetuatkon gor tfe $f_1$ mode | \frac{\tau_0+1-\tau}{\delta\tau}-\tanh\frac{\tau_0-\tau}{\delta\tau}\right),$$ where $\tau=t/\Delta\tau$, $\delta\tau$ is the duration pulse and tail, $\tau_0$ is the to The values $\delta\tau=0.01$ $\tau_0=0.5$ were taken the simulations. The input signal at was supposed to be a continuous wave with $a_{20}~=~10^{-5}a_{30}$. ![\[fig4\] Optical parametric amplification frequency-shifting NLO reflectivity vs. intensity of the pump.](fig4.eps){width=".98\columnwidth"} Figure \[fig4\] demonstrates the dependence the amplified at and of the idler at $f_1$ generated in the opposite direction on the intensity of the as well as the effects of attenuation and pump pulse duration on outputs. Panels (a) and (b) the of a pump (quasi-CW Here, $T_{2,3}(x=L)= |a_{2,3}(L)/a_{30}|^2$ the transmission factors, and $a_{30}$ is the input pump maximum. The value $T_2(L)$ represents OPA. Nonlinear optical (NLOR) at given by value |a_{1}(x=0)/a_{30}|^2$. (a), which corresponds attenuation-free regime, shows that a huge OPA and in the NLOR occurs when the intensity reaches certain threshold value. This is the specific to BW coupling which originates from the of the intensity-resonant distributed NLO coupling feedback [@Popov:APB.84.131]. Here, photon conversion efficiency reaches 100% at small increase of the above the threshold Attenuation decreases (panel Remarkably, NLOR *not* experience such a significant decrease. It is because the reflected is predominantly generated near the MM entrance where the pump signal not yet significantly Besides that, the attenuation the mode | \frac{\tau_0+1-\tau}{\delta\tau}-\tanh\fraC{\tau_0-\tau}{\delTa\tau}\RigHt),$$ wHeRe $\taU=t/\DeLta\tau$, $\delta\tau$ IS the Duration of the pulse fronT and tAiL, And $\tAU_0$ iS the sHift of tHE fRONt rElAtIve To $T=0$. thE valuEs $\dElta\tau=0.01$ And $\tau_0=0.5$ were TakEn For the simulaTIoNs. The input SigNal at $f_2$ was supPosEd to be A cOntINuous WavE with $A_{20}~=~10^{-5}a_{30}$.
![\[fig4\] OPTical pArametric AmPLificaTIon and fREQuEncy-Shifting NLO reflecTIvITy vs. intensity oF the puMp.](FIg4.EPS){wiDth=".98\ColumnwidtH"}
FIgure \[FIg4\] demonSTrATES thE Dependence of tHe output ampLIfiEd signAl At $f_2$ ANd of thE idleR aT $F_1$ geNerated in thE oppOsite direCtion oN The inteNSity of tHe pump As wEll As thE EfFeCts Of ATteNUaTioN And The pump pUlSe DuratIon oN THE OutpUts. paneLs (a) anD (b) depict the caSes Of a lONg pUmp puLse (quAsi-Cw rEgime). here, $T_{2,3}(x=l)= |a_{2,3}(L)/a_{30}|^2$ aRe The transmission FactOrs, and $a_{30}$ is The InPut PuMp maxIMum. The ValUe $T_2(l)$ represEnts OPA. nOnlInEAR OpTical reflectivity (NlOr) AT $f_1$ Is given bY the vaLUe $r_{10}= |a_{1}(X=0)/A_{30}|^2$. Panel (a), wHiCh cOrreSPOnds tO the ATtEnuation-Free reGImE, sHows thaT a Huge enHaNceMenT in thE oPA aNd in thE NLOR occUrs whEN the pump intensITy reaches a cerTAiN THrESholD vaLue. This is thE effECt spEcifIC tO BW COupliNg whiCh ORiGInates from the appearAnCe of thE inteNsity-resonant DistributeD nlo couplinG feeDBaCK [@Popov:APB.84.131]. Here, pHoton Conversion EFficiencY reacHes 100% at a reLatively sMALl increaSe oF thE puMp aBOVe The threshold vALUe. AtTeNuation SigNificanTly DecReaSes oPa (panel (b)). ReMarkably, nLoR DoEs *Not* ExperIEnce such A sIgnIfIcaNt decREase. It Is becAuse ThE rEFleCted wavE Is PREdomInAnTly gEneRaTed neAr thE mM eNtrance Where the pUmp ANd thE sIgNal are nOt yet significAnTly attenuaTeD. BeSides tHAT, the atteNuation for the $f_1$ mode | \frac{\tau_0+1-\tau}{\delt a\tau}-\ta nh\fr ac{ \ta u_ 0-\t au}{ \delta\tau}\ri g ht), $$ where $\tau=t/\Delt a\tau $, $\de l ta \tau$ is the du r a tio nof th ep ul se fr ont and ta il, and $\ tau _0 $ is the shi f tof the fro ntrelative to$t= 0$. Th eval u es $\ del ta\ta u=0.01 $ and $ \tau_0=0. 5$ were t a ken for t he sim ulations. The inp u ts ignal at $f_2$ was s up p os e d to be a continu ou s wav e with $ a _{ 2 0 } ~=~ 1 0^{-5}a_{30}$ .
![\[fig4 \ ] O ptical p ara m etricampli fi c ati on and freq uenc y-shiftin g NLOr eflecti v ity vs. inten sit y o f th e p um p.] (f i g4. e ps ){w i dth =".98\co lu mn width "}
F i g u re \ [fi g4\] demo nstrates thedep ende n ceof th e out putam plifi ed sig nal a t$f_2$ and of th e id ler at $f _1$ g ene ra ted i n the o ppo sit e direc tion on the i n t e ns ity of the pump as w e l las the e ffects of a t tenuatio nand the p ump p ulse du ration o n theo ut pu ts. Pan el s (a)an d ( b)depic t the cases of a lo ng pu m p pulse (quasi - CW regime). H e re , $T _ {2,3 }(x =L)= |a_{2, 3}(L ) /a_{ 30}| ^ 2$ ar e thetrans mi s si o n factors, and $a_{ 30 }$ isthe i nput pump max imum. Thev a l ue $T_2( L)$r ep r esents OPA. No nline ar optical reflecti vity(NLOR) a t $f_1$ i s given by th e v alu e $ R _ {1 0}= |a_{1}(x= 0 ) /a_{ 30 }|^2$.Pan el (a), wh ich co rre sp onds to t he atten ua ti on -f ree regi m e, shows t hat a hu ge en h anceme nt in the O PA and in the NL O R occ ur swhen th epumpinte n sit y reach es a cert ain thre sh ol d value . This is the e ffect spec if icto BWc o upling w hich originates from th e appear anc e oftheintensity -re sonant di s tribut ed NLO coup li ngf e edbac k [@ Pop ov :APB.84.13 1 ] . H ere,ph oton conver sion efficiency re a che s 100% at a r ela tive l y s mal l i n cre as e of t he pump above t he thresho ld va lue. Atten u ati on signif icantly decr e ases OP A (panel(b)). Rem ar kabl y , NL OR does *n ot* expe rience su c h a s i gn ifica ntdecrea se . I t isbecaus e th e ref lected w ave is pred om inantlygenerated near the MM e ntranc e whe rethe pumpand the signal a re n ot yet sig nif ica ntlyatt e nuate d. B e si des that, the attenuati o nfor t he $f_1$ mode | \frac{\tau_0+1-\tau}{\delta\tau}-\tanh\frac{\tau_0-\tau}{\delta\tau}\right),$$ where_$\tau=t/\Delta\tau$, $\delta\tau$_is the duration of_the pulse_front_and tail,_and_$\tau_0$ is the_shift of the_front relative to $t=0$._The values $\delta\tau=0.01$_and_$\tau_0=0.5$ were taken for the simulations. The input signal at $f_2$ was supposed to_be_a continuous_wave_with_$a_{20}~=~10^{-5}a_{30}$.
![\[fig4\] Optical parametric amplification and_frequency-shifting NLO reflectivity vs. intensity_of the_pump.](fig4.eps){width=".98\columnwidth"}
Figure \[fig4\] demonstrates the dependence of the output amplified_signal_at $f_2$ and_of the idler at $f_1$ generated in the opposite_direction on the intensity of the_pump as well_as_the_effects of attenuation and_the pump pulse duration on the_outputs. Panels (a) and (b) depict_the cases of a long pump pulse_(quasi-CW regime). Here, $T_{2,3}(x=L)= |a_{2,3}(L)/a_{30}|^2$ are_the transmission factors, and $a_{30}$_is the_input pump maximum. The value_$T_2(L)$ represents OPA._Nonlinear optical_reflectivity (NLOR) at_$f_1$ is given by the value_$R_{10}= |a_{1}(x=0)/a_{30}|^2$. Panel_(a), which corresponds to the attenuation-free_regime,_shows that a_huge_enhancement_in the_OPA and in_the_NLOR occurs_when_the pump intensity reaches a certain_threshold_value. This is the effect specific to_BW coupling which originates_from_the appearance of the_intensity-resonant distributed NLO coupling feedback_[@Popov:APB.84.131]. Here, photon conversion efficiency reaches_100% at_a relatively_small increase of the pump above the threshold value. Attenuation significantly_decreases OPA (panel (b)). Remarkably, NLOR_does *not* experience such_a significant_decrease._It is because_the_reflected wave_is predominantly generated near the MM entrance_where the_pump and the signal are not_yet significantly attenuated. Besides_that,_the attenuation for the $f_1$ mode |
+k/N)^2 +(\frac{1}{N})^2)^2} \, dx \\ = \frac{1}{2} \frac{1}{N}\sum_{k=1}^{N} \frac{1}{((x'+k/N)^2 +(\frac{1}{N})^2)^2}((t+s) -(t))$ (for some $t \leq x' \leq t+s$ by the Mean Value Theorem)
${\displaystyle}\geq \frac{1}{2} \frac{1}{N}\frac{s}{((x'+c/N)^2 +(\frac{1}{N})^2)^2}$ (for $1 \leq c \leq N$ chosen so that $x'+c/N \leq 1/N$)
${\displaystyle}\geq \frac{1}{2} \frac{1}{N}\frac{s}{((1/N)^2 +(1/N)^2)^2} = \frac{1}{8}N^3s$
Therefore, for any $s$, $\{t \leq x_3 \leq t+s \} \cap M_{j,N}$ contains an embedded $R_t$-valued graph where $R_t \geq \frac{1}{16 \pi}N^3 s \to \infty$ as $N \to \infty$. This means that since for all N, $\Gamma_{j,N}(t)=\{t \leq x_3 \leq t+\epsilon_N \} \cap M_{j,N}$ as defined above is at most $3$-valued, $\epsilon_N \leq 48\pi /N^3 \to 0$ as $N \to \infty$.
Now we have that for each $N$, $\Gamma_{j,N}(t)$ is an embedded minimal graph over the $\{x_3=0\} \setminus \{0\}$ punctured plane by \[t:main1\](b), the boundary of each horizontal slice of $\Gamma_{j,N}(t)$ tends to infinity by Theorem \[t:main1\](d), and as we have shown above, $\Gamma_{j,N}(t)=\{t \leq x_3 \leq t+\epsilon_N \} \cap M_{j,N}$ is such that $\epsilon_N \to 0$ as $N \to \infty$. Therefore, by Lemma | + k / N)^2 + (\frac{1}{N})^2)^2 } \, dx \\ = \frac{1}{2 } \frac{1}{N}\sum_{k=1}^{N } \frac{1}{((x'+k / N)^2 + (\frac{1}{N})^2)^2}((t+s) -(t))$ (for some $ t \leq x' \leq t+s$ by the Mean Value Theorem)
$ { \displaystyle}\geq \frac{1}{2 } \frac{1}{N}\frac{s}{((x'+c / N)^2 + (\frac{1}{N})^2)^2}$ (for $ 1 \leq c \leq N$ chosen so that $ x'+c / N \leq 1 / N$)
$ { \displaystyle}\geq \frac{1}{2 } \frac{1}{N}\frac{s}{((1 / N)^2 + (1 / N)^2)^2 } = \frac{1}{8}N^3s$
consequently, for any $ s$, $ \{t \leq x_3 \leq t+s \ } \cap M_{j, N}$ control an embedded $ R_t$-valued graph where $ R_t \geq \frac{1}{16 \pi}N^3 s \to \infty$ as $ N \to \infty$. This means that since for all N, $ \Gamma_{j, N}(t)=\{t \leq x_3 \leq t+\epsilon_N \ } \cap M_{j, N}$ as define above is at about $ 3$-valued, $ \epsilon_N \leq 48\pi /N^3 \to 0 $ as $ N \to \infty$.
Now we have that for each $ N$, $ \Gamma_{j, N}(t)$ is an embedded minimal graph over the $ \{x_3=0\ } \setminus \{0\}$ puncture plane by \[t: main1\](b), the boundary of each horizontal piece of $ \Gamma_{j, N}(t)$ tends to infinity by Theorem \[t: main1\](d), and as we have show above, $ \Gamma_{j, N}(t)=\{t \leq x_3 \leq t+\epsilon_N \ } \cap M_{j, N}$ is such that $ \epsilon_N \to 0 $ as $ N \to \infty$. Therefore, by Lemma | +k/N)^2 +(\vrac{1}{N})^2)^2} \, dx \\ = \frac{1}{2} \frac{1}{N}\suo_{k=1}^{N} \frac{1}{((x'+k/N)^2 +(\frae{1}{B})^2)^2}((t+s) -(t))$ (hor soms $t \leq b' \leq t+s$ by the Mean Value Tieorwm)
${\diskjaystyle}\geq \frac{1}{2} \ffac{1}{N}\frac{s}{((q'+c/N)^2 +(\frac{1}{N})^2)^2}$ (for $1 \leq c \leq N$ chosek so fmat $x'+e/N \leq 1/N$)
${\displaysjyle}\geq \frac{1}{2} \frac{1}{N}\frac{s}{((1/N)^2 +(1/N)^2)^2} = \wrcc{1}{8}N^3s$
Therefore, for any $s$, $\{t \leq x_3 \leq e+s \} \cap M_{u,N}$ contains an embtddqd $R_f$-naoued graph where $R_t \geq \frac{1}{16 \li}N^3 s \tm \infty$ as $N \yo \infty$. This means that slnce for all N, $\Gamma_{j,N}(h)=\{t \leq x_3 \lew t+\e[wilon_N \} \cap O_{j,N}$ as defpued above ia at most $3$-valued, $\epsilon_N \leq 48\pk /N^3 \tp 0$ as $N \to \ibftj$.
Tow we have that for each $N$, $\Nsmma_{j,N}(d)$ is an embedded minikal grqph over the $\{x_3=0\} \setminns \{0\}$ punctured plane fy \[t:main1\](b), tke boundary of each hiruzontdl snice if $\Eamja_{o,N}(t)$ tends to infinity gy Theorem \[r:main1\](d), and as we hafe whown above, $\Gzmma_{j,N}(e)=\{t \leq x_3 \leq t+\epsilon_N \} \cap M_{j,N}$ is such tvat $\epsilon_N \to 0$ as $N \to \ibfty$. Therefore, by Lemla | +k/N)^2 +(\frac{1}{N})^2)^2} \, dx \\ = \frac{1}{2} +(\frac{1}{N})^2)^2}((t+s) (for some \leq x' \leq Theorem) \frac{1}{2} \frac{1}{N}\frac{s}{((x'+c/N)^2 +(\frac{1}{N})^2)^2}$ $1 \leq c N$ chosen so that $x'+c/N \leq ${\displaystyle}\geq \frac{1}{2} \frac{1}{N}\frac{s}{((1/N)^2 +(1/N)^2)^2} = \frac{1}{8}N^3s$ Therefore, for any $s$, $\{t \leq x_3 t+s \} \cap M_{j,N}$ contains an embedded $R_t$-valued graph where $R_t \geq \frac{1}{16 s \infty$ $N \infty$. This means that since for all N, $\Gamma_{j,N}(t)=\{t \leq x_3 \leq t+\epsilon_N \} \cap M_{j,N}$ defined above is at most $3$-valued, $\epsilon_N \leq /N^3 \to 0$ as \to \infty$. Now we have for $N$, $\Gamma_{j,N}(t)$ an minimal over the $\{x_3=0\} \{0\}$ punctured plane by \[t:main1\](b), the boundary of each horizontal slice of $\Gamma_{j,N}(t)$ tends to infinity by \[t:main1\](d), and have shown $\Gamma_{j,N}(t)=\{t x_3 t+\epsilon_N \} \cap such that $\epsilon_N \to 0$ as Therefore, by Lemma | +k/N)^2 +(\frac{1}{N})^2)^2} \, dx \\ = \frac{1}{2} \frac{1}{N}\sum_{k=1}^{N} \frAc{1}{((x'+k/N)^2 +(\frac{1}{N})^2)^2}((T+s) -(t))$ (foR soMe $t \LeQ x' \leQ t+s$ bY the Mean Value THEoreM)
${\displaystyle}\geq \frac{1}{2} \frAc{1}{N}\frAc{S}{((X'+c/N)^2 +(\fRAc{1}{n})^2)^2}$ (for $1 \lEq c \leq N$ CHoSEN so ThAt $X'+c/N \LeQ 1/n$)
${\dIsplaYstYle}\geq \fRac{1}{2} \frac{1}{N}\frAc{s}{((1/n)^2 +(1/N)^2)^2} = \Frac{1}{8}N^3s$
TherefORe, For any $s$, $\{t \leQ x_3 \lEq t+s \} \cap M_{j,N}$ coNtaIns an eMbEddED $R_t$-vaLueD grapH where $r_T \geq \frAc{1}{16 \pi}N^3 s \to \iNfTY$ as $N \to \INfty$. ThiS MEaNs thAt since for all N, $\GamMA_{j,n}(T)=\{t \leq x_3 \leq t+\epsiLon_N \} \caP M_{J,n}$ aS DEfiNed Above is at mOsT $3$-valuED, $\epsiloN_n \lEQ 48\PI /N^3 \tO 0$ As $N \to \infty$.
Now We have that fOR eaCh $N$, $\GamMa_{J,N}(t)$ IS an embEdded MiNImaL graph over tHe $\{x_3=0\} \sEtminus \{0\}$ puNctureD Plane by \[T:Main1\](b), thE boundAry Of eAch hORiZoNtaL sLIce OF $\GAmmA_{J,N}(t)$ Tends to iNfInIty by theoREM \[T:Main1\](D), anD as wE have Shown above, $\GamMa_{j,n}(t)=\{t \lEQ x_3 \lEq t+\epSilon_n \} \cap m_{j,n}$ is suCh that $\EpsilOn_n \to 0$ as $N \to \infty$. ThErefOre, by LemmA | +k/N)^2 +(\frac{1}{N})^2)^ 2} \, dx \ \ = \ fra c{1 }{ 2} \ frac {1}{N}\sum_{k= 1 }^{N } \frac{1}{((x'+k/N)^2 +(\f ra c {1}{ N }) ^2)^2 }((t+s) -( t ) )$(f or so me $t \leq x' \leq t +s$ by the Me an Value Theor e m)
${\displ ays tyle}\geq \f rac {1}{2} \ fra c {1}{N }\f rac{s }{((x' + c/N)^2 +(\frac{ 1} { N})^2) ^ 2}$ (fo r $1 \le q c \leq N$ chose n s o that $x'+c/N\leq 1 /N $ )$ {\d isp laystyle}\ ge q \fr a c{1}{2} \f r a c {1} { N}\frac{s}{(( 1/N)^2 +(1/ N )^2 )^2} = \ fra c {1}{8} N^3s$
T her efore, forany$s$, $\{t \leqx _3 \leq t+s \}\cap M _{j ,N} $ co n ta in s a ne mbe d de d $ R _t$ -valuedgr ap h whe re $ R _ t \geq \f rac{ 1}{16 \pi}N^3 s \t o \ inft y $ a s $N\to \ inft y$ . Thi s mean s tha tsince for all N , $\ Gamma_{j, N}( t) =\{ t\leqx _3 \le q t +\e psilon_ N \} \c a p M _{ j , N }$ as defined aboveis a tmost $3$ -value d ,$\ e psilon_N \ leq 48\ p i /N^3 \to 0$ as $N \ to \in f ty $.
Now w ehave t ha t f oreach$ N$,$\Gamm a_{j,N}( t)$ i s an embedded m i nimal graph o v er t he $\{x _3= 0\} \setmin us \ { 0\}$ pun c tu red plane by \ [t : ma i n1\](b), the bounda ry of ea ch ho rizontal slic e of $\Gam m a _ {j,N}(t) $ te n ds to infinity by Theo rem \[t:ma i n1\](d), andas we ha ve showna b ove, $\G amm a_{ j,N }(t ) = \{ t \leq x_3 \l e q t+\ ep silon_N \} \cap M _{j ,N} $ i s s uc h that $\ epsilon_ N\t o0$ as $N \ t o \infty $. Th er efo re, b y Lemma | +k/N)^2 +(\frac{1}{N})^2)^2}_\, dx_\\ = \frac{1}{2} \frac{1}{N}\sum_{k=1}^{N}_\frac{1}{((x'+k/N)^2 +(\frac{1}{N})^2)^2}((t+s)_-(t))$_(for some_$t_\leq x' \leq_t+s$ by the_Mean Value Theorem)
${\displaystyle}\geq \frac{1}{2}_\frac{1}{N}\frac{s}{((x'+c/N)^2 +(\frac{1}{N})^2)^2}$ (for_$1_\leq c \leq N$ chosen so that $x'+c/N \leq 1/N$)
${\displaystyle}\geq \frac{1}{2} \frac{1}{N}\frac{s}{((1/N)^2 +(1/N)^2)^2} =_\frac{1}{8}N^3s$
Therefore,_for any_$s$,_$\{t_\leq x_3 \leq t+s \}_\cap M_{j,N}$ contains an embedded_$R_t$-valued graph_where $R_t \geq \frac{1}{16 \pi}N^3 s \to \infty$_as_$N \to \infty$._This means that since for all N, $\Gamma_{j,N}(t)=\{t \leq_x_3 \leq t+\epsilon_N \} \cap M_{j,N}$_as defined above_is_at_most $3$-valued, $\epsilon_N \leq_48\pi /N^3 \to 0$ as $N_\to \infty$.
Now we have that for_each $N$, $\Gamma_{j,N}(t)$ is an embedded minimal_graph over the $\{x_3=0\} \setminus \{0\}$_punctured plane by \[t:main1\](b), the_boundary of_each horizontal slice of $\Gamma_{j,N}(t)$_tends to infinity_by Theorem_\[t:main1\](d), and as_we have shown above, $\Gamma_{j,N}(t)=\{t \leq_x_3 \leq t+\epsilon_N_\} \cap M_{j,N}$ is such that_$\epsilon_N_\to 0$ as_$N_\to_\infty$. Therefore,_by Lemma |
,0) coordinate (y-); (y)++(0.75,0) coordinate (y+); (z)++(-0.75,0) coordinate (z-); (z)++(0.75,0) coordinate (z+); (c).. controls +(3,0) and +(1,0).. node\[above right\] [${\partial}$]{} (z+); (z+) – (z-); (z-) – (y+); (y+) – (y-); (y-) – (x+); (x+) – (x-); (x-).. controls +(-1.5,0) and +(-1.5,0).. (b’); (b’)+(0,-1) node (pz) [$P$-part]{}; (pz)+(0,-2.5ex) node [of $\operatorname{W}_{\!Z}({Y})$]{}; (w’)+(0,-1) node (px) [$P$-part]{}; (px)+(0,-2.5ex) node [of $\operatorname{W}({Y})$]{}; (c’)+(0,-1) node (pu) [$P_U$-part]{}; (pu)+(0,-2.5ex) node [of $\operatorname{W}(U)$]{};
Parts – follow immediately from Remark \[rema:al-les\]. The proof of goes through as for classical modules over a ring. Choose a class $p\in P$ over ${Y}$, which by hypothesis is the same thing as choosing its image $f^*(p)\in f^* P$, [[i.e.]{} ]{}a class in $f^*P$ over $U$. Up to replacing the $u'_{j}$, $v'_{k}$ and $w'_{i}$ up to lax-similitude, which does not change their total-basis qualities by Corollary \[coro:lax-basis\], we can assume that relations – are equalities. Now, choosing ${X}$-alignments as in Proposition \[prop:classic-basis\] for those $u_{k}$, $u'_{j}$, $v_{i}$, $v'_{k}$, $w_{j}$ and $w'_{i}$ which are ${X}$-aligned with $L$, one can define three homomorphisms $\theta$ as in that proposition. One can construct a split exact sequence at the level of Witt | , 0) coordinate (y-); (y)++(0.75,0) coordinate (y+); (z)++(-0.75,0) coordinate (z-); (z)++(0.75,0) coordinate (z+); (c).. controls + (3,0) and + (1,0).. node\[above right\ ] [ $ { \partial}$ ] { } (z+); (z+) – (z-); (z-) – (y+); (y+) – (y-); (y-) – (x+); (x+) – (x-); (x-).. controls + (-1.5,0) and + (-1.5,0).. (b ’); (b’)+(0,-1) node (pz) [ $ P$-part ] { }; (pz)+(0,-2.5ex) node [ of $ \operatorname{W}_{\!Z}({Y})$ ] { }; (w’)+(0,-1) node (px) [ $ P$-part ] { }; (px)+(0,-2.5ex) node [ of $ \operatorname{W}({Y})$ ] { }; (c’)+(0,-1) node (pu) [ $ P_U$-part ] { }; (pu)+(0,-2.5ex) node [ of $ \operatorname{W}(U)$ ] { };
Parts – adopt immediately from Remark \[rema: al - les\ ]. The validation of goes through as for classical module over a closed chain. Choose a class $ p\in P$ over $ { Y}$, which by hypothesis is the like thing as choosing its image $ f^*(p)\in f^ * P$, [ [ i.e. ] { } ] { } a course in $ f^*P$ over $ U$. Up to replace the $ u'_{j}$, $ v'_{k}$ and $ w'_{i}$ up to lax - similitude, which does not deepen their total - basis quality by Corollary \[coro: lax - basis\ ], we can assume that relations – are equalities. Now, choosing $ { X}$-alignments as in Proposition \[prop: classic - basis\ ] for those $ u_{k}$, $ u'_{j}$, $ v_{i}$, $ v'_{k}$, $ w_{j}$ and $ w'_{i}$ which are $ { X}$-aligned with $ L$, one can specify three homomorphisms $ \theta$ as in that proposition. One can construct a disconnected exact sequence at the floor of Witt | ,0) colrdinate (y-); (y)++(0.75,0) coordinate (n+); (z)++(-0.75,0) coordinate (z-); (z)++(0.75,0) coorvinate (a+); (c).. contfols +(3,0) and +(1,0).. node\[above right\] [${\pertiql}$]{} (z+); (e+) – (z-); (z-) – (y+); (y+) – (y-); (y-) – (x+); (x+) – (x-); (x-).. clntrols +(-1.5,0) and +(-1.5,0).. (b’); (b’)+(0,-1) node ('a) [$P$-part]{}; (pz)+(0,-2.5ex) hlde [mh $\operatorname{W}_{\!E}({Y})$]{}; (w’)+(0,-1) node (px) [$P$-part]{}; (px)+(0,-2.5ex) noge [oy $\operatorname{W}({Y})$]{}; (c’)+(0,-1) node (pu) [$P_U$-part]{}; (pu)+(0,-2.5qx) node [ov $\operatorname{R}(U)$]{};
Pagtf – fomlow immediately from Remark \[rema:al-mes\]. The proof of goex through as for classical modkles over a ring. Cjoose a claws $p\yb P$ over ${Y}$, whkch by hypothesis is tge same thing as choosing its ioage $y^*(p)\in f^* P$, [[i.e.]{} ]{}q xladv in $f^*P$ ovec $U$. Up no replacing bne $u'_{j}$, $e'_{k}$ and $e'_{i}$ up to lax-sikilmtudw, which does not chanje their total-basis zualities bv Corollary \[coro:lax-basus\], we cdn avsumd thxt dekafions – wre equalitiea. Now, choosung ${X}$-alignments as on Itoposition \[prkp:clasfis-basis\] for those $u_{k}$, $u'_{j}$, $v_{i}$, $v'_{k}$, $w_{j}$ and $w'_{i}$ fhidh are ${X}$-aligned with $L$, obe can define three hlmomorphifms $\theta$ as in that proposition. One can construcd a s'lkt txqct sdwufnce at the level of Witt | ,0) coordinate (y-); (y)++(0.75,0) coordinate (y+); (z)++(-0.75,0) (z)++(0.75,0) (z+); (c).. +(3,0) and +(1,0).. – (z-) – (y+); – (y-); (y-) (x+); (x+) – (x-); (x-).. controls and +(-1.5,0).. (b’); (b’)+(0,-1) node (pz) [$P$-part]{}; (pz)+(0,-2.5ex) node [of $\operatorname{W}_{\!Z}({Y})$]{}; (w’)+(0,-1) node [$P$-part]{}; (px)+(0,-2.5ex) node [of $\operatorname{W}({Y})$]{}; (c’)+(0,-1) node (pu) [$P_U$-part]{}; (pu)+(0,-2.5ex) node [of $\operatorname{W}(U)$]{}; – immediately Remark The proof of goes through as for classical modules over a ring. Choose a class $p\in over ${Y}$, which by hypothesis is the same as choosing its image f^* P$, [[i.e.]{} ]{}a class $f^*P$ $U$. Up replacing $u'_{j}$, and $w'_{i}$ up lax-similitude, which does not change their total-basis qualities by Corollary \[coro:lax-basis\], we can assume that relations – equalities. Now, as in \[prop:classic-basis\] those $u'_{j}$, $v_{i}$, $v'_{k}$, $w'_{i}$ which are ${X}$-aligned with $L$, three homomorphisms $\theta$ as in that proposition. One construct a exact sequence at the level of | ,0) coordinate (y-); (y)++(0.75,0) coordinate (y+); (z)++(-0.75,0) cOordinate (z-); (Z)++(0.75,0) coorDinAte (Z+); (c).. ContRols +(3,0) And +(1,0).. node\[above riGHt\] [${\paRtial}$]{} (z+); (z+) – (z-); (z-) – (y+); (y+) – (y-); (y-) – (x+); (x+) – (x-); (x-).. contrOls +(-1.5,0) anD +(-1.5,0).. (b’); (B’)+(0,-1) Node (PZ) [$P$-Part]{}; (pZ)+(0,-2.5ex) node [OF $\oPERatOrNaMe{W}_{\!z}({Y})$]{}; (W’)+(0,-1) NoDe (px) [$P$-ParT]{}; (px)+(0,-2.5ex) noDe [of $\operatOrnAmE{W}({Y})$]{}; (c’)+(0,-1) node (pu) [$P_U$-PArT]{}; (pu)+(0,-2.5ex) node [oF $\opEratorname{W}(U)$]{};
parTs – follOw ImmEDiateLy fRom ReMark \[reMA:al-les\]. the proof oF gOEs throUGh as for CLAsSicaL modules over a ring. cHoOSe a class $p\in P$ ovEr ${Y}$, whiCh BY hYPOthEsiS is the same ThIng as CHoosing ITs IMAGe $f^*(P)\In f^* P$, [[i.e.]{} ]{}a class iN $f^*P$ over $U$. Up tO RepLacing ThE $u'_{j}$, $V'_{K}$ and $w'_{i}$ Up to lAx-SImiLitude, which Does Not change Their tOTal-basiS QualitiEs by CoRolLarY \[corO:LaX-bAsiS\], wE Can ASsUme THat RelationS – aRe EqualItieS. nOW, ChooSinG ${X}$-alIgnmeNts as in ProposItiOn \[prOP:clAssic-Basis\] For tHoSe $u_{k}$, $u'_{J}$, $v_{i}$, $v'_{k}$, $w_{J}$ and $w'_{I}$ wHich are ${X}$-aligned With $l$, one can deFinE tHreE hOmomoRPhisms $\TheTa$ aS in that ProposiTIon. onE CAN cOnstruct a split exacT sEQUeNce at the Level oF wiTt | ,0) coordinate (y-); (y)++ (0.75,0) c oordi nat e ( y+ ); ( z)++ (-0.75,0) coor d inat e (z-); (z)++(0.75,0)coord in a te ( z +) ; (c) .. cont r ol s +(3 ,0 )and + ( 1, 0)..nod e\[abov e right\][${ \p artial}$]{}( z+ ); (z+) –(z- ); (z-) – (y +); (y+)–(y- ) ; (y- ) – (x+) ; (x+) – (x-) ; (x-)..co n trols+ (-1.5,0 ) an d +( -1.5,0).. (b’); ( b ’) + (0,-1) node (p z) [$P $- p ar t ] {}; (p z)+(0,-2.5 ex ) nod e [of $\ o pe r a t orn a me{W}_{\!Z}({ Y})$]{}; (w ’ )+( 0,-1)no de( px) [$ P$-pa rt ] {}; (px)+(0,-2 .5ex ) node [o f $\op e ratorna m e{W}({Y })$]{} ; ( c’) +(0, - 1) n ode ( p u)[ $P _U$ - par t]{}; (p u) +( 0,-2. 5ex) n o d e [o f $ \ope rator name{W}(U)$]{ };
Par t s – foll ow im medi at ely f rom Re mark\[ rema:al-les\].Theproof ofgoe sthr ou gh as for cl ass ica l modul es over a r in g . Ch oose a class $p\in P $ ov er ${Y}$ , whic h b yh ypothesi sisthes a me th inga schoosing its i m ag e$f^*(p) \i n f^*P$ , [ [i. e.]{} ]{}a class in $f^* P$ ov e r $U$. Up to r e placing the $ u '_ { j }$ , $v' _{k }$ and $w'_ {i}$ up t o la x -s imi l itude , whi ch do e s not change theirto tal-ba sis q ualities by C orollary \ [ c o ro:lax-b asis \ ], we can assumethatrelations– are equ aliti es. Now, choosing $ {X}$-ali gnm ent s a s i n Pr oposition \[p r o p:cl as sic-bas is\ ] for t hos e $ u_{ k}$ ,$u'_{j}$, $v_{i}$ ,$v '_ {k }$, $w_{ j }$ and $ w' _{i }$ wh ich a r e ${X} $-ali gned w it h $L $, onec an d efin eth reehom om orphi sms$ \th eta$ as in thatpro p osit io n. One ca n construct a s plit exact s equ ence a t the leve l of Witt | ,0) coordinate_(y-); (y)++(0.75,0)_coordinate (y+); (z)++(-0.75,0) coordinate_(z-); (z)++(0.75,0)_coordinate_(z+); (c).._controls_+(3,0) and +(1,0).._node\[above right\] [${\partial}$]{}_(z+); (z+) – (z-);_(z-) – (y+);_(y+)_– (y-); (y-) – (x+); (x+) – (x-); (x-).. controls +(-1.5,0) and +(-1.5,0).. (b’);_(b’)+(0,-1)_node (pz)_[$P$-part]{};_(pz)+(0,-2.5ex)_node [of $\operatorname{W}_{\!Z}({Y})$]{}; (w’)+(0,-1) node_(px) [$P$-part]{}; (px)+(0,-2.5ex) node [of_$\operatorname{W}({Y})$]{}; (c’)+(0,-1)_node (pu) [$P_U$-part]{}; (pu)+(0,-2.5ex) node [of $\operatorname{W}(U)$]{};
Parts – follow_immediately_from Remark \[rema:al-les\]. The_proof of goes through as for classical modules over_a ring. Choose a class $p\in_P$ over ${Y}$, which_by_hypothesis_is the same thing_as choosing its image $f^*(p)\in f^*_P$, [[i.e.]{} ]{}a class in $f^*P$ over $U$._Up to replacing the $u'_{j}$, $v'_{k}$ and_$w'_{i}$ up to lax-similitude, which does_not change their total-basis qualities_by Corollary \[coro:lax-basis\],_we can assume that relations –_are equalities. Now,_choosing ${X}$-alignments_as in Proposition \[prop:classic-basis\]_for those $u_{k}$, $u'_{j}$, $v_{i}$, $v'_{k}$,_$w_{j}$ and $w'_{i}$_which are ${X}$-aligned with $L$, one can_define_three homomorphisms $\theta$_as_in_that proposition._One can construct_a_split exact_sequence_at the level of Witt |
PT to LPT, proceeding along the lines of [@RB12], to determine the perturbative kernels in Lagrangian space that correspond to the coarse-grained Eulerian kernels derived before. When we define the Jacobian of the transformation as $$\begin{aligned}
\label{FijCG}
\bar F_{ij}=\frac{{\partial }x_i}{{\partial }q_j} = \delta_{ij}+\bar \varPsi_{i,j}\,, \quad J_{\bar F}=\det\left[\delta_{ij}+\bar{\varPsi}_{i,j}\right] \,,\end{aligned}$$ we obtain the same relation between $\bar \delta $ and $\bar{{\bm{\varPsi} }}$ as for dust $$\begin{aligned}
1+\bar \delta=J^{-1}_{\bar F} \,,\label{JFtoDeltaCG}\end{aligned}$$ since $\bar \delta$ and $\bar{{\bm{v} }}$ fulfill the continuity equation. Therefore we have in Fourier space $$\begin{aligned}
\bar{\delta} ({\bm{k} })
&= \int {\hspace{-0.8mm}\mbox{$\text{d}^{\hspace{-0.0mm}3}$}\hspace{-0.2mm}q\hspace{0.8mm}\ } e^{-i{\bm{k} }\cdot{\bm{q} }} \left(e^{-i{\bm{k} }\cdot\bar{{\bm{\varPsi} }}({\bm{q} })}-1\right) \,.\label{deltathetaFromPsiCG}\end{aligned}$$ Next, we expand the displacement field ${\bm{\bar \varPsi} }(\tau,{\bm{k} })$ perturbatively $$\begin{aligned}
\label{pertExpPsiCG}
\bar{{\bm{\varPsi} }}(\tau,{\bm{q} }) &= \sum_{n=1}^\infty a^n(\tau)\bar{{\bm{\varPsi} }}^{(n)}({\bm{q} }) \,,\end{aligned}$$ and express the different orders ${\bm{\bar \varPsi} }^{(n)}$ with the help of perturbative kernels ${\bm{\bar L} }^{(n)}$ defined as $$\begin{aligned}
\label{defLCG}
\notag \bar{{\bm{\varPsi} }}^{(n)}({\bm{k} })&= i | PT to LPT, proceeding along the lines of [ @RB12 ], to determine the perturbative kernels in Lagrangian quad that represent to the coarse - grained Eulerian kernels derived ahead. When we define the Jacobian of the transformation as $ $ \begin{aligned }
\label{FijCG }
\bar F_{ij}=\frac{{\partial } x_i}{{\partial } q_j } = \delta_{ij}+\bar \varPsi_{i, j}\, , \quad J_{\bar F}=\det\left[\delta_{ij}+\bar{\varPsi}_{i, j}\right ] \,,\end{aligned}$$ we obtain the like relation between $ \bar \delta $ and $ \bar{{\bm{\varPsi } } } $ as for dust $ $ \begin{aligned }
1+\bar \delta = J^{-1}_{\bar F } \,,\label{JFtoDeltaCG}\end{aligned}$$ since $ \bar \delta$ and $ \bar{{\bm{v } } } $ fulfill the continuity equation. consequently we take in Fourier space $ $ \begin{aligned }
\bar{\delta } ({ \bm{k } })
& = \int { \hspace{-0.8mm}\mbox{$\text{d}^{\hspace{-0.0mm}3}$}\hspace{-0.2mm}q\hspace{0.8mm}\ } e^{-i{\bm{k } } \cdot{\bm{q } } } \left(e^{-i{\bm{k } } \cdot\bar{{\bm{\varPsi } } } ({ \bm{q } }) } -1\right) \,.\label{deltathetaFromPsiCG}\end{aligned}$$ Next, we extend the displacement field $ { \bm{\bar \varPsi } } (\tau,{\bm{k } }) $ perturbatively $ $ \begin{aligned }
\label{pertExpPsiCG }
\bar{{\bm{\varPsi } } } (\tau,{\bm{q } }) & = \sum_{n=1}^\infty a^n(\tau)\bar{{\bm{\varPsi } } } ^{(n)}({\bm{q } }) \,,\end{aligned}$$ and express the unlike orders $ { \bm{\bar \varPsi } } ^{(n)}$ with the help of perturbative kernels $ { \bm{\bar L } } ^{(n)}$ defined as $ $ \begin{aligned }
\label{defLCG }
\notag \bar{{\bm{\varPsi } } } ^{(n)}({\bm{k } }) & = i | PT ho LPT, proceeding along uhe lines of [@RB12], to deterkine tge pertufbative kernels in Lagrangiai spqce tyat correspond to the zoarse-grapned Euleeian jernels decjved benjre. Smen wz vefine the Jaconian of the transformatiot xs $$\begin{aligned}
\label{FijCG}
\bar F_{ij}=\frac{{\[artial }x_l}{{\partial }q_j} = \dglta_{ik}+\far \barPsi_{i,j}\,, \quad J_{\bar F}=\det\left[\delta_{in}+\bar{\varKsi}_{i,j}\right] \,,\end{aligmed}$$ we obtain the same relwtioj between $\bar \deltw $ and $\bar{{\bn{\var[wi} }}$ as for djst $$\begin{aligned}
1+\bar \demta=J^{-1}_{\bar F} \,,\label{JFtoDeltaCG}\end{alkgned}$$ since $\bar \dwltw$ and $\bar{{\bm{v} }}$ fulsill the conbpnuity aquatiom. Therefore we hate ib Fourier space $$\begin{eligned}
\bar{\delta} ({\bm{k} })
&= \int {\hspdcz{-0.8mm}\mbox{$\text{d}^{\hspace{-0.0mm}3}$}\hwpqce{-0.2mm}x\hspdce{0.8mo}\ } e^{-k{\bm{i} }\rdof{\bm{q} }} \leht(e^{-i{\bm{k} }\cdof\bar{{\bm{\varPsu} }}({\bm{q} })}-1\right) \,.\label{dektwnnetaFromPsiCF}\end{alygged}$$ Next, we expand the displacement fiend ${\gm{\bar \varPsi} }(\tau,{\bm{k} })$ peeturbatively $$\begin{alihned}
\label{[ertExpPsiCG}
\bar{{\bm{\varPsi} }}(\tau,{\bm{q} }) &= \sum_{n=1}^\infty a^n(\tau)\bdr{{\bm{\verOsi} }}^{(n)}({\nm{q} }) \,,\wnf{aligned}$$ and express the different orders ${\bm{\bwd \fagPsi} }^{(n)}$ with the hclp of perturbativr netgels ${\bm{\bar L} }^{(v)}$ defiusd as $$\begin{aligned}
\lahel{defLSG}
\notqg \bar{{\bm{\vwrPso} }}^{(n)}({\bm{k} })&= i | PT to LPT, proceeding along the lines to the perturbative in Lagrangian space Eulerian derived before. When define the Jacobian the transformation as $$\begin{aligned} \label{FijCG} \bar }x_i}{{\partial }q_j} = \delta_{ij}+\bar \varPsi_{i,j}\,, \quad J_{\bar F}=\det\left[\delta_{ij}+\bar{\varPsi}_{i,j}\right] \,,\end{aligned}$$ we obtain the same between $\bar \delta $ and $\bar{{\bm{\varPsi} }}$ as for dust $$\begin{aligned} 1+\bar \delta=J^{-1}_{\bar \,,\label{JFtoDeltaCG}\end{aligned}$$ $\bar and }}$ fulfill the continuity equation. Therefore we have in Fourier space $$\begin{aligned} \bar{\delta} ({\bm{k} }) &= {\hspace{-0.8mm}\mbox{$\text{d}^{\hspace{-0.0mm}3}$}\hspace{-0.2mm}q\hspace{0.8mm}\ } e^{-i{\bm{k} }\cdot{\bm{q} }} \left(e^{-i{\bm{k} }\cdot\bar{{\bm{\varPsi} }}({\bm{q} \,.\label{deltathetaFromPsiCG}\end{aligned}$$ Next, we expand displacement field ${\bm{\bar \varPsi} }(\tau,{\bm{k} perturbatively \label{pertExpPsiCG} \bar{{\bm{\varPsi} }) \sum_{n=1}^\infty }}^{(n)}({\bm{q} }) \,,\end{aligned}$$ express the different orders ${\bm{\bar \varPsi} }^{(n)}$ with the help of perturbative kernels ${\bm{\bar L} }^{(n)}$ defined $$\begin{aligned} \label{defLCG} }}^{(n)}({\bm{k} })&= | PT to LPT, proceeding along the Lines of [@RB12], tO deteRmiNe tHe PertUrbaTive kernels in LAGranGian space that corresponD to thE cOArse-GRaIned EUlerian KErNELs dErIvEd bEfORe. when wE deFine the jacobian of The TrAnsformation AS $$\bEgin{aligneD}
\laBel{FijCG}
\bar F_{Ij}=\fRac{{\parTiAl }x_I}{{\PartiAl }q_J} = \deltA_{ij}+\bar \VArPsi_{i,J}\,, \quad J_{\bar f}=\dET\left[\dELta_{ij}+\baR{\VArpsi}_{i,J}\right] \,,\end{aligned}$$ wE ObTAin the same relaTion beTwEEn $\BAR \deLta $ And $\bar{{\bm{\vaRPSi} }}$ as fOR dust $$\beGIn{ALIGneD}
1+\Bar \delta=J^{-1}_{\bar F} \,,\Label{JFtoDeLTaCg}\end{alIgNed}$$ SInce $\baR \deltA$ aND $\baR{{\bm{v} }}$ fulfill The cOntinuity EquatiON. TherefORe we havE in FouRieR spAce $$\bEGiN{aLigNeD}
\Bar{\DElTa} ({\bM{K} })
&= \inT {\hspace{-0.8mM}\mBoX{$\text{D}^{\hspACE{-0.0MM}3}$}\hspAce{-0.2Mm}q\hSpace{0.8Mm}\ } e^{-i{\bm{k} }\cdot{\bm{Q} }} \leFt(e^{-i{\BM{k} }\cDot\baR{{\bm{\vaRPsi} }}({\Bm{Q} })}-1\righT) \,.\label{DeltaThEtaFromPsiCG}\end{AligNed}$$ Next, we ExpAnD thE dIsplaCEment fIelD ${\bm{\Bar \varPSi} }(\tau,{\bm{K} })$ PerTuRBATiVely $$\begin{aligned}
\laBeL{PErTExpPsiCg}
\bar{{\bm{\VArpsI} }}(\Tau,{\bm{q} }) &= \suM_{n=1}^\InfTy a^n(\TAU)\bar{{\bM{\varpSi} }}^{(N)}({\bm{q} }) \,,\end{aLigned}$$ ANd ExPress thE dIffereNt OrdErs ${\Bm{\bar \VArPsI} }^{(n)}$ with The help oF pertURbative kernels ${\BM{\bar L} }^{(n)}$ defined AS $$\bEGIn{ALignEd}
\lAbel{defLCG}
\nOtag \BAr{{\bm{\VarPSI} }}^{(n)}({\Bm{k} })&= I | PT to LPT, proceeding alon g the line s of[@R B12 ], todete rmine the pert u rbat ive kernels in Lagrang ian s pa c e th a tcorre spond t o t h e co ar se -gr ai n ed Eule ria n kerne ls derived be fo re. When wed ef ine the Ja cob ian of the t ran sforma ti ona s $$\ beg in{al igned} \labe l{FijCG}\b a r F_{i j }=\frac { { \p arti al }x_i}{{\partia l } q _j} = \delta_{ ij}+\b ar \v a r Psi _{i ,j}\,, \qu ad J_{ \ bar F}= \ de t \ l eft [ \delta_{ij}+\ bar{\varPsi } _{i ,j}\ri gh t]\ ,,\end {alig ne d }$$ we obtainthesame rela tion b e tween $ \ bar \de lta $and $\ bar{ { \b m{ \va rP s i}} }$ as for dust $$ \b eg in{al igne d } 1 +\ba r \ delt a=J^{ -1}_{\bar F}\,, \lab e l{J FtoDe ltaCG }\en d{ align ed}$$since $ \bar \delta$ an d $\ bar{{\bm{ v}}} $ f ul fillt he con tin uit y equat ion. Th e ref or e w ehave in Fourier sp ac e $$ \begin{a ligned }
\ ba r {\delta} ( {\b m{k} } )
&= \in t { \hspace{ -0.8mm } \m bo x{$\tex t{ d}^{\h sp ace {-0 .0mm} 3 }$}\ hspace {-0.2mm} q\hsp a ce{0.8mm}\ } e ^ {-i{\bm{k} }\ c do t { \b m {q}}} \left(e^{- i{\b m {k}}\cd o t\ bar { {\bm{ \varP si } } } ({\bm{q} })}-1\righ t) \,.\l abel{ deltathetaFro mPsiCG}\en d { a ligned}$ $ Ne x t, we expand thedispl acement fi e ld ${\bm {\bar \varPsi } }(\tau, { \ bm{k} }) $ p ert urb ati v e ly $$\begin{ali g n ed}\l abel{pe rtE xpPsiCG }
\ bar {{\ bm{ \v arPsi} }} (\tau,{\ bm {q }}) &= \sum _ {n=1}^\i nf tya^ n(\ tau)\ b ar{{\b m{\va rPsi }}} ^ {(n )}({\bm { q} } ) \, ,\ en d{al ign ed }$$ a nd e x pre ss thedifferent or d ers${ \b m{\bar\varPsi} }^{( n) }$ with th ehel p of p e r turbativ e kernels ${\bm{\bar L} }^{(n)} $ d efine d as $$\begin {al igned}
\l a bel{de fLCG}\nota g\ba r { {\bm{ \ v ar Psi }}}^{(n)}({ \ b m{k } })& =i | PT to_LPT, proceeding_along the lines of_[@RB12], to_determine_the perturbative_kernels_in Lagrangian space_that correspond to_the coarse-grained Eulerian kernels_derived before. When_we_define the Jacobian of the transformation as $$\begin{aligned}
\label{FijCG}
\bar F_{ij}=\frac{{\partial }x_i}{{\partial }q_j} = \delta_{ij}+\bar_\varPsi_{i,j}\,,_\quad _J_{\bar_F}=\det\left[\delta_{ij}+\bar{\varPsi}_{i,j}\right]_\,,\end{aligned}$$ we obtain the same_relation between $\bar \delta $_and $\bar{{\bm{\varPsi}_}}$ as for dust $$\begin{aligned}
1+\bar \delta=J^{-1}_{\bar F} \,,\label{JFtoDeltaCG}\end{aligned}$$_since_$\bar \delta$ and_$\bar{{\bm{v} }}$ fulfill the continuity equation. Therefore we have_in Fourier space $$\begin{aligned}
\bar{\delta} ({\bm{k} })_
&= \int {\hspace{-0.8mm}\mbox{$\text{d}^{\hspace{-0.0mm}3}$}\hspace{-0.2mm}q\hspace{0.8mm}\_}_e^{-i{\bm{k}_}\cdot{\bm{q} }} \left(e^{-i{\bm{k}_}\cdot\bar{{\bm{\varPsi} }}({\bm{q} })}-1\right) \,.\label{deltathetaFromPsiCG}\end{aligned}$$ Next, we_expand the displacement field ${\bm{\bar \varPsi}_}(\tau,{\bm{k} })$ perturbatively $$\begin{aligned}
\label{pertExpPsiCG}
\bar{{\bm{\varPsi} }}(\tau,{\bm{q} }) &=_\sum_{n=1}^\infty a^n(\tau)\bar{{\bm{\varPsi} }}^{(n)}({\bm{q} }) \,,\end{aligned}$$ and_express the different orders ${\bm{\bar_\varPsi} }^{(n)}$_with the help of perturbative_kernels ${\bm{\bar L}_}^{(n)}$ defined_as $$\begin{aligned}
\label{defLCG}
\notag \bar{{\bm{\varPsi}_}}^{(n)}({\bm{k} })&= i |
:2017aqq]:
1. $q\lesssim$ 40 MeV: The response function $S_{\rm pasta}(q)$ $\ll 1$, as the scattering is incoherent due to density fluctuations, which results in destructive interference of scattering amplitudes. In this region it is given by $S_{\rm pasta} \rightarrow T(\partial\rho/\partial P)_{T}$, where $T$ is the pasta temperature. For example, at $T = 1000$ K, $S_{\rm pasta}(q)$ is $\mathcal{O}(10^{-6})$.
2. $q$ $\simeq$ 60 MeV: The response function $S_{\rm pasta}(q)$ peaks, due to coherent enhancement from scattering on multiple nucleons.
3. $q\gtrsim$ 200 MeV: The response function $S_{\rm pasta}(q) \rightarrow 1$, corresponding to quasi-elastic scattering on weakly bound individual neutrons.
To calculate the dark matter-pasta scattering sensitivities, we modify Eq. for the pasta layers as $$\tau_{\rm DM} =\frac{\sigma_{\rm n\chi}}{g_{s}}\int_{\rm pasta} \langle S_{\rm pasta}(q) \rangle_{q} \ \frac{n_{\rm n}(\rho)}{\rho}\frac{dP}{d\rho} d\rho~,
\label{eq:Nscatterspasta}$$ where $n_{\rm n}$ is the nucleon number density. The $q$-averaged structure function is given by $$\langle S_{\rm pasta}(q) \rangle_{q} = \frac{1}{q_{\rm max}}\int_{0}^{q_{\rm max}} S_{\rm pasta}(q) \ dq~,$$ with the maximum momentum transfer $$q_{\rm max}=\sqrt{\frac{4\mu_{\rm n\chi}{m_{\rm DM}}^{2}m_{\rm n}\gamma^{2} v_{\rm esc}^{2}}{{m_{\rm DM}}^{2}+m_{\rm n}^{2}+2{m_{\rm DM}}m_{\rm n}\gamma}}~.$$ We then integrate Eq. over the pasta layer densities, which extend from 0.1 $\rho_0$ to 0.3 $\rho_0$. This range is a conservative choice, as various pasta structures are expected to appear near the saturation density $\rho_0$, | : 2017aqq ]:
1. $ q\lesssim$ 40 MeV: The response function $ S_{\rm pasta}(q)$ $ \ll 1 $, as the scattering is incoherent due to density fluctuation, which result in destructive interference of scattering amplitudes. In this area it is given by $ S_{\rm pasta } \rightarrow T(\partial\rho/\partial P)_{T}$, where $ T$ is the pasta temperature. For example, at $ T = 1000 $ K, $ S_{\rm pasta}(q)$ is $ \mathcal{O}(10^{-6})$.
2. $ q$ $ \simeq$ 60 MeV: The reaction function $ S_{\rm pasta}(q)$ peaks, due to coherent enhancement from disperse on multiple nucleons.
3. $ q\gtrsim$ 200 MeV: The response officiate $ S_{\rm pasta}(q) \rightarrow 1 $, corresponding to quasi - elastic scattering on decrepit bound individual neutrons.
To calculate the dark matter - pasta scattering sensitivities, we change Eq. for the pasta layers as $ $ \tau_{\rm DM } = \frac{\sigma_{\rm n\chi}}{g_{s}}\int_{\rm pasta } \langle S_{\rm pasta}(q) \rangle_{q } \ \frac{n_{\rm n}(\rho)}{\rho}\frac{dP}{d\rho } d\rho~,
\label{eq: Nscatterspasta}$$ where $ n_{\rm n}$ is the nucleon number density. The $ q$-averaged social organization function is given by $ $ \langle S_{\rm pasta}(q) \rangle_{q } = \frac{1}{q_{\rm max}}\int_{0}^{q_{\rm max } } S_{\rm pasta}(q) \ dq~,$$ with the maximum momentum transfer $ $ q_{\rm max}=\sqrt{\frac{4\mu_{\rm n\chi}{m_{\rm DM}}^{2}m_{\rm n}\gamma^{2 } v_{\rm esc}^{2}}{{m_{\rm DM}}^{2}+m_{\rm n}^{2}+2{m_{\rm DM}}m_{\rm n}\gamma}}~.$$ We then integrate Eq. over the pasta level densities, which exsert from 0.1 $ \rho_0 $ to 0.3 $ \rho_0$. This range is a conservative choice, as diverse pasta structures are expected to appear near the saturation density $ \rho_0 $, | :2017aqq]:
1. $q\lesssim$ 40 MeV: The responre function $S_{\rm pasta}(q)$ $\ll 1$, aa the scxttering is incoherent due tl eensiuj fluctuations, which fesults ij destruxtivt interference of scatterlug amlpitubew. In this regipn it is ghven by $S_{\rm pavtx} \xightarrow T(\partial\rho/\partial P)_{T}$, whewe $T$ is tje pasta tempetaturt. Fjr esample, at $T = 1000$ K, $S_{\rm pasta}(q)$ is $\mathczl{O}(10^{-6})$.
2. $q$ $\spmeq$ 60 MeV: The respomse function $S_{\rm pasta}(q)$ pewks, fue to coherent enjancement ftkm fxattering on multiple nucleons.
3. $q\gjrsim$ 200 MeV: The response function $S_{\fm paxta}(q) \rightqreow 1$, correspondmng to quasi-elastig scattaring om weakly bound invivieual neutrons.
To calcunate the dark mattgr-pasta scdtcering sensitivities, qe modixy Ex. fof thd pzsva mayers as $$\tau_{\rm DM} =\frac{\sigma_{\rn n\chi}}{g_{s}}\int_{\rm pasta} \lwbgle S_{\rm pastz}(q) \randlq_{q} \ \frac{n_{\rm n}(\rho)}{\rho}\frac{dP}{d\rho} d\rho~,
\laben{eq:Hscatterspasta}$$ where $n_{\rn n}$ is the nucleon nulber densyty. The $q$-averaged structure function is given by $$\nanglx R_{\rm pwrra}(e) \rangle_{q} = \frac{1}{q_{\rm max}}\int_{0}^{q_{\rm max}} S_{\rm pasta}(q) \ qs~,$$ einh the maximum moientum tranxffr $$z_{\rm max}=\sqrt{\frxc{4\mu_{\rm n\cgi}{m_{\rm DM}}^{2}m_{\rm n}\gamma^{2} v_{\rm ess}^{2}}{{m_{\rm EM}}^{2}+m_{\rm n}^{2}+2{m_{\ri DM}}k_{\rm n}\gamma}}~.$$ We then integratw Eq. over thv pawta layer densitied, which extznd frpm 0.1 $\rhp_0$ to 0.3 $\rho_0$. This range is c conssrvative chlice, as vzfious pasta struztuges dre expected to appear neaw the satnratipn denskty $\tho_0$, | :2017aqq]: 1. $q\lesssim$ 40 MeV: The response pasta}(q)$ 1$, as scattering is incoherent results destructive interference of amplitudes. In this it is given by $S_{\rm pasta} T(\partial\rho/\partial P)_{T}$, where $T$ is the pasta temperature. For example, at $T = K, $S_{\rm pasta}(q)$ is $\mathcal{O}(10^{-6})$. 2. $q$ $\simeq$ 60 MeV: The response function pasta}(q)$ due coherent from scattering on multiple nucleons. 3. $q\gtrsim$ 200 MeV: The response function $S_{\rm pasta}(q) \rightarrow 1$, to quasi-elastic scattering on weakly bound individual neutrons. calculate the dark matter-pasta sensitivities, we modify Eq. for pasta as $$\tau_{\rm =\frac{\sigma_{\rm pasta} S_{\rm pasta}(q) \rangle_{q} \frac{n_{\rm n}(\rho)}{\rho}\frac{dP}{d\rho} d\rho~, \label{eq:Nscatterspasta}$$ where $n_{\rm n}$ is the nucleon number density. The $q$-averaged structure function is by $$\langle \rangle_{q} = max}}\int_{0}^{q_{\rm S_{\rm \ dq~,$$ with momentum transfer $$q_{\rm max}=\sqrt{\frac{4\mu_{\rm n\chi}{m_{\rm DM}}^{2}m_{\rm DM}}^{2}+m_{\rm n}^{2}+2{m_{\rm DM}}m_{\rm n}\gamma}}~.$$ We then integrate Eq. the pasta densities, which extend from 0.1 $\rho_0$ 0.3 $\rho_0$. This range is a conservative choice, various pasta structures are expected to appear near the saturation density $\rho_0$, | :2017aqq]:
1. $q\lesssim$ 40 MeV: The response Function $S_{\rM pastA}(q)$ $\lL 1$, as ThE scaTterIng is incoherenT Due tO density fluctuations, whIch reSuLTs in DEsTructIve inteRFeRENce Of ScAttErINg AmpliTudEs. In thiS region it iS giVeN by $S_{\rm pasta} \rIGhTarrow T(\parTiaL\rho/\partial P)_{t}$, whEre $T$ is ThE paSTa temPerAture. for exaMPle, at $T = 1000$ k, $S_{\rm pasta}(Q)$ iS $\MathcaL{o}(10^{-6})$.
2. $q$ $\simeq$ 60 mEv: THe reSponse function $S_{\rm PAsTA}(q)$ peaks, due to coHerent EnHAnCEMenT frOm scatteriNg On mulTIple nucLEoNS.
3. $Q\GtrSIm$ 200 MeV: The respoNse function $s_{\Rm pAsta}(q) \rIgHtaRRow 1$, corRespoNdINg tO quasi-elastIc scAttering oN weaklY Bound inDIvidual NeutroNs.
TO caLculATe ThE daRk MAttER-pAstA ScaTtering sEnSiTivitIes, wE MODIfy EQ. foR the Pasta Layers as $$\tau_{\rm dM} =\fRac{\sIGma_{\Rm n\chI}}{g_{s}}\inT_{\rm pAsTa} \lanGle S_{\rm Pasta}(Q) \rAngle_{q} \ \frac{n_{\rm n}(\rHo)}{\rhO}\frac{dP}{d\rHo} d\RhO~,
\laBeL{eq:NsCAttersPasTa}$$ wHere $n_{\rm N}$ is the nUCleOn NUMBeR density. The $q$-averagEd STRuCture funCtion iS GiVeN By $$\langle s_{\rM paSta}(q) \RANgle_{q} = \Frac{1}{Q_{\Rm Max}}\int_{0}^{q_{\rM max}} S_{\rM PaStA}(q) \ dq~,$$ witH tHe maxiMuM moMenTum trANsfeR $$q_{\rm maX}=\sqrt{\fraC{4\mu_{\rm N\Chi}{m_{\rm DM}}^{2}m_{\rm n}\gaMMa^{2} v_{\rm esc}^{2}}{{m_{\rm DM}}^{2}+M_{\Rm N}^{2}+2{M_{\Rm dm}}m_{\rm N}\gaMma}}~.$$ We then inTegrATe Eq. Over THe PasTA layeR densItIEs, WHich extend from 0.1 $\rho_0$ to 0.3 $\RhO_0$. This rAnge iS a conservativE choice, as vARIOus pasta StruCTuREs are expected tO appeAr near the sATuration DensiTy $\rho_0$, | :2017aqq]:
1. $q\lesssim $ 40 MeV:The r esp ons efunc tion $S_{\rm pasta } (q)$ $\ll 1$, as the scatt ering i s inc o he rentdue tod en s i tyfl uc tua ti o ns , whi chresults in destru cti ve interferenc e o f scatteri ngamplitudes.Inthis r eg ion it is gi ven b y $S_{ \ rm pas ta} \righ ta r row T( \ partial \ r ho /\pa rtial P)_{T}$, wh e re $T$ is the pas ta tem pe r at u r e.For example,at $T = 1000$ K , $ S _ { \rm pasta}(q)$ is $\mathcal{ O }(1 0^{-6} )$ .
2 . $q$ $\si me q $ 6 0 MeV: Theresp onse func tion $ S _{\rm p a sta}(q) $ peak s,due toc oh er ent e n han c em ent fro m scatte ri ng on m ulti p l e nucl eon s.
3. $ q\gtrsim$ 200 Me V: T h e r espon se fu ncti on $S_{ \rm pa sta}( q) \rightarrow 1$ , co rrespondi ngto qu as i-ela s tic sc att eri ng on w eakly b o und i n d i vi dual neutrons.
To c a l cu late the darkm at te r -pasta s ca tte ring s ensit ivit i es , we mod ify Eq . fo r the p as ta lay er s a s $ $\tau _ {\rm DM} =\frac{\ sigma _ {\rm n\chi}}{g _ {s}}\int_{\rm pa s t a} \lan gle S_{\rm pas ta}( q ) \r angl e _{ q}\ \fra c{n_{ \r m n } (\rho)}{\rho}\frac{ dP }{d\rh o} d\ rho~,
\label {eq:Nscatt e r s pasta}$$ whe r e$ n_{\rm n}$ isthe n ucleon num b er densi ty. T he $q$-a veraged s t r ucture f unc tio n i s g i v en by $$\langle S _{\r mpasta}( q)\rangle _{q } = \f rac {1 }{q_{\rmmax}}\in t_ {0 }^ {q _{\ rm ma x }} S_{\r mpas ta }(q ) \ d q ~,$$ w ith t he m ax im u m m omentum tr a n sfer $ $q _{\r m m ax }=\sq rt{\ f rac {4\mu_{ \rm n\chi }{m _ {\rm D M} }^{2}m_ {\rm n}\gamma ^{ 2} v_{\rmes c}^ {2}}{{ m _ {\rm DM} }^{2}+m_{\rm n}^{2}+2{m _ {\rm DM }}m _{\rm n}\ gamma}}~. $$We the n i n tegrat e Eq. over t hep a sta l a y er de ns ities, whi c h ex tendfr om 0 .1 $\rh o_0$ to 0.3 $\rho_ 0 $.This range is acons e r va tiv e c h oic e, asv a rious pasta str uctures ar ee xp ected to a p pea rnear th e satur ation density $\rho_0$ , | :2017aqq]:
1. _$q\lesssim$ 40 MeV: The_response function $S_{\rm pasta}(q)$_$\ll 1$,_as_the scattering_is_incoherent due to_density fluctuations, which_results in destructive interference_of scattering amplitudes._In_this region it is given by $S_{\rm pasta} \rightarrow T(\partial\rho/\partial P)_{T}$, where $T$ is_the_pasta temperature._For_example,_at $T = 1000$ K, $S_{\rm_pasta}(q)$ is $\mathcal{O}(10^{-6})$.
2. $q$ $\simeq$ 60 MeV:_The response_function $S_{\rm pasta}(q)$ peaks, due to coherent enhancement_from_scattering on multiple_nucleons.
3. $q\gtrsim$ 200 MeV: The response function $S_{\rm pasta}(q) \rightarrow_1$, corresponding to quasi-elastic scattering on_weakly bound individual_neutrons.
To_calculate_the dark matter-pasta scattering_sensitivities, we modify Eq. for the_pasta layers as $$\tau_{\rm DM} _=\frac{\sigma_{\rm n\chi}}{g_{s}}\int_{\rm pasta} \langle S_{\rm pasta}(q) \rangle_{q}_\ \frac{n_{\rm n}(\rho)}{\rho}\frac{dP}{d\rho} d\rho~,
\label{eq:Nscatterspasta}$$ where_$n_{\rm n}$ is the nucleon_number density._The $q$-averaged structure function is_given by $$\langle_S_{\rm pasta}(q)_\rangle_{q} = \frac{1}{q_{\rm_max}}\int_{0}^{q_{\rm max}} S_{\rm pasta}(q) \ dq~,$$_with the maximum_momentum transfer $$q_{\rm max}=\sqrt{\frac{4\mu_{\rm n\chi}{m_{\rm DM}}^{2}m_{\rm_n}\gamma^{2}_v_{\rm esc}^{2}}{{m_{\rm DM}}^{2}+m_{\rm_n}^{2}+2{m_{\rm_DM}}m_{\rm_n}\gamma}}~.$$ We_then integrate Eq. _over_the pasta_layer_densities, which extend from 0.1 $\rho_0$ to_0.3 $\rho_0$._This range is a conservative choice, as_various pasta structures are_expected_to appear near the_saturation density $\rho_0$, |
$\beta_d$ $\rho_d(2)$ $\rho_d(3)$ $\rho_d(5)$ $\rho_d(7)$
\[-2.3ex\]
\[-2.3ex\] ${\mathsf{M}_{14}}$ $4$ $-$ $-$ $19$ $2$ $-$
$5$ $112$ $-$ $-$ $-$ $-$
\[-2.3ex\]
\[-2.3ex\] ${\mathsf{M}_{16}}$ $5$ $6$ $-$ $43$ $8$ $-$
$6$ $76$ $-$ $-$ $-$ $-$
\[-2.3ex\]
\[-2.3ex\] ${\mathsf{M}_{18}}$ $6$ $20$ $-$ $18$ $2$ $-$
$7$ $20$ $-$ $-$ $-$ $-$
\[-2.3ex\]
\[-2.3ex\] $6$ $28$ $-$ $44$ $8$ $-$
$7$ $28$ $-$ $- $ $-$ $-$
\[-2.3ex\]
\[-2.3ex\] $6$ $40$ $-$ $90$ $20$ $-$
$7$ $40$ $-$ $-$ $-$ $-$
\[-2.3ex\]
\[-2.3ex\] ${\mathsf{M}_{20}}$ $7$ $76$ $-$ $43$ $8$ $-$
$8$ $6 $ $-$ $-$ $-$ $-$
\[-2.3ex\]
\[-2. | $ \beta_d$ $ \rho_d(2)$ $ \rho_d(3)$ $ \rho_d(5)$ $ \rho_d(7)$
\[-2.3ex\ ]
\[-2.3ex\ ] $ { \mathsf{M}_{14}}$ $ 4 $ $ -$ $ -$ $ 19 $ $ 2 $ $ -$
$ 5 $ $ 112 $ $ -$ $ -$ $ -$ $ -$
\[-2.3ex\ ]
\[-2.3ex\ ] $ { \mathsf{M}_{16}}$ $ 5 $ $ 6 $ $ -$ $ 43 $ $ 8 $ $ -$
$ 6 $ $ 76 $ $ -$ $ -$ $ -$ $ -$
\[-2.3ex\ ]
\[-2.3ex\ ] $ { \mathsf{M}_{18}}$ $ 6 $ $ 20 $ $ -$ $ 18 $ $ 2 $ $ -$
$ 7 $ $ 20 $ $ -$ $ -$ $ -$ $ -$
\[-2.3ex\ ]
\[-2.3ex\ ] $ 6 $ $ 28 $ $ -$ $ 44 $ $ 8 $ $ -$
$ 7 $ $ 28 $ $ -$ $ - $ $ -$ $ -$
\[-2.3ex\ ]
\[-2.3ex\ ] $ 6 $ $ 40 $ $ -$ $ 90 $ $ 20 $ $ -$
$ 7 $ $ 40 $ $ -$ $ -$ $ -$ $ -$
\[-2.3ex\ ]
\[-2.3ex\ ] $ { \mathsf{M}_{20}}$ $ 7 $ $ 76 $ $ -$ $ 43 $ $ 8 $ $ -$
$ 8 $ $ 6 $ $ -$ $ -$ $ -$ $ -$
\[-2.3ex\ ]
\[-2. | $\heta_d$ $\rho_d(2)$ $\rho_d(3)$ $\rhu_d(5)$ $\rho_d(7)$
\[-2.3xx\]
\[-2.3tx\] ${\iathaf{M}_{14}}$ $4$ $-$ $-$ $19$ $2$ $-$
$5$ $112$ $-$ $-$ $-$ $-$
\[-2.3ex\]
\[-2.3ex\] ${\mathsf{M}_{16}}$ $5$ $6$ $-$ $43$ $8$ $-$
$6$ $76$ $-$ $-$ $-$ $-$
\[-2.3ex\]
\[-2.3ex\] ${\oatmsf{K}_{18}}$ $6$ $20$ $-$ $18$ $2$ $-$
$7$ $20$ $-$ $-$ $-$ $-$
\[-2.3ex\]
\[-2.3ex\] $6$ $28$ $-$ $44$ $8$ $-$
$7$ $28$ $-$ $- $ $-$ $-$
\[-2.3ex\]
\[-2.3ex\] $6$ $40$ $-$ $90$ $20$ $-$
$7$ $40$ $-$ $-$ $-$ $-$
\[-2.3ex\]
\[-2.3ex\] ${\mauhsf{M}_{20}}$ $7$ $76$ $-$ $43$ $8$ $-$
$8$ $6 $ $-$ $-$ $-$ $-$
\[-2.3ex\]
\[-2. | $\beta_d$ $\rho_d(2)$ $\rho_d(3)$ $\rho_d(5)$ $\rho_d(7)$ \[-2.3ex\] \[-2.3ex\] $-$ $19$ $2$ $5$ $112$ $-$ ${\mathsf{M}_{16}}$ $6$ $-$ $43$ $-$ $6$ $76$ $-$ $-$ $-$ \[-2.3ex\] \[-2.3ex\] ${\mathsf{M}_{18}}$ $20$ $-$ $18$ $2$ $-$ $7$ $20$ $-$ $-$ $-$ $-$ \[-2.3ex\] \[-2.3ex\] $28$ $-$ $44$ $8$ $-$ $7$ $28$ $-$ $- $ $-$ $-$ \[-2.3ex\] $6$ $-$ $20$ $7$ $40$ $-$ $-$ $-$ $-$ \[-2.3ex\] \[-2.3ex\] ${\mathsf{M}_{20}}$ $7$ $76$ $-$ $43$ $8$ $-$ $8$ $ $-$ $-$ $-$ $-$ \[-2.3ex\] \[-2. | $\beta_d$ $\rho_d(2)$ $\rho_d(3)$ $\rho_d(5)$ $\rho_d(7)$
\[-2.3ex\]
\[-2.3ex\] ${\Mathsf{M}_{14}}$ $4$ $-$ $-$ $19$ $2$ $-$
$5$ $112$ $-$ $-$ $-$ $-$
\[-2.3ex\]
\[-2.3eX\] ${\mathSf{M}_{16}}$ $5$ $6$ $-$ $43$ $8$ $-$
$6$ $76$ $-$ $-$ $-$ $-$
\[-2.3Ex\]
\[-2.3eX\] ${\mAthsF{M}_{18}}$ $6$ $20$ $-$ $18$ $2$ $-$
$7$ $20$ $-$ $-$ $-$ $-$
\[-2.3ex\]
\[-2.3Ex\] $6$ $28$ $-$ $44$ $8$ $-$
$7$ $28$ $-$ $- $ $-$ $-$
\[-2.3ex\]
\[-2.3ex\] $6$ $40$ $-$ $90$ $20$ $-$
$7$ $40$ $-$ $-$ $-$ $-$
\[-2.3ex\]
\[-2.3ex\] ${\mathSF{M}_{20}}$ $7$ $76$ $-$ $43$ $8$ $-$
$8$ $6 $ $-$ $-$ $-$ $-$
\[-2.3ex\]
\[-2. | $\beta_d$ $\rho_d(2)$ $\rho_d (3)$ $ \rh o_ d(5) $ $\rho_d(7)$
\[-2.3ex\ ]
\[-2. 3ex\]${ \ ma t h sf{ M}_ {14}}$ $4$ $- $ $ -$ $1 9 $ $ 2 $ $- $
$5 $ $ 112$ $-$ $- $ $-$ $- $
\[ -2 . 3 ex \] \[- 2.3ex\]${\ma t hsf{M}_{16}}$ $5$ $ 6$ $ - $ $43$ $8$ $ -$
$6 $ $76$ $-$ $-$ $-$ $-$
\[ -2.3ex\ ]
\ [ - 2.3ex\]${\mathsf{M}_{18}}$ $6 $ $20 $ $-$ $1 8 $ $ 2$ $ -$ $7$ $2 0$ $-$ $-$ $-$ $-$
\[-2 .3ex\ ]
\ [ -2.3e x\] $6 $ $28$ $-$ $44$ $8 $ $ -$
$7$ $28$ $ -$ $- $ $- $ $-$
\ [-2.3ex\]
\ [-2 . 3 ex\] $ 6 $ $4 0$ $-$ $90$ $20$ $ - $
$7$ $40 $ $-$ $ -$ $- $ $- $
\ [-2 . 3ex\]
\[-2 .3ex\ ]${\mathsf { M}_{20} }$ $7$ $7 6 $ $-$ $43$ $8$ $-$ $8$ $ 6 $ $-$ $ -$ $ - $ $- $
\[- 2. 3ex \]
\[-2. | _ $\beta_d$_ $\rho_d(2)$ _ $\rho_d(3)$__ $\rho_d(5)$__ $\rho_d(7)$
_ _ _ __ __ ___ _ _ _ __ _ _ _ ___ _ _ _ _ _ _ _ _ _ _
_ _ _ \[-2.3ex\] __ ___ _ __ __ __ _ __ _ _ _ _ _ _ _ _ __ __ _ _ _
_ \[-2.3ex\] ${\mathsf{M}_{14}}$ __ _ _ ___ _ $4$ _ _ $-$ __ $-$ __ _$19$__ _ $2$ _ _$-$
_ ___ _ _ _ _ _ _ _ $5$ _ $112$ __ $-$ ____ _ _$-$ _ _ _ _ $-$ __ $-$
_ _ __ _ _ _ __ _ _ _ _ _ _ _ _ __ ____ _ _ _ __ _ _ _ _
\[-2.3ex\] _ _ _ __ _ _ __ ___ __ _____ _ __ _ __ _ _ __
___\[-2.3ex\]_${\mathsf{M}_{16}}$ _ __ _ _$5$_ _ $6$ _ $-$ _ _ $43$ $8$ _ _$-$
_ __ _ _ _ _ _ _ _ _ $6$ $76$ _ $-$ __ $-$ _ _$-$ _ __ $-$
_ _ _ _ _ _ _ _ __ _ _ _ ___ __ _ _ _ _ _ _
_ __ \[-2.3ex\] _ __ _ __ _ _ __ _ __ _ __ _ _ _ __ _ _ _ _ _ _ _ __ _ _
__\[-2.3ex\] ${\mathsf{M}_{18}}$ _ _ _ _ _ _ $6$ ___ $20$ _ $-$ _ _ __$18$ _ _ _$2$ __ __$-$
_ _ _ _ ___ _ __ _ _ _ $7$ ___$20$ __ _$-$ __ $-$ _ $-$ $-$
__ ___ _ _ _ _ _ _ _ _ _ _ __ _ _ _ _ __ _ _ _ _ __ _ _ _
_ _ _ \[-2.3ex\] _ ___ _ _ _ _ _ _ _ __ __ _ _
_ ___\[-2.3ex\] _ _ _ _ ____ _$6$ _ $28$_ __ $-$ $44$ __ _ $8$ _ __$-$
_ _ _ _ _ _ _ _ _ __ _ $7$ $28$ _$-$ _ $- $_ _ __ _ $-$ ___ _$-$
__ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ ___ _ _ _ __ _ _ _ _
_ _ __\[-2.3ex\]_ __ _ _ _ _ __ _ _ _ _ _ _ _ ___ _ _ ____ _ __ _
_ _ \[-2.3ex\]_ _ ___ _ _ _ $6$ $40$_ _ _ _ $-$ _ $90$ __ $20$ _ $-$
_ _ __ _ _ _ _ _ ____ _ _ ___$7$ $40$_ __ __ $-$ _ _$-$__ _ $-$ _ _ $-$
_ __ __ __ _ _ _ _ _ _ _ _ __ _ _ _ _ _ _ _ _
_ \[-2.3ex\] __ _ _ _ _ _ ___ _ _ _ _ _ __ _ ___ _ _ _ _ _ __ __
\[-2.3ex\]_${\mathsf{M}_{20}}$ __ _ _ _ $7$_____$76$ $-$_ _ _ $43$ _$8$ _ $-$
_ _ _ __ _ _ _ _ _ _ _ _ __ _ __ _ _ _ $8$_ _ $6 $ _ __ $-$ _ _ $-$ _ _$-$ _ $-$
__ __ _ _ _ __ _ _ __ _ _ _ _ _ _ _ _ _ _ _ __ _ __ _ _
_ _ _ _ \[-2.3ex\] ___ _ _ __ __ _ __ _ __ _ _ __ _ __ _ _ _ _ ___ _ _ _ __
\[-2. |
$P=2.15\:$d by @Hartman2011. Having a host star brightness of $V=11.3\:$mag and a planetary transit depth of $21\:$mmag the sensitivity of medium-sized telescopes is sufficient to achieve high timing precision, therefore it is an optimal target for the YETI telescopes. The RV signal of HAT-P-32 is dominated by high jitter of $>60\:$ms$^{-1}$. @Hartman2011 claim that ’a possible cause of the jitter is the presence of one or more additional planets’. @Knutson2013 also analysed the RV signature of HAT-P-32 and found a long term trend indicating a companion with a minimum mass of $5-500\:$M$_{jup}$ at separations of $3.5-12\:$AU. However, such a companion could not yet explain the short time-scale jitter as seen in the Hartman data.
Besides the circular orbit fit, an eccentric solution with $e=0.163$ also fits the observed data. Though @Hartman2011 mention that the probability of a real non-zero eccentricity is only $\sim3\%$, it could be mimicked or triggered by a second body in the system. Thus, HAT-P-32b is an ideal candidate for further monitoring to look for Transit Timing Variations induced by a planetary companion.
Data acquisition and reduction {#sec:DataAquisitionAndReduction}
==============================
Between 2011 October and 2013 January we performed 30 complete and 15 partial transit observations (see Tables \[tab:H32completeObservations\] and \[tab:H32partialObservations\]) from 10 different observatories: our own one in Jena, as well as from telescopes at Torun (Poland), Trebur (Germany), Gettysburg and Swarthmore (USA), Tenerife and Sierra Nevada (Spain), Antalya and Ankara (Turkey), and Rozhen (Bulgaria) mostly throughout the YETI network [@YETI]. In addition, three literature data points from @Sada, and two observations from @Gibson2013 are available. The telescopes and abbreviations used hereafter are summarized in Table \[tab:H32Telescopes\], a short description of each observing site can be found below, sorted by the number of observations.
\# Observatory Telescope (abbreviation) $\ | $ P=2.15\:$d by @Hartman2011. Having a host star brightness of $ V=11.3\:$mag and a global passage depth of $ 21\:$mmag the sensitivity of medium - sized telescopes is sufficient to achieve gamey timing precision, therefore it is an optimum target for the YETI telescopes. The RV signal of HAT - P-32 is dominated by eminent jitter of $ > 60\:$ms$^{-1}$. @Hartman2011 claim that ’ a possible lawsuit of the jitter is the presence of one or more additional planet ’. @Knutson2013 also analysed the RV signature of HAT - P-32 and found a long condition trend indicating a companion with a minimum bulk of $ 5 - 500\:$M$_{jup}$ at separations of $ 3.5 - 12\:$AU. However, such a companion could not yet explain the inadequate time - scale jitter as seen in the Hartman data.
Besides the circular eye socket fit, an eccentric solution with $ e=0.163 $ also fits the observed data. Though @Hartman2011 mention that the probability of a real non - zero eccentricity is only $ \sim3\%$, it could be mimicked or triggered by a second body in the system. Thus, HAT - P-32b is an ideal candidate for further monitoring to look for Transit Timing Variations induced by a planetal companion.
Data acquisition and reduction { # sec: DataAquisitionAndReduction }
= = = = = = = = = = = = = = = = = = = = = = = = = = = = = =
Between 2011 October and 2013 January we do 30 accomplished and 15 partial transit observations (determine Tables \[tab: H32completeObservations\ ] and \[tab: H32partialObservations\ ]) from 10 different observatories: our own one in Jena, as well as from telescopes at Torun (Poland), Trebur (Germany), Gettysburg and Swarthmore (USA), Tenerife and Sierra Nevada (Spain), Antalya and Ankara (Turkey), and Rozhen (Bulgaria) largely throughout the YETI network [ @YETI ]. In addition, three literature data points from @Sada, and two observations from @Gibson2013 are available. The telescopes and abbreviation used hereafter are summarized in Table \[tab: H32Telescopes\ ], a short description of each note web site can be found below, sorted by the number of observations.
\ # Observatory Telescope (abbreviation) $ \ | $P=2.15\:$d by @Hartman2011. Having a hosu star brightness of $V=11.3\:$maj and a planetafy transit depth of $21\:$mmag the swnsituvity of medium-sized tdlescopes is suffucieit to achieve hijg timinn predlsion, vherefore it is an optiman target for tve YZTI telescopes. The RV signal of HAT-P-32 is domonwted by high jytteg jf $>60\:$ma$^{-1}$. @Hartman2011 claim that ’a possible cahse of uhe jitter is the lresence of one or more adfitilnal planets’. @Knutsln2013 also anaoyseq the RV signxture of HAT-P-32 and founs a long term trend indicating x com'anion with a mijhmum mass oh $5-500\:$M$_{jup}$ at separations of $3.5-12\:$AG. Howevrr, such a compsnimn xould not yet explain the short time-scaje jitter aa seen in the Harrmqn daja.
Beshdes rhe cidcnlad orbih fmt, an eccenfric solutiin with $e=0.163$ also fits trv observed dafa. Thotgr @Hartman2011 mention that the probability mf z real non-zero eccentrixity is only $\sim3\%$, it cluld be mymicked or triggered by a second body in the systam. Thns, HAU-P-32n kw wn ideal candidate for further monitoring to jkol nor Transit Timikg Variations induvef ni a planetary zompannkn.
Sata acquisition ajd redustion {#sec:DataAzuisotionAndReduction}
==============================
Between 2011 Oxtober and 2013 Lanyary we performed 30 complete aud 15 pattial yransit observations (sez Tablss \[tab:H32complfteObservzgions\] and \[tab:H32paftislMbservations\]) from 10 differegt observetorizs: our odn ome in Tena, as wepl as from telescopes at Toruu (Poldnd), Trebur (Germany), Gettysburg and Swarthmore (USA), Tenerige ang Sierra Nevaca (Spain), Antajya and Ankara (Turkey), and Fozhen (Bulfaria) mmstly throudhout the YETH network [@YETM]. In addieion, thrwe litefxture data poimts from @Sada, and tqo observations frpm @Ejbson2013 are availcylt. Rhe telescopes ana afbgevmatiogv used hereaxter ard summxrized in Tcbue \[tan:H32Telescopes\], a short desdription of each onscrving sije can be found below, xorted by the numbtr of mbsxrvatipns.
\# Observatory Telescope (aybreviation) $\ | $P=2.15\:$d by @Hartman2011. Having a host star $V=11.3\:$mag a planetary depth of $21\:$mmag is to achieve high precision, therefore it an optimal target for the YETI The RV signal of HAT-P-32 is dominated by high jitter of $>60\:$ms$^{-1}$. @Hartman2011 that ’a possible cause of the jitter is the presence of one or additional @Knutson2013 analysed RV signature of HAT-P-32 and found a long term trend indicating a companion with a minimum of $5-500\:$M$_{jup}$ at separations of $3.5-12\:$AU. However, such companion could not yet the short time-scale jitter as in Hartman data. the orbit an eccentric solution $e=0.163$ also fits the observed data. Though @Hartman2011 mention that the probability of a real non-zero eccentricity only $\sim3\%$, be mimicked triggered a body in the HAT-P-32b is an ideal candidate for look for Transit Timing Variations induced by a companion. Data and reduction {#sec:DataAquisitionAndReduction} ============================== Between 2011 and 2013 January we performed 30 complete and partial transit observations (see Tables \[tab:H32completeObservations\] and \[tab:H32partialObservations\]) from 10 different observatories: our own one as well as from at Torun (Poland), (Germany), and (USA), and Sierra (Spain), Antalya and Ankara (Turkey), and Rozhen (Bulgaria) mostly throughout the network [@YETI]. In addition, three literature data points from @Sada, observations @Gibson2013 are available. telescopes and abbreviations used are in Table \[tab:H32Telescopes\], a of observing found sorted the number of observations. Observatory Telescope (abbreviation) $\ | $P=2.15\:$d by @Hartman2011. Having a host staR brightnesS of $V=11.3\:$mAg aNd a PlAnetAry tRansit depth of $21\:$mMAg thE sensitivity of medium-siZed teLeSCopeS Is SuffiCient to AChIEVe hIgH tImiNg PReCisioN, thErefore It is an optiMal TaRget for the YEti tElescopes. THe Rv signal of HAT-p-32 is DominaTeD by HIgh jiTteR of $>60\:$ms$^{-1}$. @hartmaN2011 Claim tHat ’a possiBlE Cause oF The jittER Is The pResence of one or morE AdDItional planets’. @knutsoN2013 aLSo ANAlySed The RV signaTuRe of Hat-P-32 and foUNd A LONg tERm trend indicaTing a companIOn wIth a miNiMum MAss of $5-500\:$M$_{Jup}$ at SePAraTions of $3.5-12\:$AU. HoWeveR, such a comPanion COuld not YEt explaIn the sHorT tiMe-scALe JiTteR aS SeeN In The hArtMan data.
BEsIdEs the CircULAR OrbiT fiT, an eCcentRic solution wiTh $e=0.163$ Also FIts The obServeD datA. THough @hartmaN2011 mentIoN that the probabiLity Of a real noN-zeRo EccEnTriciTY is onlY $\siM3\%$, it Could be MimickeD Or tRiGGEReD by a second body in thE sYSTeM. Thus, HAT-p-32b is an IDeAl CAndidate FoR fuRtheR MOnitoRing TO lOok for TrAnsit TIMiNg variatiOnS inducEd By a PlaNetarY CompAnion.
DAta acquiSitioN And reduction {#seC:dataAquisitioNanDrEdUCtioN}
==============================
BeTween 2011 OctobeR and 2013 jAnuaRy we PErForMEd 30 comPlete AnD 15 PaRTial transit observatIoNs (see TAbles \[Tab:H32completeOBservationS\] AND \[tab:H32parTialoBsERvations\]) from 10 diFfereNt observatORies: our oWn one In Jena, as Well as froM TElescopeS at torUn (POlaND), trEbur (Germany), GeTTYsbuRg And SwarThmOre (USA), TEneRifE anD SiErRa Nevada (SPain), AntaLyA aNd anKarA (TurkEY), and RozhEn (bulGaRia) MostlY ThrougHout tHe YEtI NeTWorK [@YETI]. In ADdITIon, tHrEe LiteRatUrE data PoinTS frOm @Sada, aNd two obseRvaTIons FrOm @gibson2013 aRe available. ThE tElescopes aNd AbbReviatIONs used heReafter are summarized in TABle \[tab:H32telEscopEs\], a sHort descrIptIon of eAch OBserviNg site Can be FoUnd BELow, soRTEd By tHe Number of obSERvaTions.
\# obServAtory TeLescope (abbreviatioN) $\ | $P=2.15\:$d by @Hartman20 11. Having a ho ststa rbrig htne ss of $V=11.3\ : $mag and a planetary trans it de pt h of$ 21 \:$mm ag thes en s i tiv it yofme d iu m-siz edtelesco pes is suf fic ie nt to achiev e h igh timing pr ecision, the ref ore it i s a n opti mal targ et for the YE TI telesc op e s. The RV sign a l o f HA T-P-32 is dominat e db y high jitterof $>6 0\ : $m s $ ^{- 1}$ . @Hartman 20 11 cl a im that ’a p o ssi b le cause of t he jitter i s th e pres en ceo f oneor mo re add itional pla nets ’. @Knuts on2013 also an a lysed t he RVsig nat ureo fHA T-P -3 2 an d f oun d along ter mtr end i ndic a t i n g acom pani on wi th a minimummas s of $5- 500\: $M$_{ jup} $at se parati ons o f$3.5-12\:$AU. H owev er, sucha c om pan io n cou l d notyet ex plain t he shor t ti me - s c al e jitter as seen i nt h eHartmandata.Be si d es the c ir cul ar o r b it fi t, a n e ccentric solut i on w ith $e= 0. 163$ a ls o f its theo bser ved da ta. Thou gh @H a rtman2011 ment i on that the p r ob a b il i ty o f a real non-z eroe ccen tric i ty is only$\sim 3\ % $, it could be mimicke dor tri ggere d by a second body in t h e system.Thus , H A T-P-32b is anideal candidate for furt her m onitorin g to look f or Trans itTim ing Va r i at ions inducedb y a p la netarycom panion.
D ata ac qui si tion andreductio n{# se c: Dat aAqui s itionAnd Re duc ti on}
==== = ====== ===== ==== == == = === ==
Bet w ee n 2011 O ct ober an d2013Janu a rywe perf ormed 30com p lete a nd 15 par tial transitob servations ( see Table s \[tab:H3 2completeObservations\] and \[t ab: H32pa rtia lObservat ion s\]) f rom 10 dif ferent obse rv ato r i es: o u r o wnon e in Jena, a s w ell a sfrom telesc opes at Torun (Pol a nd) , Trebur (Ger man y),G e tt ysb u rg and S w art h m ore (USA), Tene rife and S ie r ra Nevada (S p ain ), Antaly a and A nkara (Turkey ), and Ro zhen (Bul ga ria) m ost ly through out theYETI netw o rk [@ Y ET I]. I n a dditio n, th ree l iterat u redatapoints f rom @S ada,an d two ob servations from @Gibson 2013 a re av ail able. The te l esc opes andabbr eviationsuse d h ereaf ter are s umma r iz edi n Tab le \ [ tab:H32Te l es cop e s \] , a short d e s c rip tionofe ach ob serv ing site can be f o und below, sor tedb y th e n u mber o f observations .
\ # Obser va tory T el e scope(abb rev iation) $\ | $P=2.15\:$d_by @Hartman2011._Having a host star_brightness of_$V=11.3\:$mag_and a_planetary_transit depth of_$21\:$mmag the sensitivity_of medium-sized telescopes is_sufficient to achieve_high_timing precision, therefore it is an optimal target for the YETI telescopes. The RV_signal_of HAT-P-32_is_dominated_by high jitter of $>60\:$ms$^{-1}$._@Hartman2011 claim that ’a possible_cause of_the jitter is the presence of one or_more_additional planets’. @Knutson2013_also analysed the RV signature of HAT-P-32 and found_a long term trend indicating a_companion with a_minimum_mass_of $5-500\:$M$_{jup}$ at separations_of $3.5-12\:$AU. However, such a companion_could not yet explain the short_time-scale jitter as seen in the Hartman_data.
Besides the circular orbit fit, an_eccentric solution with $e=0.163$ also_fits the_observed data. Though @Hartman2011 mention_that the probability_of a_real non-zero eccentricity_is only $\sim3\%$, it could be_mimicked or triggered_by a second body in the_system._Thus, HAT-P-32b is_an_ideal_candidate for_further monitoring to_look_for Transit_Timing_Variations induced by a planetary companion.
Data_acquisition_and reduction {#sec:DataAquisitionAndReduction}
==============================
Between 2011 October and 2013_January we performed 30_complete_and 15 partial transit_observations (see Tables \[tab:H32completeObservations\] and \[tab:H32partialObservations\])_from 10 different observatories: our own_one in_Jena, as_well as from telescopes at Torun (Poland), Trebur (Germany), Gettysburg and_Swarthmore (USA), Tenerife and Sierra Nevada_(Spain), Antalya and Ankara_(Turkey), and_Rozhen_(Bulgaria) mostly throughout_the_YETI network_[@YETI]. In addition, three literature data points_from @Sada,_and two observations from @Gibson2013 are_available. The telescopes and_abbreviations_used hereafter are summarized in Table \[tab:H32Telescopes\],_a short description of each observing_site can be found below,_sorted_by_the number of observations.
_\# Observatory _ _ __ __ ___ _ _ Telescope (abbreviation) _ _ ___ $\ |
right)},\label{eq:lax_vector}$$ the frequency spectrum is related to the roots of the equation $\sqrt{L^{2}\left(u\right)}=0$. We analyze them numerically in the limit $N\gg1$, and find that, in the high coupling regime $\Omega\gg\Delta$, all roots merge into a continuous line except two complex conjugated pairs, see example in Figure \[fig:L\_roots\]. Note that all coefficients of the polynomial $L^{2}\left(u\right)$ are real thus every complex root has a complex conjugated partner. The dynamical variables that correspond to the continuous band form a decay part of the solution and the two discreet frequencies give an oscillating part that we will be interested in. The discreet roots can be found by turning the summation over $j$ in Eq. (\[eq:lax\_vector\]) into an integral, $\sum_{j}\rightarrow\frac{N}{\Delta}\int_{\omega-\Delta/2}^{\omega+\Delta/2}d\epsilon$. Then, the equation $\sqrt{L^{2}\left(u\right)}=0$ turns into $$\pm2iga\left(0\right)=u-\frac{V^{2}N}{\Delta}\ln\left(\frac{u+\Delta/2}{u-\Delta/2}\right),\label{eq:lax2}$$ where the total width of splittings $\Delta=\epsilon_{N}-\epsilon_1$ is finite.
In the limit $M=N$, opposite to $M\ll N$, the roots of Eq. (\[eq:lax2\]) have zero real part. Parametrizing the roots as $u=iu_{0}\Delta/2$ we obtain
$$\frac{\pm4Va\left(0\right)}{\Delta}=u_{0}+\frac{2V^{2}N}{\Delta^{2}}\left(\pi-2\tan^{-1}u_{0}\right).$$
Then a $1/u_{0}$-expansion gives the imaginary parts as $u_{1,2}=u_{0}\Delta/2=\pm\left(\frac{V\sqrt{N}}{2}\pm\frac{\Delta}{2\sqrt{3}}\right)$.
In the case of two discreet roots the hyperelliptic function of many variables reduces to an elliptic function of only one variable [@relaxation] which corresponds to an effective model of a single | right)},\label{eq: lax_vector}$$ the frequency spectrum is related to the roots of the equation $ \sqrt{L^{2}\left(u\right)}=0$. We analyze them numerically in the limit $ N\gg1 $, and witness that, in the eminent coupling regime $ \Omega\gg\Delta$, all roots blend into a continuous line except two complex conjugated pair, experience example in Figure \[fig: L\_roots\ ]. notice that all coefficient of the polynomial $ L^{2}\left(u\right)$ are real thus every complex etymon has a complex conjugated partner. The dynamic variables that correspond to the continuous isthmus form a decay part of the solution and the two discerning frequencies give an oscillating part that we will be concerned in. The discreet roots can be find by turn the summation over $ j$ in Eq. (\[eq: lax\_vector\ ]) into an built-in, $ \sum_{j}\rightarrow\frac{N}{\Delta}\int_{\omega-\Delta/2}^{\omega+\Delta/2}d\epsilon$. Then, the equation $ \sqrt{L^{2}\left(u\right)}=0 $ turns into $ $ \pm2iga\left(0\right)=u-\frac{V^{2}N}{\Delta}\ln\left(\frac{u+\Delta/2}{u-\Delta/2}\right),\label{eq: lax2}$$ where the total width of splittings $ \Delta=\epsilon_{N}-\epsilon_1 $ is finite.
In the limit $ molarity = N$, opposite to $ M\ll N$, the roots of Eq. (\[eq: lax2\ ]) have zero real part. Parametrizing the etymon as $ u = iu_{0}\Delta/2 $ we obtain
$ $ \frac{\pm4Va\left(0\right)}{\Delta}=u_{0}+\frac{2V^{2}N}{\Delta^{2}}\left(\pi-2\tan^{-1}u_{0}\right).$$
Then a $ 1 / u_{0}$-expansion gives the imaginary parts as $ u_{1,2}=u_{0}\Delta/2=\pm\left(\frac{V\sqrt{N}}{2}\pm\frac{\Delta}{2\sqrt{3}}\right)$.
In the case of two discreet roots the hyperelliptic function of many variables shrink to an elliptic function of only one variable star [ @relaxation ] which represent to an effective model of a single | rigjt)},\label{eq:lax_vector}$$ the fvequency spectrum is renated fo the ruots of the equation $\sqrt{L^{2}\leht(u\rught)}=0$. Qe analyze them numerizally in nhe limit $N\gg1$, qnd find tizt, in tmz higg cou'lmng regime $\Omegs\gg\Delta$, anl roots merge ivtl a continuous line except two compjex conkuhated pairs, seg exak[le jn Figure \[fig:L\_roots\]. Note that all doefficpents of the polymomial $L^{2}\left(u\right)$ are reap thks every complex rlot has a cimplqz conjugated partner. The dynamical variables that correspond to tfe coutinuous babd fogk a decay pert of the solutiok and tve two ciscreet frequcnciev guve an oscillating pact that we will be igterested iu. The discreet roots xab be xoung by rurvinf vhe summahioi over $j$ in Eq. (\[eq:lax\_vextor\]) into an integrsl, $\wum_{j}\rightarros\frac{N}{\Qejta}\int_{\omega-\Delta/2}^{\omega+\Delta/2}d\epsilon$. Then, ths equation $\sqrt{L^{2}\left(u\ritht)}=0$ turns into $$\pm2iga\lgft(0\right)=u-\fwac{V^{2}N}{\Delta}\ln\left(\frac{u+\Delta/2}{u-\Delta/2}\right),\label{eq:lax2}$$ wvere vhd titwu wldth of splittings $\Delta=\epsilon_{N}-\epsilon_1$ is figjtt.
In the limit $M=N$, jpposite to $M\pl G$, the roots ow Eq. (\[eq:las2\]) have zero real pwrt. Parwmetruzing the rooys as $u=iu_{0}\Delta/2$ we obtain
$$\frax{\pm4Va\left(0\rigkt)}{\Dwlta}=u_{0}+\frac{2V^{2}N}{\Delta^{2}}\leyt(\pi-2\tan^{-1}u_{0}\righc).$$
Then s $1/u_{0}$-exlansion gives the imagiuary pzrts as $u_{1,2}=u_{0}\Dflta/2=\pm\leff(\wrac{V\sqrt{N}}{2}\pm\frac{\Aelna}{2\sqst{3}}\right)$.
In the case of two qiscreet coots the hyoerekliptis function of many variables reducfs to at elliptic function of only one variable [@cxlaxation] whivh cogresponds to ak effective modql of a single | right)},\label{eq:lax_vector}$$ the frequency spectrum is related to of equation $\sqrt{L^{2}\left(u\right)}=0$. analyze them numerically find in the high regime $\Omega\gg\Delta$, all merge into a continuous line except complex conjugated pairs, see example in Figure \[fig:L\_roots\]. Note that all coefficients of polynomial $L^{2}\left(u\right)$ are real thus every complex root has a complex conjugated partner. dynamical that to continuous band form a decay part of the solution and the two discreet frequencies give an part that we will be interested in. The roots can be found turning the summation over $j$ Eq. into an $\sum_{j}\rightarrow\frac{N}{\Delta}\int_{\omega-\Delta/2}^{\omega+\Delta/2}d\epsilon$. the $\sqrt{L^{2}\left(u\right)}=0$ turns into where the total width of splittings $\Delta=\epsilon_{N}-\epsilon_1$ is finite. In the limit $M=N$, opposite to $M\ll N$, roots of have zero part. the as $u=iu_{0}\Delta/2$ we Then a $1/u_{0}$-expansion gives the imaginary In the case of two discreet roots the function of variables reduces to an elliptic function only one variable [@relaxation] which corresponds to an model of a single | right)},\label{eq:lax_vector}$$ the fRequency spEctruM is RelAtEd to The rOots of the equatIOn $\sqRt{L^{2}\left(u\right)}=0$. We analyze Them nUmERicaLLy In the Limit $N\gG1$, AnD FInd ThAt, In tHe HIgH coupLinG regime $\omega\gg\DelTa$, aLl Roots merge inTO a Continuous LinE except two coMplEx conjUgAteD Pairs, See ExampLe in FiGUre \[fig:l\_roots\]. NotE tHAt all cOEfficieNTS oF the Polynomial $L^{2}\left(u\rIGhT)$ Are real thus eveRy compLeX RoOT Has A coMplex conjuGaTed paRTner. The DYnAMICal VAriables that cOrrespond to THe cOntinuOuS baND form a Decay PaRT of The solution And tHe two discReet frEQuencieS Give an oScillaTinG paRt thAT wE wIll Be INteREsTed IN. ThE discreeT rOoTs can Be foUND BY turNinG the SummaTion over $j$ in Eq. (\[Eq:lAx\_veCTor\]) Into aN inteGral, $\SuM_{j}\rigHtarroW\frac{n}{\DElta}\int_{\omega-\DelTa/2}^{\omEga+\Delta/2}d\EpsIlOn$. THeN, the eQUation $\SqrT{L^{2}\lEft(u\rigHt)}=0$ turns INto $$\Pm2IGA\LeFt(0\right)=u-\frac{V^{2}N}{\DeltA}\lN\LEfT(\frac{u+\DeLta/2}{u-\DeLTa/2}\RiGHt),\label{eQ:lAx2}$$ wHere THE totaL widTH oF splittiNgs $\DelTA=\ePsIlon_{N}-\epSiLon_1$ is fInIte.
in tHe limIT $M=N$, oPpositE to $M\ll N$, tHe rooTS of Eq. (\[eq:lax2\]) have ZEro real part. PaRAmETRiZIng tHe rOots as $u=iu_{0}\DeLta/2$ wE ObtaIn
$$\frAC{\pM4Va\LEft(0\riGht)}{\DeLtA}=U_{0}+\fRAc{2V^{2}N}{\Delta^{2}}\left(\pi-2\tan^{-1}u_{0}\RiGht).$$
TheN a $1/u_{0}$-exPansion gives tHe imaginarY PARts as $u_{1,2}=u_{0}\DElta/2=\PM\lEFt(\frac{V\sqrt{N}}{2}\pm\Frac{\DElta}{2\sqrt{3}}\riGHt)$.
In the cAse of Two discrEet roots tHE HyperellIptIc fUncTioN OF mAny variables rEDUces To An ellipTic FunctioN of OnlY onE vaRiAble [@relaxAtion] whiCh CoRrEsPonDs to aN EffectivE mOdeL oF a sIngle | right)},\label{eq:lax_vect or}$$ thefrequ enc y s pe ctru m is related to th e roo ts of the equation $\s qrt{L ^{ 2 }\le f t( u\rig ht)}=0$ . W e ana ly ze th em nu meric all y in th e limit $N \gg 1$ , and find t h at , in the h igh coupling re gim e $\Om eg a\g g \Delt a$, allrootsm erge i nto a con ti n uous l i ne exce p t t wo c omplex conjugated pa i rs, see exampl e in F ig u re \ [fi g:L \_roots\]. N ote t h at allc oe f f i cie n ts of the pol ynomial $L^ { 2}\ left(u \r igh t )$ are real t h usevery compl ex r oot has a compl e x conju g ated pa rtner. Th e d ynam i ca lvar ia b les th atc orr espond t oth e con tinu o u s band fo rm a deca y part of the so luti o n a nd th e two dis cr eet f requen ciesgi ve an oscillati ng p art thatwewi llbe inte r estedin. Th e discr eet roo t s c an b e f ound by turning th es u mm ation ov er $j$ in E q . (\[eq: la x\_ vect o r \]) i ntoa nintegral , $\su m _{ j} \righta rr ow\fra c{ N}{ \De lta}\ i nt_{ \omega -\Delta/ 2}^{\ o mega+\Delta/2} d \epsilon$. Th e n, t he equa tio n $\sqrt{L^ {2}\ l eft( u\ri g ht )}= 0 $ tur ns in to $$ \ pm2iga\left(0\right )= u-\fra c{V^{ 2}N}{\Delta}\ ln\left(\f r a c {u+\Delt a/2} { u- \ Delta/2}\right ),\la bel{eq:lax 2 }$$ wher e the total w idth of s p l ittings$\D elt a=\ eps i l on _{N}-\epsilon _ 1 $ is f inite.
In the li mit $M =N$ , o pp osite to$M\ll N$ ,th ero ots of E q . (\[eq: la x2\ ]) ha ve ze r o real part . Pa ra me t riz ing the ro o t s as $ u= iu_{ 0}\ De lta/2 $ we obt ain
$$ \frac{\pm 4Va \ left (0 \r ight)}{ \Delta}=u_{0} +\ frac{2V^{2 }N }{\ Delta^ { 2 }}\left( \pi-2\tan^{-1}u_{0}\rig h t).$$
The n a $ 1/u_ {0}$-expa nsi on giv est he ima ginary part sas$ u _{1,2 } = u_ {0} \D elta/2=\pm \ l eft (\fra c{ V\sq rt{N}}{ 2}\pm\frac{\Delta} { 2\s qrt{3}}\right )$.
In t he ca s eo f t wo dis c r eet roots the h yperellipt ic fu nction ofm any v ariable s reduc es to an elli ptic func tion of o nl y on e var iable [@re laxation ] which c o rresp o nd s toaneffect iv e m odelof a s i ngl e | right)},\label{eq:lax_vector}$$ the_frequency spectrum_is related to the_roots of_the_equation $\sqrt{L^{2}\left(u\right)}=0$._We_analyze them numerically_in the limit_$N\gg1$, and find that,_in the high_coupling_regime $\Omega\gg\Delta$, all roots merge into a continuous line except two complex conjugated pairs,_see_example in_Figure_\[fig:L\_roots\]._Note that all coefficients of_the polynomial $L^{2}\left(u\right)$ are real_thus every_complex root has a complex conjugated partner. The_dynamical_variables that correspond_to the continuous band form a decay part of_the solution and the two discreet_frequencies give an_oscillating_part_that we will be_interested in. The discreet roots can_be found by turning the summation_over $j$ in Eq. (\[eq:lax\_vector\]) into an_integral, $\sum_{j}\rightarrow\frac{N}{\Delta}\int_{\omega-\Delta/2}^{\omega+\Delta/2}d\epsilon$. Then, the equation $\sqrt{L^{2}\left(u\right)}=0$_turns into $$\pm2iga\left(0\right)=u-\frac{V^{2}N}{\Delta}\ln\left(\frac{u+\Delta/2}{u-\Delta/2}\right),\label{eq:lax2}$$ where the_total width_of splittings $\Delta=\epsilon_{N}-\epsilon_1$ is finite.
In_the limit $M=N$,_opposite to_$M\ll N$, the_roots of Eq. (\[eq:lax2\]) have zero_real part. Parametrizing_the roots as $u=iu_{0}\Delta/2$ we obtain
$$\frac{\pm4Va\left(0\right)}{\Delta}=u_{0}+\frac{2V^{2}N}{\Delta^{2}}\left(\pi-2\tan^{-1}u_{0}\right).$$
Then_a_$1/u_{0}$-expansion gives the_imaginary_parts_as $u_{1,2}=u_{0}\Delta/2=\pm\left(\frac{V\sqrt{N}}{2}\pm\frac{\Delta}{2\sqrt{3}}\right)$.
In_the case of_two_discreet roots_the_hyperelliptic function of many variables reduces_to_an elliptic function of only one variable_[@relaxation] which corresponds to_an_effective model of a_single |
theories, see [@Derendinger_linear] and references therein.
---
abstract: 'The notion of $2$-almost Gorenstein local ring ($2$-$\AGL$ ring for short) is a generalization of the notion of almost Gorenstein local ring from the point of view of Sally modules of canonical ideals. In this paper, for further developments of the theory, we discuss three different topics on $2$-$\AGL$ rings. The first one is to clarify the structure of minimal presentations of canonical ideals, and the second one is the study of the question of when certain fiber products, so called amalgamated duplications are $2$-$\AGL$ rings. We also explore Ulrich ideals in $2$-$\AGL$ rings, mainly two-generated ones.'
address:
- 'Department of Mathematics, School of Science and Technology, Meiji University, 1-1-1 Higashi-mita, Tama-ku, Kawasaki 214-8571, Japan'
- 'Department of Mathematics and Informatics, Graduate School of Science and Technology, Chiba University, 1-33 Yayoi-cho, Inage-ku, Chiba, 263-8522, Japan'
- 'Global Education Center, Waseda University, 1-6-1 Nishi-Waseda, Shinjuku-ku, Tokyo 169-8050, Japan'
author:
- Shiro Goto
- Ryotaro Isobe
- Naoki Taniguchi
title: 'Ulrich ideals and $2$-$\AGL$ rings'
---
Introduction
============
The series [@CGKM; @GGHV; @GMP; @GMTY1; @GMTY2; @GMTY3; @GMTY4; @GRTT; @GTT; @GTT2] of researches are motivated and supported by the strong desire to stratify Cohen-Macaulay rings, finding new and interesting classes which naturally include that of Gorenstein rings. As is already pointed out by these works, the class of [*almost Gorenstein local rings*]{} (AGL rings for short) could be a very nice candidate for such classes. The prototype of AGL rings is found in the work [@BF] of V. Barucci and R. Fröberg in 1997, where they introduced the notion of AGL ring for one-dimensional analytically unramified local rings, developing | theories, see [ @Derendinger_linear ] and references therein.
---
abstract:' The notion of $ 2$-almost Gorenstein local gang ($ 2$-$\AGL$ gang for short) is a generalization of the notion of about Gorenstein local ring from the point of view of Sally faculty of canonical ideals. In this composition, for further developments of the theory, we hash out three different topics on $ 2$-$\AGL$ hoop. The first one is to clarify the structure of minimal presentations of canonical ideals, and the second one is the study of the doubt of when certain fiber products, indeed called amalgamated duplications are $ 2$-$\AGL$ rings. We besides explore Ulrich ideal in $ 2$-$\AGL$ rings, mainly two - generated one.'
address:
-' Department of Mathematics, School of Science and Technology, Meiji University, 1 - 1 - 1 Higashi - mita, Tama - ku, Kawasaki 214 - 8571, Japan'
-' Department of Mathematics and Informatics, Graduate School of Science and Technology, Chiba University, 1 - 33 Yayoi - cho, Inage - ku, Chiba, 263 - 8522, Japan'
-' Global Education Center, Waseda University, 1 - 6 - 1 Nishi - Waseda, Shinjuku - ku, Tokyo 169 - 8050, Japan'
author:
- Shiro Goto
- Ryotaro Isobe
- Naoki Taniguchi
title:' Ulrich ideals and $ 2$-$\AGL$ rings'
---
Introduction
= = = = = = = = = = = =
The series [ @CGKM; @GGHV; @GMP; @GMTY1; @GMTY2; @GMTY3; @GMTY4; @GRTT; @GTT; @GTT2 ] of researches are motivated and supported by the strong desire to stratify Cohen - Macaulay rings, finding new and interesting classes which naturally include that of Gorenstein rings. As is already point out by these works, the class of [ * about Gorenstein local hoop * ] { } (AGL rings for short) could be a very nice campaigner for such classes. The prototype of AGL ring is found in the work [ @BF ] of V. Barucci and R. Fröberg in 1997, where they introduce the notion of AGL ring for one - dimensional analytically unramified local rings, developing | thfories, see [@Derendinger_likear] and referenews thecein.
---
abatract: 'Tfe notion of $2$-almost Gorenstemn lical eing ($2$-$\AGL$ ring for shorg) is a gejeralizarion if the notmkn of almost Flrenvvein local ring from the [oint of view mf Sclly modules of canonical ideals. In ehis paleg, for further qevekjpmehns of the theory, we discuss thrse difftrent topics on $2$-$\AGK$ rings. The first one is tl clwrify the structurf of minimao prqwentations ow canonical ideals, and the second one is the study of the auestion od qhej certain fiuer prjducts, so called amangamatec duplications arx $2$-$\AGO$ rings. We also exploce Ulrich ideals in $2$-$\WGL$ rings, mcinly two-generated onws.'
qddrevs:
- 'Dapargneng or Kafhematlcs, School of Science ane Technology, Meiji Inynrrsity, 1-1-1 Higaahi-mitw, Eama-ku, Kawasaki 214-8571, Japan'
- 'Department of Mauhemafics and Informatics, Grqduate School of Sciejce and Tqchnology, Chiba University, 1-33 Yayoi-cho, Inage-ku, Chibd, 263-8522, Ja'av'
- 'Goonal Deufation Center, Waseda University, 1-6-1 Nishi-Waseda, Fgimjlku-ku, Tokyo 169-8050, Japak'
author:
- Shiro Goto
- Rjoywro Isobe
- Naoyi Tannfudhi
title: 'Ulrich idfals anq $2$-$\AGL$ rings'
---
Intwoduvtion
============
The series [@CGKM; @GGHV; @GNP; @GMTY1; @GMTY2; @TMTY3; @GMTY4; @GRTT; @GTT; @GTT2] of reszarchex are motivated and supporteb by tge strong dfsire to agratify Cohen-Macxulsy rings, finding new and intqresting rlasszs which natorally ynclude thwt of Gorenstein rings. Ad is clreagy pointed out by these works, the class oh [*almost Goremsdeit local xings*]{} (SGL rings for short) could bg a very uice cxndidate fkr such classes. Thq prototype ox AGL rings iv found yn tye wirk [@BF] uw V. Barucci anc R. Fröbery in 1997, where they introdmced jhs notion of AGL ribg for one-dimenxioval ajakyehcally unramhfiea luval rkngs, develoiine | theories, see [@Derendinger_linear] and references therein. --- notion $2$-almost Gorenstein ring ($2$-$\AGL$ ring of notion of almost local ring from point of view of Sally modules canonical ideals. In this paper, for further developments of the theory, we discuss different topics on $2$-$\AGL$ rings. The first one is to clarify the structure minimal of ideals, the second one is the study of the question of when certain fiber products, so called duplications are $2$-$\AGL$ rings. We also explore Ulrich in $2$-$\AGL$ rings, mainly ones.' address: - 'Department of School Science and Meiji 1-1-1 Tama-ku, Kawasaki 214-8571, - 'Department of Mathematics and Informatics, Graduate School of Science and Technology, Chiba University, 1-33 Yayoi-cho, Inage-ku, 263-8522, Japan' Education Center, University, Nishi-Waseda, Tokyo 169-8050, Japan' Shiro Goto - Ryotaro Isobe - 'Ulrich ideals and $2$-$\AGL$ rings' --- Introduction ============ series [@CGKM; @GMP; @GMTY1; @GMTY2; @GMTY3; @GMTY4; @GRTT; @GTT2] of researches are motivated and supported by strong desire to stratify Cohen-Macaulay rings, finding new and interesting classes which naturally include that rings. As is already out by these the of Gorenstein rings*]{} (AGL for short) could be a very nice candidate for such classes. prototype of AGL rings is found in the work [@BF] Barucci R. Fröberg in where they introduced the of ring for one-dimensional analytically rings, | theories, see [@Derendinger_linEar] and refeRenceS thEreIn.
---
AbstRact: 'the notion of $2$-almOSt GoRenstein local ring ($2$-$\AGL$ riNg for ShORt) is A GeNeralIzation OF tHE NotIoN oF alMoST GOrensTeiN local rIng from the PoiNt Of view of SallY MoDules of canOniCal ideals. In tHis Paper, fOr FurTHer deVelOpmenTs of thE Theory, We discuss ThREe diffERent topICS oN $2$-$\AGL$ Rings. The first one iS To CLarify the strucTure of MiNImAL PreSenTations of cAnOnicaL Ideals, aND tHE SEcoND one is the studY of the questIOn oF when cErTaiN Fiber pRoducTs, SO caLled amalgamAted DuplicatiOns are $2$-$\agL$ rings. wE also exPlore ULriCh iDealS In $2$-$\aGl$ riNgS, MaiNLy Two-GEneRated oneS.'
aDdRess:
- 'DEparTMENT of MAthEmatIcs, ScHool of Science And techNOloGy, MeiJi UniVersItY, 1-1-1 HigaShi-mitA, Tama-Ku, kawasaki 214-8571, Japan'
- 'DePartMent of MatHemAtIcs AnD InfoRMatics, graDuaTe SchooL of ScieNCe aNd tECHnOlogy, Chiba UniversiTy, 1-33 yAYoI-cho, InagE-ku, ChiBA, 263-8522, JApAN'
- 'Global EDuCatIon CENTer, WaSeda uNiVersity, 1-6-1 NIshi-WaSEdA, SHinjuku-Ku, tokyo 169-8050, JApAn'
aUthOr:
- ShiRO GotO
- RyotaRo Isobe
- NAoki TANiguchi
title: 'UlRIch ideals and $2$-$\Agl$ rINGs'
---
iNtroDucTion
============
The seriEs [@CGkm; @GGHv; @GMP; @gmTy1; @GMty2; @GMTY3; @gMTY4; @GrTt; @gTt; @gTT2] of researches are mOtIvated And suPported by the sTrong desirE TO Stratify coheN-maCAulay rings, findIng neW and intereSTing clasSes whIch naturAlly incluDE That of GoRenSteIn rIngS. aS iS already pointED Out bY tHese worKs, tHe class Of [*aLmoSt GOreNsTein local Rings*]{} (AGL RiNgS fOr ShoRt) couLD be a very NiCe cAnDidAte foR Such clAsses. the pRoToTYpe Of AGL riNGs IS FounD iN tHe woRk [@Bf] oF V. BarUcci ANd R. fröberg In 1997, where thEy iNTrodUcEd The notiOn of AGL ring foR oNe-dimensioNaL anAlyticALLy unramiFied local rings, developinG | theories, see [@Derending er_linear] andref ere nc es t here in.
---
abstr a ct:'The notion of $2$-alm ost G or e nste i nlocal ring ( $ 2$ - $ \AG L$ r ing f o rshort ) i s a gen eralizatio n o fthe notion o f a lmost Gore nst ein local ri ngfrom t he po i nt of vi ew of Sally module s of cano ni c al ide a ls. Int h is pap er, for further d e ve l opments of the theor y, we d isc uss three dif fe rentt opics o n $ 2 $ - $\A G L$ rings. The first onei s t o clar if y t h e stru cture o f mi nimal prese ntat ions of c anonic a l ideal s , and t he sec ond on e is th estu dy oft he qu e sti on of wh en c ertai n fi b e r prod uct s, s o cal led amalgamat eddupl i cat ionsare $ 2$-$ \A GL$ r ings.We al so explore Ulrich ide als in $2 $-$ \A GL$ r ings, mainly tw o-g enerate d ones. '
ad dr e s s :- 'Department of M at h e ma tics, Sc hool o f S ci e nce andTe chn olog y , Meij i Un i ve rsity, 1 -1-1 H i ga sh i-mita, T ama-ku ,Kaw asa ki 21 4 -857 1, Jap an'
- 'D epart m ent of Mathema t ics and Infor m at i c s, Grad uat e School of Sci e nceandT ec hno l ogy,Chiba U n iv e rsity, 1-33 Yayoi-c ho , Inag e-ku, Chiba, 263-8 522, Japan ' - 'Global Edu c at i on Center, Was eda U niversity, 1-6-1 Ni shi-W aseda, S hinjuku-k u , Tokyo 1 69- 805 0,Jap a n 'author:
- Shi r o Got o- Ryota roIsobe
- Na oki Ta nig uc hi
title: 'Ulrich i de al sand $2$- $ \AGL$ ri ng s'-- -
Intro d uction
==== ==== == ==
Th e serie s [ @ C GKM; @ GG HV;@GM P; @GMT Y1;@ GMT Y2; @GM TY3; @GMT Y4; @GRT T; @ GTT; @G TT2] of resea rc hes are mo ti vat ed and s upported by the strong desire t o strati fyCohen -Mac aulay rin gs, findi ngn ew and inter estin gcla s s es wh i c hnat ur ally inclu d e th at of G oren stein r ings. As is alread y po inted out bythe se w o r ks , t h ec las so f [ * a lmost Gorenstei n local ri ng s *] {} (AGL ri n gsfo r short ) could be a very ni ce candid ate for s uc h cl a s ses . The prot otype of AGL ring s is f o un d inthe work[@ BF] of V . Baru c ciand R . Fröb er g in 1 997,wh ere they introduced the notionof AGL ring fo r one-dim ens i ona l analyti call y unramifi edloc al ri ngs , deve lopi n g | theories,_see [@Derendinger_linear]_and references therein.
---
abstract:_'The notion_of_$2$-almost Gorenstein_local_ring ($2$-$\AGL$ ring_for short) is_a generalization of the_notion of almost_Gorenstein_local ring from the point of view of Sally modules of canonical ideals. In_this_paper, for_further_developments_of the theory, we discuss_three different topics on $2$-$\AGL$_rings. The_first one is to clarify the structure of_minimal_presentations of canonical_ideals, and the second one is the study of_the question of when certain fiber_products, so called_amalgamated_duplications_are $2$-$\AGL$ rings. We_also explore Ulrich ideals in $2$-$\AGL$_rings, mainly two-generated ones.'
address:
- 'Department of_Mathematics, School of Science and Technology, Meiji_University, 1-1-1 Higashi-mita, Tama-ku, Kawasaki 214-8571,_Japan'
- 'Department of Mathematics and_Informatics, Graduate_School of Science and Technology,_Chiba University, 1-33_Yayoi-cho, Inage-ku,_Chiba, 263-8522, Japan'
-_'Global Education Center, Waseda University, 1-6-1_Nishi-Waseda, Shinjuku-ku, Tokyo_169-8050, Japan'
author:
- Shiro Goto
- Ryotaro Isobe
-_Naoki_Taniguchi
title: 'Ulrich ideals_and_$2$-$\AGL$_rings'
---
Introduction
============
The series_[@CGKM; @GGHV; @GMP;_@GMTY1;_@GMTY2; @GMTY3;_@GMTY4;_@GRTT; @GTT; @GTT2] of researches are_motivated_and supported by the strong desire to_stratify Cohen-Macaulay rings, finding_new_and interesting classes which_naturally include that of Gorenstein_rings. As is already pointed out_by these_works, the_class of [*almost Gorenstein local rings*]{} (AGL rings for short) could_be a very nice candidate for_such classes. The prototype_of AGL_rings_is found in_the_work [@BF]_of V. Barucci and R. Fröberg in_1997, where_they introduced the notion of AGL_ring for one-dimensional analytically_unramified_local rings, developing |
1}$.
#### Causal Consistency (CC).
Causal consistency (CC) holds when transactions observe consistent snapshots of the system, and the client application processes are monotonic. CC is a weak consistency model and it does not allow solving consensus. It is in fact the strongest model that is available under partition [@syn:rep:1738]. Historically, CC refers to the consistency of single operations on a shared memory [@syn:mat:1025]. Causally consistent transactions (Causal HAT) is a consistency model that extends CC to transactional data stores [@db:syn:1751].
#### Snapshot Isolation (SI).
SI is a widely-used consistency model [@syn:bd:1759]. This model is strong, but allows more interleavings of concurrent read transactions than SER. Furthermore, SI is causal (i.e., SI $\subseteq$ CC), whereas SER is not.
Under SI, a transaction observes a *snapshot* of the state of the data store at some point prior in time. Strong snapshot isolation (SSI) requires this snapshot to contain all the preceding transactions in real time [@Daudjee:2006]. Two transactions may commit under SI as long as they do not write the same item concurrent. SI avoids the anomalies listed in, but exhibits the *write-skew* anomaly, illustrated in history $h_4$ below.
In this history, an application using data items $x$ and $y$ wishes to maintain the invariant $x \geq y$. The invariant holds initially, and each of the two transactions ${\color{blue}{T_1}}$ and ${\color{red}{T_2}}$ guarantees the invariant individually. As illustrated in history $h_4$, running them concurrently under SI may violate the invariant.
An application is *robust* against a consistency model $M$ when, it produces serialisable histories [@Cerone:2016], despite running atop a data store providing $M$, It is known [@syn:1687] that an application is robust against SI when every invariant is materialised by a data item.
#### Parallel [/]{} Non-Monotonic Snapshot Isolation (PSI[/]{}NMSI).
Parallel and non-monotonic snapshot isolation are scalable variations of SI. These models retain two core properties of SI, namely
each transaction observes a consistent snapshot, and
no | 1}$.
# # # # Causal Consistency (CC).
Causal consistency (CC) holds when transactions observe consistent snapshot of the arrangement, and the client application summons are flat. CC is a weak consistency model and it does not give up solving consensus. It is in fact the strongest exemplar that is available under partition [ @syn: rep:1738 ]. Historically, CC refer to the consistency of single operations on a partake memory [ @syn: mat:1025 ]. Causally consistent transactions (Causal HAT) is a consistency model that extends CC to transactional data shop [ @db: syn:1751 ].
# # # # Snapshot Isolation (SI).
SI is a widely - used consistency model [ @syn: bd:1759 ]. This mannequin is strong, but allows more interleavings of concurrent read transactions than SER. Furthermore, SI is causal (i.e., SI $ \subseteq$ CC), whereas SER is not.
Under SI, a transaction detect a * snapshot * of the state of the data store at some point prior in time. potent snapshot isolation (SSI) requires this snapshot to contain all the preceding transactions in real meter [ @Daudjee:2006 ]. Two transactions may commit under SI as long as they do not write the same item concurrent. SI avoids the anomalies listed in, but exhibits the * write - skew * anomaly, exemplify in history $ h_4 $ below.
In this history, an application use datum items $ x$ and $ y$ wishes to maintain the invariant $ x \geq y$. The invariant holds initially, and each of the two transactions $ { \color{blue}{T_1}}$ and $ { \color{red}{T_2}}$ guarantees the changeless individually. As illustrated in history $ h_4 $, running them concurrently under SI may violate the invariant.
An application is * robust * against a consistency model $ M$ when, it produces serialisable histories [ @Cerone:2016 ], despite running atop a data store provide $ M$, It is known [ @syn:1687 ] that an lotion is full-bodied against SI when every changeless is materialised by a data item.
# # # # Parallel [ / ] { } Non - Monotonic Snapshot Isolation (PSI[/]{}NMSI).
Parallel and non - monotonic snapshot isolation are scalable variations of SI. These models retain two core properties of SI, namely
each transaction note a coherent snapshot, and
no | 1}$.
#### Caksal Consistency (CC).
Causau consistency (CE) holds when fransactkons observe consistent snapdhits od the system, and the cuient apppication proresses are monotonic. CC lf a scak cmisistency model and it doas not allow smlxiug consensus. It is in fact the strondest mocep that is avaijablt ugder iavtition [@syn:rep:1738]. Historically, CC rsfers tm the consistrncy of single operations ln a shared memory [@syn:lat:1025]. Causalli cogwistent tranractions (Causal HAT) is a consistency model that extenas CC to transaxtuonwn data storxs [@db:sjn:1751].
#### Snapshot Isolation (SI).
SI ix a widely-used coisisrency model [@syn:bd:1759]. Thiv model is strong, fut allowv jore interleavingw if cotcursent eeaa tdaisadtions then SER. Furtgermore, SI us causal (i.e., SI $\subxeeva$ CC), whereas SER if got.
Under SI, a transaction observes a *sndpsgot* of the state of the data store at some plint priow in time. Strong snapshot isolation (SSI) requires dhis avapwhjg tl contain all the preceding transactions in rqzl tpme [@Daudjee:2006]. Two tvansactions may cokmlt onder SI as lovg as cgeg do not write the same ijem cobcurrent. FI afoids the anomalies listed un, but exhibpts rhe *write-skew* anomcly, illustraced in histpry $h_4$ below.
In this histury, zn applicatlon using aata items $x$ and $y$ eivhes to maintain the invaryant $x \gew y$. Che invafianj holds initially, and cdch of the two trajsactnons ${\wolor{blue}{T_1}}$ and ${\color{red}{T_2}}$ guarantees the iitariant indivodgalny. As iljustrsted in histowy $h_4$, running tkem concbrrentuy under SP may vionate the indariant.
An appnlcation is *rmbust* agwinsr a xonsistdvcy model $M$ whrn, it probbces seriqlisable histories [@Cetohe:2016], despite runnnug atop a data stpre prjvpdiig $M$, Yd is known [@sfn:1687] tfat sn apolication iw robist against SI when averg invariant is matrrlalised bi a data ytem.
#### Parallel [/]{} Non-Monotonic Snakshot Msolatmon (PSO[/]{}NMFI).
Parallel and non-monotonic snzpshot isllabion are scalwble variations pf SI. These models retain two core propxrties of SI, namely
each transaction observgs s consistenv snapfhot, and
nm | 1}$. #### Causal Consistency (CC). Causal consistency when observe consistent of the system, are CC is a consistency model and does not allow solving consensus. It in fact the strongest model that is available under partition [@syn:rep:1738]. Historically, CC to the consistency of single operations on a shared memory [@syn:mat:1025]. Causally consistent (Causal is consistency that extends CC to transactional data stores [@db:syn:1751]. #### Snapshot Isolation (SI). SI is a widely-used model [@syn:bd:1759]. This model is strong, but allows interleavings of concurrent read than SER. Furthermore, SI is (i.e., $\subseteq$ CC), SER not. SI, a transaction a *snapshot* of the state of the data store at some point prior in time. Strong snapshot (SSI) requires to contain the transactions real time [@Daudjee:2006]. may commit under SI as long not write the same item concurrent. SI avoids anomalies listed but exhibits the *write-skew* anomaly, illustrated history $h_4$ below. In this history, an application data items $x$ and $y$ wishes to maintain the invariant $x \geq y$. The invariant and each of the transactions ${\color{blue}{T_1}}$ and guarantees invariant As in history running them concurrently under SI may violate the invariant. An application *robust* against a consistency model $M$ when, it produces serialisable despite atop a data providing $M$, It is [@syn:1687] an application is robust when invariant a item. Parallel [/]{} Non-Monotonic Snapshot (PSI[/]{}NMSI). Parallel and non-monotonic snapshot are scalable variations of core properties of SI, namely each transaction observes consistent snapshot, and no | 1}$.
#### Causal Consistency (CC).
Causal ConsistencY (CC) hoLds WheN tRansActiOns observe consIStenT snapshots of the system, aNd the ClIEnt aPPlIcatiOn proceSSeS ARe mOnOtOniC. Cc Is A weak ConSistencY model and iT doEs Not allow solvINg Consensus. IT is In fact the strOngEst modEl ThaT Is avaIlaBle unDer parTItion [@sYn:rep:1738]. HistOrICally, Cc Refers tO THe ConsIstency of single opERaTIons on a shared mEmory [@sYn:MAt:1025]. cAUsaLly Consistent TrAnsacTIons (CauSAl hat) Is a COnsistency modEl that extenDS CC To tranSaCtiONal datA storEs [@DB:syN:1751].
#### Snapshot IsOlatIon (SI).
SI is A widelY-Used conSIstency Model [@sYn:bD:1759]. ThIs moDEl Is StrOnG, But ALlOws MOre InterleaViNgS of coNcurRENT Read TraNsacTions Than SER. FurtheRmoRe, SI IS caUsal (i.E., SI $\suBsetEq$ cC), wheReas SEr is noT.
UNder SI, a transactIon oBserves a *sNapShOt* oF tHe staTE of the DatA stOre at soMe point PRioR iN TIMe. strong snapshot isolAtION (SsI) requirEs this SNaPsHOt to contAiN alL the PREcediNg trANsActions iN real tIMe [@daUdjee:2006]. TwO tRansacTiOns May CommiT UndeR SI as lOng as theY do noT Write the same itEM concurrent. SI AVoIDS tHE anoMalIes listed in, But eXHibiTs thE *WrIte-SKew* anOmaly, IlLUsTRated in history $h_4$ beloW.
IN this hIstorY, an applicatioN using data ITEMs $x$ and $y$ wIsheS To MAintain the invaRiant $X \geq y$. The inVAriant hoLds inItially, aNd each of tHE Two transActIonS ${\coLor{BLUe}{t_1}}$ and ${\color{red}{T_2}}$ GUAranTeEs the inVarIant indIviDuaLly. as iLlUstrated iN history $H_4$, rUnNiNg TheM concURrently uNdEr Si mAy vIolatE The invArianT.
An aPpLiCAtiOn is *robUSt* AGAinsT a CoNsisTenCy Model $m$ wheN, It pRoduces SerialisaBle HIstoRiEs [@cerone:2016], dEspite running AtOp a data stoRe ProViding $m$, iT is known [@Syn:1687] that an application is rOBust agaInsT SI whEn evEry invariAnt Is mateRiaLIsed by A data iTem.
#### PaRaLleL [/]{} nOn-MonOTOnIc SNaPshot IsolaTIOn (PsI[/]{}NMSi).
PAralLel and nOn-monotonic snapshoT IsoLation are scalAblE varIATiOns OF Si. theSe MOdeLS Retain two core prOperties of sI, NAmEly
each traNSacTiOn obserVes a conSisteNT snapshOt, and
no | 1}$.
#### Causal Consiste ncy (CC).
Caus alcon si sten cy ( CC) holds when tran sactions observe consi stent s n apsh o ts of t he syst e m, a ndth ecli en t a pplic ati on proc esses aremon ot onic. CC isa w eak consis ten cy model and it doesno t a l low s olv ing c onsens u s. Itis in fac tt he str o ngest m o d el tha t is available un d er partition [@sy n:rep: 17 3 8] . His tor ically, CC r efers to thec on s i s ten c y of single o perations o n ashared m emo r y [@sy n:mat :1 0 25] . Causallycons istent tr ansact i ons (Ca u sal HAT ) is a co nsi sten c ymo del t h ate xt end s CC to tran sa ct ional dat a s t ores [@ db:s yn:17 51].
#### Sn aps hotI sol ation (SI) .
S Iis awidely -used c onsistency mode l [@ syn:bd:17 59] .Thi smodel is str ong , b ut allo ws more int er l e a vi ngs of concurrentre a d t ransacti ons th a nSE R . Furthe rm ore , SI i s cau sal( i. e., SI $ \subse t eq $CC), wh er eas SE Risnot .
Un d er S I, a t ransacti on ob s erves a *snaps h ot* of the st a te o ft he d ata store at s omep oint pri o rint ime.Stron gs na p shot isolation (SSI )requir es th is snapshot t o containa l l the pre cedi n gt ransactions in real time [@Da u djee:200 6]. T wo trans actions m a y commitund erSIasl o ng as they do n o t wri te the sa meitem co ncu rre nt. SI a voids the anomali es l is te d i n, bu t exhibit sthe * wri te-sk e w* ano maly, ill us tr a ted in his t or y $h_4 $be low.
I nthishist o ry, an app licationusi n g da ta i tems $x $ and $y$ wis he s to maint ai n t he inv a r iant $x\geq y$. The invarianth olds in iti ally, and each ofthe two t ran s action s ${\c olor{ bl ue} { T _1}}$ a nd ${ \c olor{red}{ T _ 2}} $ gua ra ntee s the i nvariant individua l ly. As illustrat edin h i s to ry$ h_ 4 $,ru n nin g them concurrent ly under S Im ay violate t h e i nv ariant.
An ap plica t ion is*robust*against a c onsi s t enc y model $M $ when,it produc e s ser i al isabl e h istori es [@ Ceron e:2016 ] , d espit e runn in g atop a da ta store p roviding $M$, It is kno wn [@s yn:16 87] that anapp l ica tion is r obus t againstSIwhe n eve ryi nvari anti smat e riali sedb y a datai te m.# ## # Parallel[ / ] {}Non-M ono t onic S naps hot Isolation (PS I [/]{}NMSI).
P aral l e l a ndn on-m on otonic snapsho t i so l a tion are s calable var iationsof SI. T hese m odelsretaint w oc ore pr oper tie s of SI,nam el y
eachtr an s action obs er ves aconsis t ents n apshot, and
no | 1}$.
#### Causal_Consistency (CC).
Causal_consistency (CC) holds when_transactions observe_consistent_snapshots of_the_system, and the_client application processes_are monotonic. CC is_a weak consistency_model_and it does not allow solving consensus. It is in fact the strongest model_that_is available_under_partition_[@syn:rep:1738]. Historically, CC refers to_the consistency of single operations_on a_shared memory [@syn:mat:1025]. Causally consistent transactions (Causal HAT)_is_a consistency model_that extends CC to transactional data stores [@db:syn:1751].
#### Snapshot_Isolation (SI).
SI is a widely-used consistency_model [@syn:bd:1759]. This_model_is_strong, but allows more_interleavings of concurrent read transactions than_SER. Furthermore, SI is causal (i.e.,_SI $\subseteq$ CC), whereas SER is not.
Under_SI, a transaction observes a *snapshot*_of the state of the_data store_at some point prior in_time. Strong snapshot_isolation (SSI)_requires this snapshot_to contain all the preceding transactions_in real time_[@Daudjee:2006]. Two transactions may commit under_SI_as long as_they_do_not write_the same item_concurrent._SI avoids_the_anomalies listed in, but exhibits the_*write-skew*_anomaly, illustrated in history $h_4$ below.
In this_history, an application using_data_items $x$ and $y$_wishes to maintain the invariant_$x \geq y$. The invariant holds_initially, and_each of_the two transactions ${\color{blue}{T_1}}$ and ${\color{red}{T_2}}$ guarantees the invariant individually. As_illustrated in history $h_4$, running them_concurrently under SI may_violate the_invariant.
An_application is *robust*_against_a consistency_model $M$ when, it produces serialisable histories_[@Cerone:2016], despite_running atop a data store providing_$M$, It is known_[@syn:1687]_that an application is robust against_SI when every invariant is materialised_by a data item.
#### Parallel_[/]{}_Non-Monotonic_Snapshot Isolation (PSI[/]{}NMSI).
Parallel and non-monotonic_snapshot isolation are scalable variations of_SI. These models_retain two core properties of SI, namely
each_transaction_observes a consistent snapshot, and
no |
}$ generate the cumulants of $\xi(t)$ (see e.g. [@JacobsStochasticBook]). Fourier transforming these cumulants yields the polyspectra, which completely characterizes the noise in the frequency domain [@Rao1984book]. In the quantum case, our noise is a Heisenberg picture operator $\hat{\xi}(t)$ whose evolution is generated by the Hamiltonian of some bath; we take $\hat{\xi}(t)$ to be Hermitian for simplicity. Defining correlation functions now has some subtlety, as $\hat{\xi}(t)$ will not in general commute with itself at different times; hence, different time-ordering choices yield different results. Correlation functions at a given order describe both how the bath responds to external perturbations, as well as its intrinsic fluctuations [@Kamenev2011book]. We are interested here in characterizing the latter quantity, and asking whether these fluctuations are equivalent to an effective classical noise process.
The well-developed machinery of Keldysh quantum field theory provides a precise method for accomplishing our task [@Levitov1996; @Nazarov2003; @BelzigPRL2010; @Clerk2011; @Hofer2017]. While this approach is completely general, the simplest derivation is to imagine coupling an ancilla qubit to $\hat{\xi}$, such that the only qubit dynamics is from the interaction picture Hamiltonian $\hat H_{\mathrm{int}} (t) = \frac{1}{2} F(t) \hat{\xi}(t) \hat \sigma_z$. We then use the dephasing of the qubit to [*define*]{} the MGF of the noise in the quantum case, exactly like we would if the noise were classical: $$\begin{aligned}
\label{Eq:MGFqu}
& \Lambda[F( t );t_f] \equiv
\langle \hat{\sigma}_{-}(t) \rangle / \langle \hat{\sigma}_{-}(0) \rangle,
\\
\label{Eq:MGFK}
= &
\textrm{Tr} \left[
\mathcal{T} e^{- \frac{i}{2} \int_0^{t_f} dt' F(t') \hat{\xi}(t')}
\hat{\rho}_{\rm B}
\tilde{\mathcal{T}} e^{-\frac{i}{2} \int_ | } $ generate the cumulants of $ \xi(t)$ (see e.g. [ @JacobsStochasticBook ]). Fourier transforming these cumulants render the polyspectra, which wholly characterizes the noise in the frequency knowledge domain [ @Rao1984book ]. In the quantum lawsuit, our noise is a Heisenberg picture hustler $ \hat{\xi}(t)$ whose evolution is generated by the Hamiltonian of some bath; we carry $ \hat{\xi}(t)$ to be Hermitian for simplicity. Defining correlation coefficient functions now take some subtlety, as $ \hat{\xi}(t)$ will not in general commute with itself at different times; hence, different clock time - ordering choices yield unlike results. Correlation function at a given ordering describe both how the bath responds to external perturbations, as well as its intrinsic fluctuation [ @Kamenev2011book ]. We are interested here in characterizing the latter quantity, and asking whether these fluctuations are equivalent to an effective classical noise process.
The well - developed machinery of Keldysh quantum field theory provides a precise method for accomplishing our task [ @Levitov1996; @Nazarov2003; @BelzigPRL2010; @Clerk2011; @Hofer2017 ]. While this approach is wholly cosmopolitan, the simplest derivation is to imagine pair an ancilla qubit to $ \hat{\xi}$, such that the only qubit dynamics is from the interaction picture Hamiltonian $ \hat H_{\mathrm{int } } (t) = \frac{1}{2 } F(t) \hat{\xi}(t) \hat \sigma_z$. We then practice the dephasing of the qubit to [ * define * ] { } the MGF of the noise in the quantum shell, exactly like we would if the noise were authoritative: $ $ \begin{aligned }
\label{Eq: MGFqu }
& \Lambda[F (t); t_f ] \equiv
\langle \hat{\sigma}_{-}(t) \rangle / \langle \hat{\sigma}_{-}(0) \rangle,
\\
\label{Eq: MGFK }
= &
\textrm{Tr } \left [
\mathcal{T } e^{- \frac{i}{2 } \int_0^{t_f } dt' F(t') \hat{\xi}(t') }
\hat{\rho}_{\rm B }
\tilde{\mathcal{T } } e^{-\frac{i}{2 } \int _ | }$ gejerate the cumulants of $\wi(t)$ (see e.g. [@JacobsStochasvicBook]). Fourier transforming these cumulantd tieldw the polyspectra, whicf completvly charaxtermzes the noise ii the frczuendn domcii [@Rao1984book]. In thg quantum cave, our noise iv x Keisenberg picture operator $\hat{\xi}(t)$ wrose evplktion is generwted fy tgv Mamiltonian of some bath; we take $\hat{\xi}(t)$ to be Hermitoan for simplicity. Defininh cogrelation functiond now has sime fybtlety, as $\hxt{\xi}(t)$ will not in genetal commute with itself at diffefent cimes; hence, duffftent time-ordxring bhoices yield differett resukts. Correlatiok funrtiobs at a given order dxscribe both how the bath res[ouds to external pertuebqtionv, as welu as ita mntdinsic flnctuations [@Iamenev2011book]. We are interested newv in charactedizing tre latter quantity, and asking whether tvess fluctuations are equicalent to an effectivg classicaj noise process.
The well-developed machinery of Kelgysh sjanuun fieue hheory provides a precise method for accomplifgimg our task [@Levibov1996; @Nazarov2003; @BelzigLRP2010; @Vjerk2011; @Hofer2017]. Whkle thna zpproach is complehely gegeral, the simpjest derivation is to imagine ciupling an auciola qubit to $\hat{\xi}$, such that che onky qunit dynamics is from thz intedaction pichure Hamimgonian $\hat H_{\mathfm{ikt}} (d) = \frac{1}{2} F(t) \hat{\xi}(t) \hat \sigia_z$. We thxn usz the deohasong of the qubit to [*dcxine*]{} the MGF of thf noide in the quwntum case, exactly like we woulv if the noisg ware classiccl: $$\begln{aligned}
\label{Ez:MGFqu}
& \Lamyda[F( t );t_y] \equix
\langle \hat{\sigma}_{-}(t) \wangle / \langla \hat{\sigma}_{-}(0) \raigle,
\\
\label{Eq:MGFL}
= &
\rextrm{Tr} \left[
\mathcal{T} e^{- \fxcc{u}{2} \int_0^{t_f} dt' F(t') \hst{\xk}(t')}
\hat{\rho}_{\rm B}
\tilde{\mathccl{G}} e^{-\ftac{i}{2} \int_ | }$ generate the cumulants of $\xi(t)$ (see Fourier these cumulants the polyspectra, which the domain [@Rao1984book]. In quantum case, our is a Heisenberg picture operator $\hat{\xi}(t)$ evolution is generated by the Hamiltonian of some bath; we take $\hat{\xi}(t)$ to Hermitian for simplicity. Defining correlation functions now has some subtlety, as $\hat{\xi}(t)$ will in commute itself different times; hence, different time-ordering choices yield different results. Correlation functions at a given order describe how the bath responds to external perturbations, as as its intrinsic fluctuations We are interested here in the quantity, and whether fluctuations equivalent to an classical noise process. The well-developed machinery of Keldysh quantum field theory provides a precise method for accomplishing task [@Levitov1996; @Clerk2011; @Hofer2017]. this is general, the simplest to imagine coupling an ancilla qubit that the only qubit dynamics is from the picture Hamiltonian H_{\mathrm{int}} (t) = \frac{1}{2} F(t) \hat{\xi}(t) \sigma_z$. We then use the dephasing of the to [*define*]{} the MGF of the noise in the quantum case, exactly like we would noise were classical: $$\begin{aligned} & \Lambda[F( t \equiv \hat{\sigma}_{-}(t) / \hat{\sigma}_{-}(0) \rangle, \label{Eq:MGFK} = & \textrm{Tr} \left[ \mathcal{T} e^{- \frac{i}{2} \int_0^{t_f} dt' F(t') \hat{\rho}_{\rm B} \tilde{\mathcal{T}} e^{-\frac{i}{2} \int_ | }$ generate the cumulants of $\xi(t)$ (See e.g. [@JacobSStocHasTicboOk]). FoUrieR transforming tHEse cUmulants yields the polysPectrA, wHIch cOMpLetelY characTErIZEs tHe NoIse In THe FrequEncY domain [@rao1984book]. In tHe qUaNtum case, our nOIsE is a HeisenBerG picture operAtoR $\hat{\xi}(T)$ wHosE EvoluTioN is geNerateD By the HAmiltoniaN oF Some baTH; we take $\HAT{\xI}(t)$ to Be Hermitian for simPLiCIty. Defining corRelatiOn FUnCTIonS noW has some suBtLety, aS $\Hat{\xi}(t)$ wILl NOT In gENeral commute wIth itself at DIffErent tImEs; hENce, difFerenT tIMe-oRdering choiCes yIeld diffeRent reSUlts. CorRElation FunctiOns At a GiveN OrDeR deScRIbe BOtH hoW The Bath respOnDs To extErnaL PERTurbAtiOns, aS well As its intrinsiC flUctuATioNs [@KamEnev2011bOok]. WE aRe intEresteD here In Characterizing tHe laTter quantIty, AnD asKiNg wheTHer theSe fLucTuationS are equIValEnT TO An Effective classical NoISE pRocess.
ThE well-dEVeLoPEd machinErY of keldYSH quanTum fIElD theory pRovideS A pReCise metHoD for acCoMplIshIng ouR Task [@levitoV1996; @Nazarov2003; @belziGpRL2010; @Clerk2011; @Hofer2017]. WHIle this approaCH iS COmPLeteLy gEneral, the siMpleST derIvatIOn Is tO ImagiNe couPlINg AN ancilla qubit to $\hat{\xI}$, sUch thaT the oNly qubit dynamIcs is from tHE INteractiOn piCTuRE Hamiltonian $\haT H_{\matHrm{int}} (t) = \fraC{1}{2} f(t) \hat{\xi}(t) \Hat \siGma_z$. We thEn use the dEPHasing of The QubIt tO [*deFINe*]{} The MGF of the noISE in tHe Quantum CasE, exactlY liKe wE woUld If The noise wEre classIcAl: $$\BeGiN{alIgned}
\LAbel{Eq:MGfqU}
& \LaMbDa[F( T );t_f] \eqUIv
\langLe \hat{\SigmA}_{-}(t) \RaNGle / \Langle \hAT{\sIGMa}_{-}(0) \raNgLe,
\\
\LabeL{Eq:mGfK}
= &
\texTrm{TR} \LefT[
\mathcaL{T} e^{- \frac{i}{2} \iNt_0^{t_F} Dt' F(t') \HaT{\xI}(t')}
\hat{\rhO}_{\rm B}
\tilde{\mathCaL{T}} e^{-\frac{i}{2} \inT_ | }$ generate the cumulantsof $\xi(t) $ (se e e .g. [ @Jac obsS tochasticBook] ) . Fo urier transforming the se cu mu l ants yi eldsthe pol y sp e c tra ,wh ich c o mp letel y c haracte rizes thenoi se in the freq u en cy domain[@R ao1984book]. In the q ua ntu m case , o ur no ise is a Heis enberg pi ct u re ope r ator $\ h a t{ \xi} (t)$ whose evolut i on is generated b y theHa m il t o nia n o f some bat h; we t a ke $\ha t {\ x i } (t) $ to be Hermit ian for sim p lic ity. D ef ini n g corr elati on fun ctions nowhassome subt lety,a s $\hat { \xi}(t) $ will no t i n ge n er al co mm u tew it h i t sel f at dif fe re nt ti mes; h e n ce,dif fere nt ti me-ordering c hoi cesy iel d dif feren t re su lts.Correl ation f unctions at a g iven order de scr ib e b ot h how the ba thres ponds t o exter n alpe r t u rb ations, as well as i t s i ntrinsic fluct u at io n s [@Kame ne v20 11bo o k ]. We are in terested herei nch aracter iz ing th elat ter quan t ity, and a sking wh ether these fluctuat i ons are equiv a le n t t o aneff ective clas sica l noi se p r oc ess .
The well -d e ve l oped machinery of K el dysh q uantu m field theor y provides a precisemeth o df or accomplishi ng ou r task [@L e vitov199 6; @N azarov20 03; @Belz i g PRL2010; @C ler k20 11; @ Ho fer2017]. Whi l e thi sapproac h i s compl ete lygen era l, the simp lest der iv at io nisto im a gine cou pl ing a n a ncill a qubit to $ \hat {\ xi } $,such th a tt h e on ly q ubit dy na micsis f r omthe int eractionpic t ureHa mi ltonian $\hat H_{\ma th rm{int}} ( t) =\frac{ 1 } {2} F(t) \hat{\xi}(t) \hat \si g ma_z$.Wethenusethe depha sin g of t heq ubit t o [*de fine* ]{ } t h e MGFo f t heno ise in the q uan tum c as e, e xactlylike we would if t h e n oise were cla ssi cal: $ $\ beg i n{ a lig ne d }
\ l a bel{Eq:MGFqu}
& \Lamb da [ F( t );t_f]\ equ iv
\la ngle\ hat{\si gma}_{-}( t) \rangl e/ \l a n gle \hat{\sig ma}_{-}( 0) \rangl e ,
\\
\l abel{E q :MG FK}
= &
\ textrm{T r} \left[
\ mathca l{T}e^{ - \frac{i }{2 } \i nt_0^{t_f } dt ' F(t') \h at{ \xi }(t') }
\h at{ \ rho}_ {\rm B}
\ t il de{\mathcal { T } } e ^{-\f rac { i}{2}\int _ | }$ generate_the cumulants_of $\xi(t)$ (see e.g. [@JacobsStochasticBook])._Fourier transforming_these_cumulants yields_the_polyspectra, which completely_characterizes the noise_in the frequency domain_[@Rao1984book]. In the_quantum_case, our noise is a Heisenberg picture operator $\hat{\xi}(t)$ whose evolution is generated by_the_Hamiltonian of_some_bath;_we take $\hat{\xi}(t)$ to be_Hermitian for simplicity. Defining correlation_functions now_has some subtlety, as $\hat{\xi}(t)$ will not in_general_commute with itself_at different times; hence, different time-ordering choices yield different_results. Correlation functions at a given_order describe both_how_the_bath responds to external_perturbations, as well as its intrinsic_fluctuations [@Kamenev2011book]. We are interested here_in characterizing the latter quantity, and asking_whether these fluctuations are equivalent to_an effective classical noise process.
The_well-developed machinery_of Keldysh quantum field theory_provides a precise_method for_accomplishing our task_[@Levitov1996; @Nazarov2003; @BelzigPRL2010; @Clerk2011; @Hofer2017]. While_this approach is_completely general, the simplest derivation is_to_imagine coupling an_ancilla_qubit_to $\hat{\xi}$,_such that the_only_qubit dynamics_is_from the interaction picture Hamiltonian $\hat_H_{\mathrm{int}}_(t) = \frac{1}{2} F(t) \hat{\xi}(t) \hat_\sigma_z$. We then use_the_dephasing of the qubit_to [*define*]{} the MGF of_the noise in the quantum case,_exactly like_we would_if the noise were classical: $$\begin{aligned}
\label{Eq:MGFqu}
& \Lambda[F(_t );t_f] \equiv
_ _ _\langle_\hat{\sigma}_{-}(t) \rangle /_\langle_\hat{\sigma}_{-}(0) \rangle,
_ _ \\
_ _ __ \label{Eq:MGFK}
_ _ = &
___ \textrm{Tr} \left[
_ _ _ \mathcal{T} e^{- \frac{i}{2} \int_0^{t_f} dt'_F(t')_\hat{\xi}(t')}
__ \hat{\rho}_{\rm_B}
__ _ \tilde{\mathcal{T}}_e^{-\frac{i}{2} \int_ |
of a single element of $B$ is deferred to a later time, usually with a different procedure (according to additional information or “subjective randomization", e.g., flipping a coin). Notice that the rejection map associated to a choice may fail to be a choice, since the rejection set of some menu can be empty.
The next definition recalls the classical notion of a rationalizable choice.
\[DEF:rationalizable choice\] A choice $c \colon \Omega \rightrightarrows U$ is *rationalizable* (or *binary*) if there exists a binary relation $\precsim$ on $U$ such that the equality[^5] $c(B) = \max_{\precsim} B$ holds for all menus $B \in \Omega$.
The revealed preference theory approach postulates that preferences can be derived from choices. The preference revealed by a primitive choice is modeled by a suitable binary relation on the set of alternatives. Then a choice is rationalizable whenever the observed behavior can be fully explained (i.e., retrieved) by constructing a binary relation of revealed preference.
The rationalizability of choice is traditionally connected to the satisfaction of suitable *axioms of choice consistency*. These axioms codify rules of coherent behavior of an economic agent. Among the several axioms that are considered in the literature, the following are relevant to our analysis (a universal quantification on all the involved menus is implicit):
--------------------------------------------------------------------- --------------------------------------------------------------------------------------------------------------------- --
**axiom $(\alpha)$** \[*standard contraction consistency*\]: $A \subseteq B \;\; \Longrightarrow \;\; A \cap c(B) \subseteq c(A)$
\[.1cm\] **axiom $(\gamma)$** \[*standard expansion consistency*\]: $c(A) \cap c(B) \subseteq c(A \cup B)$
\[.1cm\] **axiom $(\beta)$** \[*symmetric expansion consistency*\]: $\big(A \subseteq B \: \wedge \: c(A) \cap c(B) \neq \emptyset \big) \;\; \Longrightarrow \;\; c(A) \subseteq c(B)$
\[.1cm\] **axiom $(\rho)$** \[*standard replacement consistency*\]: $c(A) \setminus c(A \cup B) \neq \emptyset \;\; \Longrightarrow \;\; B \cap | of a single element of $ B$ is deferred to a later meter, normally with a different procedure (harmonize to extra information or “ subjective randomization ", e.g., flip a coin). Notice that the rejection map associate to a choice may fail to be a choice, since the rejection hardening of some menu can be empty.
The next definition recalls the authoritative notion of a rationalizable choice.
\[DEF: rationalizable choice\ ] A option $ c \colon \Omega \rightrightarrows U$ is * rationalizable * (or * binary *) if there exist a binary sexual intercourse $ \precsim$ on $ U$ such that the equality[^5 ] $ c(B) = \max_{\precsim } B$ holds for all menus $ B \in \Omega$.
The reveal preference theory approach contend that preferences can be derived from choices. The preference revealed by a crude choice is modeled by a suitable binary relation on the set of alternatives. Then a choice is rationalizable whenever the observed behavior can be fully explained (i.e., retrieved) by constructing a binary relation of revealed preference.
The rationalizability of choice is traditionally get in touch to the satisfaction of suitable * axiom of option consistency *. These axioms codify rule of coherent behavior of an economic agent. Among the respective axioms that are regard in the literature, the following are relevant to our analysis (a cosmopolitan quantification on all the involved menus is implicit ):
--------------------------------------------------------------------- --------------------------------------------------------------------------------------------------------------------- --
* * axiom $ (\alpha)$ * * \[*standard contraction consistency*\ ]: $ A \subseteq B \;\; \Longrightarrow \;\; A \cap c(B) \subseteq c(A)$
\[.1cm\ ] * * axiom $ (\gamma)$ * * \[*standard expansion consistency*\ ]: $ c(A) \cap c(B) \subseteq c(A \cup B)$
\[.1cm\ ] * * axiom $ (\beta)$ * * \[*symmetric expansion consistency*\ ]: $ \big(A \subseteq B \: \wedge \: c(A) \cap c(B) \neq \emptyset \big) \;\; \Longrightarrow \;\; c(A) \subseteq c(B)$
\[.1cm\ ] * * axiom $ (\rho)$ * * \[*standard replacement consistency*\ ]: $ c(A) \setminus c(A \cup B) \neq \emptyset \;\; \Longrightarrow \;\; B \cap | of a single element of $B$ ir deferred to a later vime, ushally wigh a different procedure (acclreing uj additional inforoation or “subjectuve candomization", e.g., flippinn a ckln). Nocire that the rejgction map avsociated to a cfonce may fail to be a choice, since thq rejecyiln set of some menl san gv tmpty.
The next definition recalls fhe clavsical notion of a rationalizable choicf.
\[DEF:gationalizable cholce\] A choicg $c \silon \Omega \rkghtrightagxows U$ is *rztionalizable* (or *binary*) if therd exixts a binatv repdtion $\precsmm$ on $L$ such that tmv equalhty[^5] $c(B) = \max_{\precsim} B$ molds foe all menus $B \in \Omege$.
The revealed prefergnce theorf cpproach postulates tyar prexeretces xan be dxribed frlm rhoices. The preference revealed by a primotynr choice is jodeleq fy a suitable binary relation on the seu of zlternatives. Then a chouce is rationalizable whenever the observed behavior can be fully explained (i.e., setrixvdd) yn zindtructing a binary relation of revealed prefewsnve.
Nhe rationalizabijity of choocf of traditionaluy conuscfed to the satisfaftion os suirable *axijms pf choice consistency*. These axioms codiyy eules of coherent yehavior of cn ecomomic agent. Among the severau axjoms that age considsfed in the literxtuge, tve following are relevant eo our anelysix (a unixerssl quagtificatioj on all the involved mejus id hmplicit):
--------------------------------------------------------------------- --------------------------------------------------------------------------------------------------------------------- --
**axiom $(\alpha)$** \[*standard contrarvion consistemcf*\]: $C \subscteq B \;\; \Longrigrtarrow \;\; A \cap c(B) \subfeteq c(A)$
\[.1cm\] **axioi $(\ganma)$** \[*wtandara expansion conxistency*\]: $c(A) \cap c(V) \subseteq c(A \cup N)$
\[.1cm\] **axlom $(\bets)$** \[*symmetric expansiot cohsistency*\]: $\big(A \sibfeteq B \: \wedge \: s(A) \cap c(B) \neq \emptyset \big) \;\; \Lonhrighvarrow \;\; c(A) \xubfeteq c(B)$
\[.1cm\] **axiom $(\rho)$** \[*stansard replwceient consistqncy*\]: $c(A) \setminbs c(A \cup B) \neq \emptyset \;\; \Longrightarrow \;\; B \cap | of a single element of $B$ is a time, usually a different procedure “subjective e.g., flipping a Notice that the map associated to a choice may to be a choice, since the rejection set of some menu can be The next definition recalls the classical notion of a rationalizable choice. \[DEF:rationalizable choice\] choice \colon \rightrightarrows is *rationalizable* (or *binary*) if there exists a binary relation $\precsim$ on $U$ such that the $c(B) = \max_{\precsim} B$ holds for all menus \in \Omega$. The revealed theory approach postulates that preferences be from choices. preference by primitive choice is by a suitable binary relation on the set of alternatives. Then a choice is rationalizable whenever the behavior can explained (i.e., by a relation of revealed rationalizability of choice is traditionally connected of suitable *axioms of choice consistency*. These axioms rules of behavior of an economic agent. Among several axioms that are considered in the literature, following are relevant to our analysis (a universal quantification on all the involved menus is --------------------------------------------------------------------------------------------------------------------- -- **axiom $(\alpha)$** contraction consistency*\]: $A B \Longrightarrow A c(B) \subseteq \[.1cm\] **axiom $(\gamma)$** \[*standard expansion consistency*\]: $c(A) \cap c(B) \subseteq c(A B)$ \[.1cm\] **axiom $(\beta)$** \[*symmetric expansion consistency*\]: $\big(A \subseteq B \: \cap c(B) \neq \big) \;\; \Longrightarrow \;\; \subseteq \[.1cm\] **axiom $(\rho)$** \[*standard $c(A) c(A \emptyset \Longrightarrow B \cap | of a single element of $B$ is defeRred to a latEr timE, usUalLy With A difFerent procedurE (AccoRding to additional inforMatioN oR “SubjECtIve raNdomizaTIoN", E.G., flIpPiNg a CoIN). NOtice ThaT the rejEction map aSsoCiAted to a choicE MaY fail to be a ChoIce, since the rEjeCtion sEt Of sOMe menU caN be emPty.
The NExt defInition reCaLLs the cLAssical NOTiOn of A rationalizable chOIcE.
\[dEF:rationalizaBle choIcE\] a cHOIce $C \coLon \Omega \riGhTrighTArrows U$ IS *rATIOnaLIzable* (or *binarY*) if there exiSTs a Binary ReLatIOn $\precSim$ on $u$ sUCh tHat the equalIty[^5] $c(b) = \max_{\precsIm} B$ holDS for all MEnus $B \in \omega$.
THe rEveAled PReFeRenCe THeoRY aPprOAch PostulatEs ThAt preFereNCES Can bE deRiveD from Choices. The preFerEnce REveAled bY a priMitiVe ChoicE is modEled bY a Suitable binary rElatIon on the sEt oF aLteRnAtiveS. then a cHoiCe iS rationAlizablE WheNeVER ThE observed behavior cAn BE FuLly explaIned (i.e., REtRiEVed) by conStRucTing A BInary RelaTIoN of reveaLed preFErEnCe.
The raTiOnalizAbIliTy oF choiCE is tRaditiOnally coNnectED to the satisfacTIon of suitable *AXiOMS oF ChoiCe cOnsistency*. THese AXiomS codIFy RulES of coHerenT bEHaVIor of an economic agenT. AMong thE seveRal axioms that Are consideRED In the litEratURe, THe following are RelevAnt to our anALysis (a unIversAl quantiFication oN ALl the invOlvEd mEnuS is IMPlIcit):
--------------------------------------------------------------------- --------------------------------------------------------------------------------------------------------------------- --
**axiom $(\alphA)$** \[*STandArD contraCtiOn consiSteNcy*\]: $a \suBseTeQ B \;\; \LongrigHtarrow \;\; A \CaP c(b) \sUbSetEq c(A)$
\[.1cM\] **Axiom $(\gamMa)$** \[*StaNdArd ExpanSIon conSisteNcy*\]: $c(a) \cAp C(b) \suBseteq c(a \CuP b)$
\[.1Cm\] **axIoM $(\bEta)$** \[*sYmmEtRic exPansIOn cOnsisteNcy*\]: $\big(A \suBseTEq B \: \wEdGe \: C(A) \cap c(B) \Neq \emptyset \biG) \;\; \LOngrightarRoW \;\; c(A) \SubsetEQ C(B)$
\[.1cm\] **axioM $(\rho)$** \[*standard replacement COnsisteNcy*\]: $C(A) \setMinuS c(A \cup B) \neQ \emPtyset \;\; \lonGRightaRrow \;\; B \cAp | of a single element of $B $ is defer red t o a la te r ti me,usually with a diff erent procedure (accor dingto addi t io nal i nformat i on o r “ su bj ect iv e r andom iza tion",e.g., flip pin ga coin). Not i ce that therej ection map a sso ciated t o a choic e m ay fa il tob e a ch oice, sin ce the re j ections e tof s ome menu can be e m pt y .
The next de finiti on re c a lls th e classica lnotio n of a r a ti o n a liz a ble choice.
\[DEF:ratio n ali zablech oic e \] A c hoice $ c \c olon \Omega \ri ghtrighta rrowsU $ is *r a tionali zable* (o r * bina r y* )ifth e ree xi sts a b inary re la ti on $\ prec s i m $ on$U$ suc h tha t the equalit y[^ 5] $ c (B) = \m ax_{\ prec si m} B$ holds foral l menus $B \in\Ome ga$.
The re ve ale dprefe r ence t heo ryapproac h postu l ate st h a tpreferences can be d e r iv ed fromchoice s .Th e prefere nc e r evea l e d bya pr i mi tive cho ice is mo de led byasuitab le bi nar y rel a tion on th e set of alte r natives. Thena choice is ra t io n a li z able wh enever theobse r vedbeha v io r c a n befully e x pl a ined (i.e., retriev ed ) by c onstr ucting a bina ry relatio n o f reveal ed p r ef e rence.
The ra tiona lizability of choic e istraditio nally con n e cted tothe sa tis fac t i on of suitable* a xiom sof choi ceconsist enc y*. Th ese a xioms cod ify rule sof c oh ere nt be h avior of a n e co nom ic ag e nt. Am ong t he s ev er a l a xioms t h at a re c on si dere d i nthe l iter a tur e, thefollowing ar e rel ev an t to ou r analysis (a u niversal q ua nti ficati o n on allthe involved menus is i m plicit) :
--- ---- --------- --- ------ --- - ------ ------ ----- -- --- - - ----- - - -- --- -- -- ------- - - --- ----- -- ---- ------- ------------------ - --- ------------- --- ---- - - -- --- - -- - --- -- - --- - - --------------- -------- - - * *axiom $(\ a lph a) $** \[* standar d con t raction consiste ncy*\]: $A\subseteqB \;\; \ Longright a rrow\ ;\ ; A \ cap c(B)\s ubs eteqc(A)$
\[.1cm\]**axio m $(\ gam ma)$** \[ *st a nda rd expans ionconsistenc y*\ ]: $c( A)\ cap c (B)\ su bse t eq c( A \c u p B)$
\[. 1 c m\] ** a xiom $ (\beta)$** \[* sym me t r ic expan si on consiste ncy*\]: $ \big( A \sub seteqB \: \w e d ge \: c(A ) \c apc(B) \neq \e mp t yset \b ig )\ ;\; \L ongr ig htarro w \;\; c(A) \ subseteq c(B)$
\ [ . 1cm\] **a xiom$( \rho)$* * \[* standard r eplacementconsis tenc y*\]: $c(A )\setmi nus c (A \cup B) \neq \emp tyset \;\; \ Lo ngri ght arrow\;\; B \cap | of_a single_element of $B$ is_deferred to_a_later time,_usually_with a different_procedure (according to_additional information or “subjective_randomization", e.g., flipping_a_coin). Notice that the rejection map associated to a choice may fail to be_a_choice, since_the_rejection_set of some menu can_be empty.
The next definition recalls_the classical_notion of a rationalizable choice.
\[DEF:rationalizable choice\] A choice_$c_\colon \Omega \rightrightarrows_U$ is *rationalizable* (or *binary*) if there exists a_binary relation $\precsim$ on $U$ such_that the equality[^5]_$c(B)_=_\max_{\precsim} B$ holds for_all menus $B \in \Omega$.
The revealed_preference theory approach postulates that preferences_can be derived from choices. The preference_revealed by a primitive choice is_modeled by a suitable binary_relation on_the set of alternatives. Then_a choice is_rationalizable whenever_the observed behavior_can be fully explained (i.e., retrieved)_by constructing a_binary relation of revealed preference.
The rationalizability_of_choice is traditionally_connected_to_the satisfaction_of suitable *axioms_of_choice consistency*._These_axioms codify rules of coherent behavior_of_an economic agent. Among the several axioms_that are considered in_the_literature, the following are_relevant to our analysis (a_universal quantification on all the involved_menus is_implicit):
_--------------------------------------------------------------------- --------------------------------------------------------------------------------------------------------------------- --
**axiom $(\alpha)$** \[*standard contraction consistency*\]: _ _ $A \subseteq B_\;\; \Longrightarrow_\;\;_A \cap c(B)_\subseteq_c(A)$ _ _ _ _ __ _ _ ___ _
_\[.1cm\] **axiom $(\gamma)$**_\[*standard expansion consistency*\]: $c(A) \cap_c(B)_\subseteq c(A \cup B)$ __ ___ _ _ _ _ ___ _ _ _ _ _
\[.1cm\] **axiom $(\beta)$**_\[*symmetric expansion consistency*\]: $\big(A \subseteq B_\: \wedge \: c(A) \cap c(B) \neq_\emptyset \big) \;\; \Longrightarrow_\;\;_c(A) \subseteq c(B)$
_\[.1cm\]_**axiom_$(\rho)$**_\[*standard replacement consistency*\]:_ $c(A)_\setminus c(A \cup_B) \neq_\emptyset \;\; \Longrightarrow_\;\; B_\cap |
ATWT 2.1074 1.5489 1.4156 0.9983 0.9892
AWLP 2.0507 1.5223 1.3005 0.9985 0.9896
GLP 2.1739 1.5937 1.2016 0.9980 0.9879
P+XS 1.1031 0.7988 0.9034 0.9998 0.9946
NLV [**0.9395**]{} [**0.6794**]{} [**0.7543**]{} [**0.9999**]{} [**0.9954**]{}
NLVD 1.1546 0.8377 0.9651 0.9997 0.9934
: Quantitative evaluation of the fused products on simulated data from RGB aerial images at resolution of 30 cm per pixel. For these experiments, the low-resolution spectral components were co-registered and the panchro-spectral constraint fulfilled with $\alpha_B=\alpha_G=\alpha_R=\frac{1}{3}$. In this ideal setting, NLV provides the best numerical results although the proposed NLVD model is the closest to it. Only P+XS beats NLVD when the aliasing is not as apparent (Table \[table\_30cm\_s17\_RGB\_regist\_linear\]). Interestingly, we observe that PCA, Brovey, PRACS, HPF, SFIM, and NLVD seem to work almost independent of the amount of aliasing in the data, since the quality of their performances with respect to all metrics even increase as $\sigma$ decreases.[]{data-label="table_30cm_RGB_regist_linear"}
[0.99]{}
RMSE ERGAS SAM SSIM Q$2^n$
----------- ---------------- ---------------- ---------------- ---------------- ---------------- -- -- --
Reference 0 0 0 1 1
PCA 2.5703 1.8913 1.5554 0.9953 0.9745
Brovey 2.1917 1.6125 1.3189 0.9963 0.9766
BDSD 1.9834 1.4462 | ATWT 2.1074 1.5489 1.4156 0.9983 0.9892
AWLP 2.0507 1.5223 1.3005 0.9985 0.9896
GLP 2.1739 1.5937 1.2016 0.9980 0.9879
P+XS 1.1031 0.7988 0.9034 0.9998 0.9946
NLV [ * * 0.9395 * * ] { } [ * * 0.6794 * * ] { } [ * * 0.7543 * * ] { } [ * * 0.9999 * * ] { } [ * * 0.9954 * * ] { }
NLVD 1.1546 0.8377 0.9651 0.9997 0.9934
: Quantitative evaluation of the fused products on simulated data from RGB aeriform image at resolution of 30 centimeter per pixel. For these experiment, the low - resolution apparitional components were co - registered and the panchro - spectral restraint satisfy with $ \alpha_B=\alpha_G=\alpha_R=\frac{1}{3}$. In this ideal setting, NLV put up the best numerical solution although the nominate NLVD model is the near to it. Only P+XS beats NLVD when the aliasing is not equally apparent (Table \[table\_30cm\_s17\_RGB\_regist\_linear\ ]). Interestingly, we observe that PCA, Brovey, PRACS, HPF, SFIM, and NLVD seem to exercise almost independent of the sum of aliasing in the data, since the timbre of their performances with regard to all metrics even increase as $ \sigma$ decreases.[]{data - label="table_30cm_RGB_regist_linear " }
[ 0.99 ] { }
RMSE ERGAS SAM SSIM Q$2^n$
----------- ---------------- ---------------- ---------------- ---------------- ---------------- -- -- --
Reference 0 0 0 1 1
PCA 2.5703 1.8913 1.5554 0.9953 0.9745
Brovey 2.1917 1.6125 1.3189 0.9963 0.9766
BDSD 1.9834 1.4462 |
ATWT 2.1074 1.5489 1.4156 0.9983 0.9892
AWLP 2.0507 1.5223 1.3005 0.9985 0.9896
GLP 2.1739 1.5937 1.2016 0.9980 0.9879
P+XS 1.1031 0.7988 0.9034 0.9998 0.9946
NLV [**0.9395**]{} [**0.6794**]{} [**0.7543**]{} [**0.9999**]{} [**0.9954**]{}
NLVD 1.1546 0.8377 0.9651 0.9997 0.9934
: Quantitatpre evaluatikn of the fused products on simjlateb data from RTB wgrial images at rvsolution of 30 cm per [ixel. Fpr these experlmentv, tye low-resolution specvral components were co-registarzd and the panchro-spexteal cmnstsaing fuufimlxd sith $\apphe_B=\alpha_G=\alpga_R=\frac{1}{3}$. In rhis ideal setting, MLD provides the best guierical results although the proposed NNVD model is the closest ti it. Only P+XS beats NPVD when ehe aliasing is not as apparent (Table \[table\_30cm\_s17\_RGB\_segisv\_lknecv\]). Kbtfrestingly, we observe that PCA, Brovey, PRACS, H[R, XFPM, and NLVD seem bo work almost indrpfncgnt of the amoont of cmizsing in the data, dince tre quqlity of uheir performances with respect ro all metribs ecen increase as $\siyma$ decreaser.[]{dats-labek="table_30cm_RGB_regist_linear"}
[0.99]{}
RMSE ERGAS SAM SFIM Q$2^n$
----------- ---------------- ---------------- ---------------- ---------------- ---------------- -- -- --
Refcsence 0 0 0 1 1
PCA 2.5703 1.8913 1.5554 0.9953 0.9745
Brovey 2.1917 1.6125 1.3189 0.9963 0.9766
VDSD 1.9834 1.4462 | ATWT 2.1074 1.5489 1.4156 0.9983 0.9892 AWLP 1.3005 0.9896 GLP 1.5937 1.2016 0.9980 0.9998 NLV [**0.9395**]{} [**0.6794**]{} [**0.9999**]{} [**0.9954**]{} NLVD 0.8377 0.9651 0.9997 0.9934 : Quantitative of the fused products on simulated data from RGB aerial images at resolution 30 cm per pixel. For these experiments, the low-resolution spectral components were co-registered the constraint with In this ideal setting, NLV provides the best numerical results although the proposed NLVD model is closest to it. Only P+XS beats NLVD when aliasing is not as (Table \[table\_30cm\_s17\_RGB\_regist\_linear\]). Interestingly, we observe PCA, PRACS, HPF, and seem work almost independent the amount of aliasing in the data, since the quality of their performances with respect to all even increase decreases.[]{data-label="table_30cm_RGB_regist_linear"} [0.99]{} ERGAS SSIM ----------- ---------------- ---------------- ---------------- -- -- -- Reference 0 1 PCA 2.5703 1.8913 1.5554 0.9953 0.9745 Brovey 1.6125 1.3189 0.9766 BDSD 1.9834 1.4462 |
ATWT 2.1074 1.5489 1.4156 0.9983 0.9892
AWLP 2.0507 1.5223 1.3005 0.9985 0.9896
GLP 2.1739 1.5937 1.2016 0.9980 0.9879
P+XS 1.1031 0.7988 0.9034 0.9998 0.9946
NLV [**0.9395**]{} [**0.6794**]{} [**0.7543**]{} [**0.9999**]{} [**0.9954**]{}
NLVD 1.1546 0.8377 0.9651 0.9997 0.9934
: QuanTitative evAluatIon Of tHe FuseD proDucts on simulatED datA from RGB aerial images at ResolUtIOn of 30 CM pEr pixEl. For thESe EXPerImEnTs, tHe LOw-ResolUtiOn spectRal componeNts WeRe co-registerED aNd the panchRo-sPectral constRaiNt fulfIlLed WIth $\alPha_b=\alphA_G=\alphA_r=\frac{1}{3}$. IN this ideaL sETting, Nlv providES ThE besT numerical results ALtHOugh the proposeD NLVD mOdEL iS THe cLosEst to it. OnlY P+xS beaTS NLVD whEN tHE ALiaSIng is not as appArent (Table \[tABle\_30Cm\_s17\_RGB\_ReGisT\_Linear\]). interEsTIngLy, we observe That pCA, Brovey, pRACS, Hpf, SFIM, anD nLVD seeM to worK alMosT indEPeNdEnt Of THe aMOuNt oF AliAsing in tHe DaTa, sinCe thE QUALity Of tHeir PerfoRmances with reSpeCt to ALl mEtricS even IncrEaSe as $\sIgma$ deCreasEs.[]{Data-label="table_30cM_RGB_Regist_linEar"}
[0.99]{}
rMsE ErGaS SAM ssIM Q$2^n$
----------- ---------------- ---------------- ---------------- ---------------- ---------------- -- -- --
REfeRenCe 0 0 0 1 1
PCA 2.5703 1.8913 1.5554 0.9953 0.9745
BrOvey 2.1917 1.6125 1.3189 0.9963 0.9766
BDSd 1.9834 1.4462 |
ATWT 2.1074 1.54 89 1.4 156 0 .998 3 0.9892
A WLP 2. 0 5 07 1. 5223 1.3005 0.9985 0.9896
GLP 2.1 739 1 .5937 1 . 2016 0 .9 980 0.9879
P +XS 1.1 031 0 .7988 0 . 9 034 0.9 998 0 .9946 N L V [**0.93 95** ]{} [** 0.6794 * *]{} [ **0.754 3**]{} [** 0.99 9 9* *] {} [ **0 . 99 54* * ]{}
NLV D 1 .15 46 0.8377 0.9 651 0. 9997 0 .9 934
: Quantit ati ve ev al uatio n of th e f use d produ cts ons imu la t e d d ata from RGB aeria li m ag es at re soluti o nof 30 cm pe rpix el.F o r the se e x pe riments, the l o w- re solutio nspectr al co mpo nents were co-re gistered andt he panchro-spe c tral constrai n tf u lf i lled wi th $\alpha_ B=\a l pha_ G=\a l ph a_R = \frac {1}{3 }$ . I n this ideal setting ,NLV pr ovide s the best nu merical re s u l ts altho ught he proposed NLVDmodel is the cl o sest toit. O nly P+XS beats NL V D when th e a lia sin g i s no t as apparent ( Tabl e\[table \_3 0cm\_s1 7\_ RGB \_r egi st \_linear\ ]). Inte re st in gl y,we ob s erve tha tPCA ,Bro vey,P RACS,HPF,SFIM ,an d NL VD seem to w orkal mo st i nde pe ndent oft heamountof aliasi ngi n th eda ta, sin ce the qualit yof their p er for mances w ith resp ect to all metrics even increas e a s $\s igma $ decreas es. []{dat a-l a bel="t able_3 0cm_R GB _re g i st_li n e ar "}
[ 0.99]{}
RMSE ERGAS SAM SS I M Q $2^n$
----- -- - -- - -------- - --- -- -- ---- ------- ----- ------- --------- -------- -- ---- - - -- ---------- ---- ---- --
R efere n ce 0 0 0 1 1
PC A 2 .57 0 3 1.89 13 1.5 554 0.99 5 3 0.9 7 45 Brovey 2. 1917 1.61 25 1.31 8 9 0. 9963 0. 97 66
B DSD 1.9834 1 . 4462 |
_ _ ATWT _ __ __ 2.1074 _ _ _ 1.5489__ 1.4156 __ ___ 0.9983 _ _ _0.9892 __ _
AWLP _ _2.0507 ___ _ 1.5223 _ _ 1.3005 _ 0.9985_ _ _ 0.9896 _ _ _ _
_ GLP_ __ 2.1739___ _ __ _1.5937_ __ 1.2016 _ __ 0.9980 _ _ 0.9879 _ _ _
_P+XS _ _1.1031 __ __ _ 0.7988 _ _ 0.9034 _ __ 0.9998 _ _ 0.9946 ___ _
_ _NLV [**0.9395**]{}__ [**0.6794**]{} [**0.7543**]{} __[**0.9999**]{} [**0.9954**]{} ___
_ NLVD _ 1.1546 _ _ __0.8377_ _0.9651 _ _ 0.9997 _ _0.9934 _
: Quantitative evaluation of the_fused products on simulated data from RGB_aerial images at resolution_of_30 cm per pixel. For these experiments,_the_low-resolution_spectral_components were co-registered_and the panchro-spectral_constraint fulfilled with_$\alpha_B=\alpha_G=\alpha_R=\frac{1}{3}$. In_this ideal setting,_NLV provides_the best numerical_results_although the proposed NLVD model is the closest to it. Only P+XS beats NLVD when the aliasing is_not as apparent (Table \[table\_30cm\_s17\_RGB\_regist\_linear\]). Interestingly,_we observe that PCA, Brovey,_PRACS,_HPF, SFIM, and_NLVD seem to work_almost independent of the amount of aliasing in_the data, since the quality of their performances with_respect_to all_metrics even increase as_$\sigma$ decreases.[]{data-label="table_30cm_RGB_regist_linear"}
[0.99]{}
_ _ _ _ _ _ __RMSE ____ERGAS _ _ _ __SAM _ _ _ SSIM _ Q$2^n$
_----------- ----------------_---------------- ---------------- ---------------- ----------------_-- --_--
_ Reference _ _ 0 __ 0 ___ __ ___0__ _ __1 _ __ _ _1 __ __
__ _PCA 2.5703 __ _1.8913 __ _1.5554 _ 0.9953 _ _ 0.9745
_Brovey _ 2.1917_ __ _ 1.6125_ _ 1.3189 0.9963 _ 0.9766_ _ _
_ BDSD _ 1.9834 1.4462 |
c_s}{8 \pi \rho_0 G M_{\rm c} \chi},
\label{eq:v_MVR17}$$ where $\gamma$ is the adiabatic index of the gas, $L_{\rm c}$ and $M_{\rm c}$ are the luminosity and mass of the perturbing core and $\chi$ is the thermal conductivity of the gas that they assumed to be constant. They noted that the effects of the heating force are expected to be significant for planet formation application, via, e.g., the Earth developing a non-negligible eccentricity and inclination with respect to the protoplanetary disc for realistic disc parameters. These conclusions and the analytic result (Equation \[eq:v\_MVR17\]) were confirmed with numerical simulations by [@VelascoRomeroMasset19], [@ChrenkoLambrechts19] and [@GuileraEtal19].
In application to our particular problem, we note that the density in the central part of the protoplanet is initially homogeneous with a nearly constant $\rho_0 \approx 10^{-9}$ gcm$^{-3}$. Hence, the constant background density results of [@MassetVelascoRomero17] are applicable. However, our protoplanet is finite in extent. When the core is displaced by a distance $R$ from the centre, there is a returning force of protoplanet gravity, given by $$F_{\rm g} = - \frac{G M(R) M_{\rm c}}{R^2}\;,
\label{eq:Fg0}$$ where $M(R) \approx (4\pi/3) \rho_0 R^3$ is the enclosed mass within radius $R$.
The motion and radius of the orbiting core is then set by a balance of forces. In the azimuthal direction the heating and dynamical friction force specify the core velocity from Equation \[eq:v\_MVR17\]. Given this velocity ($V_0$), the radial position of the core ($R$) is simply set by the gravitational force of the enclosed protoplanet which provides the centrifugal force. We see that $V_0$ is independent of orbit radius $R$ whereas the gravitational force is propotional to $R$, as long as the enclosed gas density is roughly constant. The balance of these therefore establishes an equilibrium radius of the orbit for the core of mass $M_{\rm c}$ and luminosity $L | c_s}{8 \pi \rho_0 G M_{\rm c } \chi },
\label{eq: v_MVR17}$$ where $ \gamma$ is the adiabatic index of the gas, $ L_{\rm c}$ and $ M_{\rm c}$ are the luminosity and mass of the perturbing effect and $ \chi$ is the thermal conduction of the gas that they assumed to be changeless. They notice that the effects of the heating effect are expect to be significant for satellite constitution application, via, for example, the Earth developing a non - negligible eccentricity and inclination with respect to the protoplanetary disc for naturalistic disc parameters. These stopping point and the analytic result (Equation \[eq: v\_MVR17\ ]) were confirm with numerical simulations by [ @VelascoRomeroMasset19 ], [ @ChrenkoLambrechts19 ] and [ @GuileraEtal19 ].
In application to our finical problem, we note that the density in the central region of the protoplanet is initially homogeneous with a nearly constant $ \rho_0 \approx 10^{-9}$ gcm$^{-3}$. Hence, the constant background concentration results of [ @MassetVelascoRomero17 ] are applicable. However, our protoplanet is finite in extent. When the core is displaced by a distance $ R$ from the centre, there is a returning effect of protoplanet graveness, given by $ $ F_{\rm g } = - \frac{G M(R) M_{\rm c}}{R^2}\; ,
\label{eq: Fg0}$$ where $ M(R) \approx (4\pi/3) \rho_0 R^3 $ is the enclosed mass within radius $ R$.
The motion and radius of the orbiting effect is then sic by a balance of forces. In the azimuthal steering the heating system and dynamical friction force specify the core velocity from Equation \[eq: v\_MVR17\ ]. Given this velocity ($ V_0 $), the radial stead of the core ($ R$) is simply set by the gravitational force of the enclosed protoplanet which provides the centrifugal power. We see that $ V_0 $ is independent of orbit radius $ R$ whereas the gravitational force is propotional to $ R$, as long as the envelop gas density is roughly constant. The balance of these consequently establishes an equilibrium radius of the orbit for the core of mass $ M_{\rm c}$ and luminosity $ L | c_s}{8 \pi \rho_0 G M_{\rm c} \chi},
\label{tq:v_MVR17}$$ where $\gammc$ is thx adiabztic inddx of the gas, $L_{\rm c}$ and $M_{\rm r}$ arw the luminosity and mass ow the pernurbing cire end $\chi$ is the tisrmal conductjyity mh the gas that jhey assumed to be constand. Ghzy noted that the effects of the heaeing fotcf are expected to nq sifnificant for planet formation appmicatioi, via, e.g., the Eatth developing a non-negliglble eccentricity and lnclination witr respect to ghe protoplanetary disd for realistic disc parameters. Thesz conclusiobs anf the analytmc resllt (Equation \[ca:v\_MVR17\]) fere comfirmed with nmmeriral wimulations by [@VelascmRomeroMasset19], [@ChregkoLambrewhcs19] and [@GuileraEtal19].
In qpplicajion do oje pxrtjcnlad probpem, we note tgat the denwity in the central pwgy of the profoplanqt is initially homogeneous with a nearly cohstant $\rho_0 \approx 10^{-9}$ gcm$^{-3}$. Hwnce, the constant bacnground dqnsity results of [@MassetVelascoRomero17] are applicatle. Hkdevtr, our oeohoplanet is finite in extent. When the core is sixpkaced by a disbance $R$ from the crnhrr, there is a rgturniny fkrce of protoplaneh gravijy, givwn by $$F_{\rm g} = - \frac{G M(R) M_{\rm c}}{R^2}\;,
\label{eq:Dg0}$$ where $M(R) \cppeox (4\pi/3) \rho_0 R^3$ is thz enclosed mcss wijhin rsdius $R$.
The motion and rcdius kf the orbihing core ks then set by a bakatce of forces. In the azimuehal dirertion the hextinb and qynamical vrictlmn force specify tje cote velmcity from Equation \[eq:v\_MVR17\]. Given this velocity ($V_0$), the racidl iosition jf thc core ($R$) is sim[ly set by the gravitctionau force of the enrlosed proto[lanet which [tovides the cxntrifugaj foece. Qe see gfat $V_0$ is indeprndent of orbit raduus $R$ whereas the nravijafional force is pripotional to $R$, ss uond ws tra enclosed gds ddnskyy is roughly cokstxnt. Yhe balance of these thedefore establishes ak equilibtium radits of the orbot for the core of mass $M_{\ro c}$ amd juminosity $L | c_s}{8 \pi \rho_0 G M_{\rm c} \chi}, $\gamma$ the adiabatic of the gas, are luminosity and mass the perturbing core $\chi$ is the thermal conductivity of gas that they assumed to be constant. They noted that the effects of heating force are expected to be significant for planet formation application, via, e.g., Earth a eccentricity inclination with respect to the protoplanetary disc for realistic disc parameters. These conclusions and the analytic (Equation \[eq:v\_MVR17\]) were confirmed with numerical simulations by [@ChrenkoLambrechts19] and [@GuileraEtal19]. In to our particular problem, we that density in central of protoplanet is initially with a nearly constant $\rho_0 \approx 10^{-9}$ gcm$^{-3}$. Hence, the constant background density results of [@MassetVelascoRomero17] are However, our finite in When core displaced by a from the centre, there is a protoplanet gravity, given by $$F_{\rm g} = - M(R) M_{\rm \label{eq:Fg0}$$ where $M(R) \approx (4\pi/3) \rho_0 is the enclosed mass within radius $R$. The and radius of the orbiting core is then set by a balance of forces. In direction the heating and friction force specify core from \[eq:v\_MVR17\]. this velocity the radial position of the core ($R$) is simply set by gravitational force of the enclosed protoplanet which provides the centrifugal see $V_0$ is independent orbit radius $R$ whereas gravitational is propotional to $R$, as enclosed roughly The of these therefore establishes equilibrium radius of the orbit the core of mass | c_s}{8 \pi \rho_0 G M_{\rm c} \chi},
\label{eq:v_MVr17}$$ where $\gammA$ is thE adIabAtIc inDex oF the gas, $L_{\rm c}$ and $m_{\Rm c}$ aRe the luminosity and mass Of the PeRTurbINg Core aNd $\chi$ is THe THErmAl CoNduCtIViTy of tHe gAs that tHey assumed To bE cOnstant. They nOTeD that the efFecTs of the heatiNg fOrce arE eXpeCTed to Be sIgnifIcant fOR planeT formatioN aPPlicatIOn, via, e.g., THE EArth Developing a non-negLIgIBle eccentricitY and inClINaTIOn wIth Respect to tHe ProtoPLanetarY DiSC FOr rEAlistic disc paRameters. TheSE coNclusiOnS anD The anaLytic ReSUlt (equation \[eq:v\_mVR17\]) wEre confirMed witH NumericAL simulaTions bY [@VeLasCoRoMErOMAssEt19], [@cHreNKolamBRecHts19] and [@GuIlErAEtal19].
in apPLICAtioN to Our pArticUlar problem, we NotE thaT The DensiTy in tHe ceNtRal paRt of thE protOpLanet is initiallY homOgeneous wIth A nEarLy ConstANt $\rho_0 \aPprOx 10^{-9}$ gCm$^{-3}$. Hence, The consTAnt BaCKGRoUnd density results oF [@MASSeTVelascoromero17] ARe ApPLicable. HOwEveR, our PROtoplAnet IS fInite in eXtent. WHEn ThE core is DiSplaceD bY a dIstAnce $R$ FRom tHe centRe, there iS a retURning force of prOToplanet graviTY, gIVEn BY $$F_{\rm G} = - \frAc{G M(R) M_{\rm c}}{R^2}\;,
\lAbel{EQ:Fg0}$$ wHere $m(r) \aPprOX (4\pi/3) \rhO_0 R^3$ is tHe ENcLOsed mass within radiuS $R$.
the motIon anD radius of the oRbiting corE IS Then set bY a baLAnCE of forces. In the AzimuThal directIOn the heaTing aNd dynamiCal frictiON Force speCifY thE coRe vELOcIty from EquatiON \[Eq:v\_MvR17\]. given thIs vElocity ($v_0$), thE raDiaL poSiTion of the Core ($R$) is sImPlY sEt By tHe graVItationaL fOrcE oF thE enclOSed proToplaNet wHiCh PRovIdes the CEnTRIfugAl FoRce. WE seE tHat $V_0$ iS indEPenDent of oRbit radiuS $R$ wHEreaS tHe GravitaTional force is PrOpotional tO $R$, As lOng as tHE Enclosed Gas density is roughly consTAnt. The bAlaNce of ThesE thereforE esTablisHes AN equilIbrium RadiuS oF thE ORbit fOR ThE coRe Of mass $M_{\rm c}$ AND luMinosItY $L | c_s}{8 \pi \rho_0 G M_{\r m c} \chi} ,
\la bel {eq :v _MVR 17}$ $ where $\gamm a $ is the adiabatic index o f the g a s, $ L _{ \rm c }$ and$ M_ { \ rmc} $are t h elumin osi ty andmass of th e p er turbing core an d $\chi$ i s t he thermal c ond uctivi ty of the g asthatthey a s sumedto be con st a nt. Th e y noted t ha t th e effects of theh ea t ing force areexpect ed to b e s ign ificant fo rplane t format i on a p pli c ation, via, e .g., the Ea r thdevelo pi nga non-n eglig ib l e e ccentricity and inclinat ion wi t h respe c t to th e prot opl ane tary di sc fo rr eal i st icd isc paramet er s. Thes e co n c l u sion s a nd t he an alytic result (E quat i on\[eq: v\_MV R17\ ]) were confi rmedwi th numerical si mula tions by[@V el asc oR omero M asset1 9], [@ Chrenko Lambrec h ts1 9] a n d[@GuileraEtal19].
I n ap plicatio n to o u rpa r ticularpr obl em,w e note tha t t he densi ty int he c entralpa rt ofth e p rot oplan e t is initi ally hom ogene o us with a near l y constant $\ r ho _ 0 \ a ppro x 1 0^{-9}$ gcm $^{- 3 }$.Henc e ,the const ant b ac k gr o und density results o f [@Ma ssetV elascoRomero1 7] are app l i c able. Ho weve r ,o ur protoplanet is f inite in e x tent. Wh en th e core i s displac e d by a di sta nce $R $ f r o mthe centre, t h e re i sa retur nin g force of pr oto pla ne t gravity , givenby $ $F _{ \rm g} = - \frac{ GM(R )M_{ \rm c } }{R^2} \;,
\ la be l {eq :Fg0}$$ wh e r e $M (R )\app rox ( 4\pi/ 3) \ r ho_ 0 R^3$is the en clo s ed m as swithinradius $R$.
Th e motion a nd ra dius o f the orbi ting core is then set b y a bala nce of f orce s. In the az imutha l d i rectio n theheati ng an d dynam i c al fr ic tion force s pec ify t he cor e veloc ity from Equation\ [eq :v\_MVR17\].Giv en t h i svel o ci t y ( $V _ 0$) , the radial posi tion of th ec or e ($R$) is sim pl y set b y the g ravit a tionalforce ofthe enclo se d pr o t opl anet which provide s the cen t rifug a lforce . W e seeth at$V_0$ is in d epe ndent of or bi t radi us $R $whereasthe gravitational force is pr opoti ona l to $R$, as lon g as theencl osed gas d ens ity is r oug h ly co nsta n t. Th e bala nceo f these t h er efo r e e stablishesa n equ ilibr ium radius ofthe orbit for the core of mass $ M_{\ r m c} $ a n d lu mi nosity $L | c_s}{8_\pi \rho_0_G M_{\rm c} \chi},
\label{eq:v_MVR17}$$_where $\gamma$_is_the adiabatic_index_of the gas,_$L_{\rm c}$ and_$M_{\rm c}$ are the_luminosity and mass_of_the perturbing core and $\chi$ is the thermal conductivity of the gas that they_assumed_to be_constant._They_noted that the effects of_the heating force are expected_to be_significant for planet formation application, via, e.g., the_Earth_developing a non-negligible_eccentricity and inclination with respect to the protoplanetary disc_for realistic disc parameters. These conclusions_and the analytic_result_(Equation_\[eq:v\_MVR17\]) were confirmed with_numerical simulations by [@VelascoRomeroMasset19], [@ChrenkoLambrechts19] and_[@GuileraEtal19].
In application to our particular problem,_we note that the density in the_central part of the protoplanet is_initially homogeneous with a nearly_constant $\rho_0_\approx 10^{-9}$ gcm$^{-3}$. Hence, the constant_background density results_of [@MassetVelascoRomero17]_are applicable. However,_our protoplanet is finite in extent._When the core_is displaced by a distance $R$_from_the centre, there_is_a_returning force_of protoplanet gravity,_given_by $$F_{\rm_g}_= - \frac{G M(R) M_{\rm c}}{R^2}\;,
__ \label{eq:Fg0}$$ where $M(R) \approx (4\pi/3)_\rho_0 R^3$ is the_enclosed_mass within radius $R$.
The_motion and radius of the_orbiting core is then set by_a balance_of forces._In the azimuthal direction the heating and dynamical friction force specify_the core velocity from Equation \[eq:v\_MVR17\]._Given this velocity ($V_0$),_the radial_position_of the core_($R$)_is simply_set by the gravitational force of the_enclosed protoplanet_which provides the centrifugal force. We_see that $V_0$ is_independent_of orbit radius $R$ whereas the_gravitational force is propotional to $R$,_as long as the enclosed_gas_density_is roughly constant. The balance_of these therefore establishes an equilibrium_radius of the_orbit for the core of mass $M_{\rm_c}$_and luminosity $L |
-n}{\left(}n-1{\right)}!{\left(}j/2{\right)}!\,\omega_n}{{\left(}n-2{\right)}{\left(}n+j/2-1{\right)}!\,\omega_{n-1}}\operatorname{B}{\left(}\frac{n}{2},\frac{n+j}{2}{\right)}^{-1}.$$ By using – together with the identity $$\begin{gathered}
\operatorname{B}{\left(}\frac{n}{2},\frac{n+j'+j''-2m}{2}{\right)}^{-1}\operatorname{B}{\left(}\frac{n+j'+j''-2m}{2},\frac{n+j'-j''+2l'-4k}{2}{\right)}\\
=\frac{{\left(}\frac{j'+j''}{2}+n-m-1{\right)}!}{{\left(}n+j'-m+l'-2k-1{\right)}!{\left(}\frac{j''-j'}{2}+2k-l'-1{\right)}!}\\
\times\operatorname{B}{\left(}\frac{n+j'-j''+2l'-4k}{2},\frac{j''-j'+4k-2l'}{2}{\right)},\end{gathered}$$ we obtain that if $j'+j''$ is even, then $$\begin{gathered}
\label{Step3Eq24}
\int_0^{{\left(}r_0/\mu{\right)}^2}\frac{r^{\frac{n+j'+j''-2m-2}{2}}}{{\left(}1+r{\right)}^{n+j'-m+l'-2k}}\,dr\int_{{\mathbb{S}}^{n-1}}y_{i_{\sigma(2m+1)}}\dotsm y_{i_{\sigma(j')}}y_{i_{j'+1}}\dotsm y_{i_{j'+j''}}dv_{g_0}{\left(}y{\right)}\\
=\frac{2^{1-n-j'-j''+2m}{\left(}n-1{\right)}!\,\omega_n}{{\left(}n+j'-m+l'-2k-1{\right)}!{\left(}\frac{j''-j'}{2}+ | -n}{\left(}n-1{\right)}!{\left(}j/2{\right)}!\,\omega_n}{{\left(}n-2{\right)}{\left(}n+j/2 - 1{\right)}!\,\omega_{n-1}}\operatorname{B}{\left(}\frac{n}{2},\frac{n+j}{2}{\right)}^{-1}.$$ By using – together with the identity $ $ \begin{gathered }
\operatorname{B}{\left(}\frac{n}{2},\frac{n+j'+j''-2m}{2}{\right)}^{-1}\operatorname{B}{\left(}\frac{n+j'+j''-2m}{2},\frac{n+j'-j''+2l'-4k}{2}{\right)}\\
= \frac{{\left(}\frac{j'+j''}{2}+n - m-1{\right)}!}{{\left(}n+j'-m+l'-2k-1{\right)}!{\left(}\frac{j''-j'}{2}+2k - l'-1{\right)}!}\\
\times\operatorname{B}{\left(}\frac{n+j'-j''+2l'-4k}{2},\frac{j''-j'+4k-2l'}{2}{\right)},\end{gathered}$$ we obtain that if $ j'+j''$ is even, then $ $ \begin{gathered }
\label{Step3Eq24 }
\int_0^{{\left(}r_0/\mu{\right)}^2}\frac{r^{\frac{n+j'+j''-2m-2}{2}}}{{\left(}1+r{\right)}^{n+j'-m+l'-2k}}\,dr\int_{{\mathbb{S}}^{n-1}}y_{i_{\sigma(2m+1)}}\dotsm y_{i_{\sigma(j')}}y_{i_{j'+1}}\dotsm y_{i_{j'+j''}}dv_{g_0}{\left(}y{\right)}\\
= \frac{2^{1 - n - j'-j''+2m}{\left(}n-1{\right)}!\,\omega_n}{{\left(}n+j'-m+l'-2k-1{\right)}!{\left(}\frac{j''-j'}{2}+ | -n}{\levt(}n-1{\right)}!{\left(}j/2{\right)}!\,\omega_n}{{\ueft(}n-2{\right)}{\left(}n+l/2-1{\eight)}!\,\okega_{n-1}}\oleratornxme{B}{\left(}\frac{n}{2},\frac{n+j}{2}{\right)}^{-1}.$$ By nsint – totether with the identigy $$\begin{gwthered}
\operauorname{B}{\left(}\frac{n}{2},\hdac{n+j'+j''-2m}{2}{\vnght)}^{-1}\olcratoxneme{B}{\left(}\frac{n+j'+j''-2k}{2},\frac{n+j'-j''+2l'-4k}{2}{\sight)}\\
=\frac{{\left(}\fsaz{j'+l''}{2}+n-m-1{\right)}!}{{\left(}n+j'-m+l'-2k-1{\right)}!{\left(}\frac{j''-j'}{2}+2k-l'-1{\ridht)}!}\\
\timex\ooeratorname{B}{\lest(}\frss{n+j'-j''+2m'-4k}{2},\frac{j''-j'+4k-2l'}{2}{\right)},\end{gathered}$$ we obtajn that if $j'+j''$ is evem, then $$\begin{gathered}
\label{Shep3Ee24}
\int_0^{{\left(}r_0/\mu{\right)}^2}\frwc{r^{\frac{n+j'+j''-2m-2}{2}}}{{\oeft(}1+w{\eight)}^{n+j'-m+l'-2k}}\,dr\knt_{{\mathbb{S}}^{n-1}}y_{i_{\sigma(2m+1)}}\dotam y_{i_{\sigma(j')}}y_{i_{j'+1}}\dotsm y_{i_{j'+j''}}dv_{g_0}{\left(}h{\righc)}\\
=\frac{2^{1-n-j'-j''+2m}{\ledt(}b-1{\rihvt)}!\,\omega_n}{{\lefv(}n+j'-m+l'-2k-1{\gight)}!{\left(}\frac{m''-k'}{2}+ | -n}{\left(}n-1{\right)}!{\left(}j/2{\right)}!\,\omega_n}{{\left(}n-2{\right)}{\left(}n+j/2-1{\right)}!\,\omega_{n-1}}\operatorname{B}{\left(}\frac{n}{2},\frac{n+j}{2}{\right)}^{-1}.$$ By using – together with the \operatorname{B}{\left(}\frac{n}{2},\frac{n+j'+j''-2m}{2}{\right)}^{-1}\operatorname{B}{\left(}\frac{n+j'+j''-2m}{2},\frac{n+j'-j''+2l'-4k}{2}{\right)}\\ \times\operatorname{B}{\left(}\frac{n+j'-j''+2l'-4k}{2},\frac{j''-j'+4k-2l'}{2}{\right)},\end{gathered}$$ we that if $j'+j''$ \int_0^{{\left(}r_0/\mu{\right)}^2}\frac{r^{\frac{n+j'+j''-2m-2}{2}}}{{\left(}1+r{\right)}^{n+j'-m+l'-2k}}\,dr\int_{{\mathbb{S}}^{n-1}}y_{i_{\sigma(2m+1)}}\dotsm y_{i_{j'+j''}}dv_{g_0}{\left(}y{\right)}\\ =\frac{2^{1-n-j'-j''+2m}{\left(}n-1{\right)}!\,\omega_n}{{\left(}n+j'-m+l'-2k-1{\right)}!{\left(}\frac{j''-j'}{2}+ | -n}{\left(}n-1{\right)}!{\left(}j/2{\right)}!\,\omegA_n}{{\left(}n-2{\rigHt)}{\lefT(}n+j/2-1{\RigHt)}!\,\OmegA_{n-1}}\opEratorname{B}{\lefT(}\Frac{N}{2},\frac{n+j}{2}{\right)}^{-1}.$$ By using – togEther WiTH the IDeNtity $$\Begin{gaTHeRED}
\opErAtOrnAmE{b}{\lEft(}\frAc{n}{2},\Frac{n+j'+j''-2M}{2}{\right)}^{-1}\operAtoRnAme{B}{\left(}\frac{N+J'+j''-2M}{2},\frac{n+j'-j''+2l'-4k}{2}{\RigHt)}\\
=\frac{{\left(}\frAc{j'+J''}{2}+n-m-1{\rigHt)}!}{{\LefT(}N+j'-m+l'-2k-1{\RigHt)}!{\lefT(}\frac{j''-J'}{2}+2K-l'-1{\righT)}!}\\
\times\opeRaTOrname{b}{\Left(}\fraC{N+J'-j''+2L'-4k}{2},\frAc{j''-j'+4k-2l'}{2}{\right)},\end{gatHErED}$$ we obtain that iF $j'+j''$ is eVeN, ThEN $$\BegIn{gAthered}
\labEl{step3EQ24}
\Int_0^{{\left(}R_0/\Mu{\RIGHt)}^2}\fRAc{r^{\frac{n+j'+j''-2m-2}{2}}}{{\leFt(}1+r{\right)}^{n+j'-m+L'-2K}}\,dr\Int_{{\matHbB{S}}^{n-1}}Y_{I_{\sigma(2M+1)}}\dotsM y_{I_{\SigMa(j')}}y_{i_{j'+1}}\dotsm Y_{i_{j'+j''}}Dv_{g_0}{\left(}y{\rIght)}\\
=\frAC{2^{1-n-j'-j''+2m}{\leFT(}n-1{\right)}!\,\Omega_n}{{\LefT(}n+j'-M+l'-2k-1{\rIGhT)}!{\lEft(}\FrAC{j''-j'}{2}+ | -n}{\left(}n-1{\right)}!{\ left(}j/2{ \righ t)} !\, \o mega _n}{ {\left(}n-2{\r i ght) }{\left(}n+j/2-1{\righ t)}!\ ,\ o mega _ {n -1}}\ operato r na m e {B} {\ le ft( }\ f ra c{n}{ 2}, \frac{n +j}{2}{\ri ght )} ^{-1}.$$ Byu si ng – toget her with the id ent ity $$ \b egi n {gath ere d}
\o perato r name{B }{\left(} \f r ac{n}{ 2 },\frac { n +j '+j' '-2m}{2}{\right)} ^ {- 1 }\operatorname {B}{\l ef t (} \ f rac {n+ j'+j''-2m} {2 },\fr a c{n+j'- j '' + 2 l '-4 k }{2}{\right)} \\
=\frac{{ \ lef t(}\fr ac {j' + j''}{2 }+n-m -1 { \ri ght)}!}{{\l eft( }n+j'-m+l '-2k-1 { \right) } !{\left (}\fra c{j ''- j'}{ 2 }+ 2k -l' -1 { \ri g ht )}! } \\\times\o pe ra torna me{B } { \ l eft( }\f rac{ n+j'- j''+2l'-4k}{2 },\ frac { j'' -j'+4 k-2l' }{2} {\ right )},\en d{gat he red}$$ we obtai n th at if $j' +j' '$ is e ven,t hen $$ \be gin {gather ed}
\la b el{ St e p 3 Eq 24}
\int_0^{{\left (} r _ 0/ \mu{\rig ht)}^2 } \f ra c {r^{\fra c{ n+j '+j' ' - 2m-2} {2}} } {{ \left(}1 +r{\ri g ht )} ^{n+j'- m+ l'-2k} }\ ,dr \in t_{{\ m athb b{S}}^ {n-1}}y_ {i_{\ s igma(2m+1)}}\d o tsm y_{i_{\si g ma ( j ') } }y_{ i_{ j'+1}}\dots m y_ { i_{j '+j' ' }} dv_ { g_0}{ \left (} y {\ r ight)}\\
=\frac{2^{ 1- n-j'-j ''+2m }{\left(}n-1{ \right)}!\ , \ o mega_n}{ {\le f t( } n+j'-m+l'-2k-1 {\rig ht)}!{\lef t (}\frac{ j''-j '}{2}+ | -n}{\left(}n-1{\right)}!{\left(}j/2{\right)}!\,\omega_n}{{\left(}n-2{\right)}{\left(}n+j/2-1{\right)}!\,\omega_{n-1}}\operatorname{B}{\left(}\frac{n}{2},\frac{n+j}{2}{\right)}^{-1}.$$ By_using –_together with the identity_$$\begin{gathered}
\operatorname{B}{\left(}\frac{n}{2},\frac{n+j'+j''-2m}{2}{\right)}^{-1}\operatorname{B}{\left(}\frac{n+j'+j''-2m}{2},\frac{n+j'-j''+2l'-4k}{2}{\right)}\\
=\frac{{\left(}\frac{j'+j''}{2}+n-m-1{\right)}!}{{\left(}n+j'-m+l'-2k-1{\right)}!{\left(}\frac{j''-j'}{2}+2k-l'-1{\right)}!}\\
\times\operatorname{B}{\left(}\frac{n+j'-j''+2l'-4k}{2},\frac{j''-j'+4k-2l'}{2}{\right)},\end{gathered}$$ we_obtain_that if_$j'+j''$_is even, then_$$\begin{gathered}
\label{Step3Eq24}
\int_0^{{\left(}r_0/\mu{\right)}^2}\frac{r^{\frac{n+j'+j''-2m-2}{2}}}{{\left(}1+r{\right)}^{n+j'-m+l'-2k}}\,dr\int_{{\mathbb{S}}^{n-1}}y_{i_{\sigma(2m+1)}}\dotsm y_{i_{\sigma(j')}}y_{i_{j'+1}}\dotsm y_{i_{j'+j''}}dv_{g_0}{\left(}y{\right)}\\
=\frac{2^{1-n-j'-j''+2m}{\left(}n-1{\right)}!\,\omega_n}{{\left(}n+j'-m+l'-2k-1{\right)}!{\left(}\frac{j''-j'}{2}+ |
primary link vector ${\vec{r}_{d,l}}^{\,0}$ is given in Eq. (\[r-r0-T\]).
For the model on the square lattice there are four primary links, one per family. Their link vectors ${\vec{r}_{d,l}}^{\,0}$ have components such that $N_1=N_2=0$ in Eqs. (\[r-r0-T\]) and (\[xd-xd\]). Therefore, the primary links have minimum length $\xi_{{\vec{r}}_{d,l}^{\,0}}=a_s$. Alike the remaining links of its family, the centre of a primary link is located at $\vec{r}=\vec{r}_{j}+{\vec{r}_{d,l}}^{\,0}$. For the square lattice there are two horizontal primary links whose centers are located at $\vec{r}_{j}+\vec{r}_{1,l}^{\,0}$ with $l=\pm 1$ and two vertical primary links whose centers are located at $\vec{r}_{j}+\vec{r}_{2,l}^{\,0}$ with $l=\pm 1$. In the case of the 1D lattice there are two primary links whose centers are located at $\vec{r}_{j}+\vec{r}_{1,l}^{\,0}$ with $l=\pm 1$.
Partitions and $g$-primary partitions
-------------------------------------
The building blocks of the $N_{s1}^h=0$ configuration state are singlet pairs of spinons on sites $\vec{r}_j^{\,-}$ and $\vec{r}_j^{\,+}$ of the spin effective lattice, $$\begin{aligned}
\vert\vec{r}_j^{\,-},\vec{r}_j^{\,+}\rangle
& = & {1\over \sqrt{2}}\left(\vert\uparrow_{\vec{r}_j^{\,-}}
\downarrow_{\vec{r}_j^{\,+}}\rangle -
\vert\downarrow_{\vec{r}_j^{\,-}}
\uparrow_{\vec{r}_j^{\,+}}\rangle\right) \,,
\nonumber \\
\vec{r}_j^{\,\mp} & = & \vec{r}_j + \vec{r}_{d,l}^{\,0}
\mp \vec{r}_{d,l}^{\,g}
\, ; \hspace{0.5cm}
d = d (j) | primary link vector $ { \vec{r}_{d, l}}^{\,0}$ is given in Eq. (\[r - r0 - T\ ]).
For the model on the straight wicket there are four primary connection, one per kin. Their link vectors $ { \vec{r}_{d, l}}^{\,0}$ have component such that $ N_1 = N_2=0 $ in Eqs. (\[r - r0 - T\ ]) and (\[xd - xd\ ]). Therefore, the primary link have minimum distance $ \xi_{{\vec{r}}_{d, l}^{\,0}}=a_s$. Alike the remaining links of its kin, the center of a primary link is located at $ \vec{r}=\vec{r}_{j}+{\vec{r}_{d, l}}^{\,0}$. For the square lattice there embody two horizontal primary links whose centers are settle at $ \vec{r}_{j}+\vec{r}_{1,l}^{\,0}$ with $ l=\pm 1 $ and two vertical primary links whose centers are locate at $ \vec{r}_{j}+\vec{r}_{2,l}^{\,0}$ with $ l=\pm 1$. In the case of the 1D lattice there are two primary links whose center are located at $ \vec{r}_{j}+\vec{r}_{1,l}^{\,0}$ with $ l=\pm 1$.
Partitions and $ g$-primary partitions
-------------------------------------
The building blocks of the $ N_{s1}^h=0 $ configuration department of state are singlet pairs of spinons on sites $ \vec{r}_j^{\,-}$ and $ \vec{r}_j^{\,+}$ of the spin effective lattice, $ $ \begin{aligned }
\vert\vec{r}_j^{\,-},\vec{r}_j^{\,+}\rangle
& = & { 1\over \sqrt{2}}\left(\vert\uparrow_{\vec{r}_j^{\,- } }
\downarrow_{\vec{r}_j^{\,+}}\rangle -
\vert\downarrow_{\vec{r}_j^{\,- } }
\uparrow_{\vec{r}_j^{\,+}}\rangle\right) \, ,
\nonumber \\
\vec{r}_j^{\,\mp } & = & \vec{r}_j + \vec{r}_{d, l}^{\,0 }
\mp \vec{r}_{d, l}^{\,g }
\, ; \hspace{0.5 cm }
d = d (j ) | prlmary link vector ${\vec{r}_{d,l}}^{\,0}$ is given in Eq. (\[r-r0-T\]).
For the mkdel on ghe square lattice there are fiur peimary links, one per fxmily. Thepr link vwctocs ${\vec{r}_{d,l}}^{\,0}$ have components such bhat $U_1=N_2=0$ in Eqs. (\[r-r0-T\]) anc (\[xd-xd\]). Thesefore, the prikafy links have minimum length $\xi_{{\vec{r}}_{d,l}^{\,0}}=w_s$. Alikr hhe remaining jinkx of jns family, the centre of a primady link is located ay $\vec{r}=\vec{r}_{j}+{\vec{r}_{d,l}}^{\,0}$. For the dquage lattice there age two horiekntwo primary livks whose bznters are mocated at $\vec{r}_{j}+\vec{r}_{1,l}^{\,0}$ with $l=\pm 1$ and cwo verticao prildry links wiose cvnters are logsted ad $\vec{r}_{j}+\fec{r}_{2,l}^{\,0}$ with $l=\pm 1$. Ii thw case of the 1D lattire there are two priiary linkv chose centers are locqtwd at $\vec{s}_{j}+\vez{e}_{1,l}^{\,0}$ ditg $k=\pj 1$.
Partltikns and $g$-pdimary partutions
-------------------------------------
The building nljbls of the $N_{s1}^g=0$ confygtration state are singlet pairs of spinmns on sites $\vec{r}_j^{\,-}$ and $\vec{e}_j^{\,+}$ of the spin effectlve lattise, $$\begin{aligned}
\vert\vec{r}_j^{\,-},\vec{r}_j^{\,+}\rangle
& = & {1\over \sqrt{2}}\lext(\verv\uoarxiw_{\vec{f}_h^{\,-}}
\dlwnarrow_{\vec{r}_j^{\,+}}\rangle -
\vert\downarrow_{\vec{r}_j^{\,-}}
\uparrow_{\dsc{t}_j^{\,+}}\gangle\right) \,,
\nonumner \\
\vec{r}_j^{\,\mp} & = & \vec{t}_j + \fgc{r}_{d,l}^{\,0}
\mp \vec{r}_{d,l}^{\,e}
\, ; \hs'zcs{0.5cm}
d = d (j) | primary link vector ${\vec{r}_{d,l}}^{\,0}$ is given in For model on square lattice there per Their link vectors have components such $N_1=N_2=0$ in Eqs. (\[r-r0-T\]) and (\[xd-xd\]). the primary links have minimum length $\xi_{{\vec{r}}_{d,l}^{\,0}}=a_s$. Alike the remaining links of its the centre of a primary link is located at $\vec{r}=\vec{r}_{j}+{\vec{r}_{d,l}}^{\,0}$. For the square there two primary whose centers are located at $\vec{r}_{j}+\vec{r}_{1,l}^{\,0}$ with $l=\pm 1$ and two vertical primary links whose centers located at $\vec{r}_{j}+\vec{r}_{2,l}^{\,0}$ with $l=\pm 1$. In the of the 1D lattice are two primary links whose are at $\vec{r}_{j}+\vec{r}_{1,l}^{\,0}$ $l=\pm Partitions $g$-primary partitions ------------------------------------- building blocks of the $N_{s1}^h=0$ configuration state are singlet pairs of spinons on sites $\vec{r}_j^{\,-}$ and $\vec{r}_j^{\,+}$ the spin $$\begin{aligned} \vert\vec{r}_j^{\,-},\vec{r}_j^{\,+}\rangle = {1\over \downarrow_{\vec{r}_j^{\,+}}\rangle - \vert\downarrow_{\vec{r}_j^{\,-}} \nonumber \\ \vec{r}_j^{\,\mp} & = & \mp \vec{r}_{d,l}^{\,g} \, ; \hspace{0.5cm} d = d | primary link vector ${\vec{r}_{d,l}}^{\,0}$ is Given in Eq. (\[r-R0-T\]).
For The ModEl On thE squAre lattice therE Are fOur primary links, one per fAmily. thEIr liNK vEctorS ${\vec{r}_{d,l}}^{\,0}$ HAvE COmpOnEnTs sUcH ThAt $N_1=N_2=0$ iN EqS. (\[r-r0-T\]) and (\[Xd-xd\]). TherefOre, ThE primary linkS HaVe minimum lEngTh $\xi_{{\vec{r}}_{d,l}^{\,0}}=a_s$. aliKe the rEmAinINg linKs oF its fAmily, tHE centrE of a primaRy LInk is lOCated at $\VEC{r}=\Vec{r}_{J}+{\vec{r}_{d,l}}^{\,0}$. For the squaRE lATtice there are tWo horiZoNTaL PRimAry Links whose CeNters ARe locatED aT $\VEC{r}_{j}+\VEc{r}_{1,l}^{\,0}$ with $l=\pm 1$ anD two verticaL PriMary liNkS whOSe centErs arE lOCatEd at $\vec{r}_{j}+\veC{r}_{2,l}^{\,0}$ wIth $l=\pm 1$. In tHe case OF the 1D laTTice theRe are tWo pRimAry lINkS wHosE cENteRS aRe lOCatEd at $\vec{r}_{J}+\vEc{R}_{1,l}^{\,0}$ witH $l=\pm 1$.
pARTItioNs aNd $g$-pRimarY partitions
-------------------------------------
ThE buIldiNG blOcks oF the $N_{S1}^h=0$ coNfIguraTion stAte arE sInglet pairs of spInonS on sites $\vEc{r}_J^{\,-}$ aNd $\vEc{R}_j^{\,+}$ of tHE spin eFfeCtiVe lattiCe, $$\begin{ALigNeD}
\VERt\Vec{r}_j^{\,-},\vec{r}_j^{\,+}\rangle
& = & {1\ovEr \SQRt{2}}\Left(\vert\UparroW_{\VeC{r}_J^{\,-}}
\DownarroW_{\vEc{r}_J^{\,+}}\ranGLE -
\vert\DownARrOw_{\vec{r}_j^{\,-}}
\uParrow_{\VEc{R}_j^{\,+}}\Rangle\rIgHt) \,,
\nonuMbEr \\
\vEc{r}_J^{\,\mp} & = & \veC{R}_j + \veC{r}_{d,l}^{\,0}
\mp \Vec{r}_{d,l}^{\,g}
\, ; \hSpace{0.5CM}
d = d (j) | primary link vector ${\ve c{r}_{d,l} }^{\, 0}$ is g iven inEq. (\[r-r0-T\ ] ).
For the model on the s quare l a ttic e t hereare fou r p r i mar yli nks ,o ne perfam ily. Th eir link v ect or s ${\vec{r}_ { d, l}}^{\,0}$ ha ve component s s uch th at $N _ 1=N_2 =0$ in E qs. (\ [ r-r0-T \]) and ( \[ x d-xd\] ) . There f o re , th e primary links h a ve minimum length $\xi_ {{ \ ve c { r}} _{d ,l}^{\,0}} =a _s$.A like th e r e m a ini n g links of it s family, t h e c entreof ap rimary link i s lo cated at $\ vec{ r}=\vec{r }_{j}+ { \vec{r} _ {d,l}}^ {\,0}$ . F orthes qu ar e l at t ice th ere are two hor iz on tal p rima r y l inks wh osecente rs are locate d a t $\ v ec{ r}_{j }+\ve c{r} _{ 1,l}^ {\,0}$ with $ l=\pm 1$ and tw o ve rtical pr ima ry li nk s who s e cent ers ar e locat ed at $ \ vec {r } _ { j} +\vec{r}_{2,l}^{\, 0} $ wi th $l=\p m 1$.I nth e case of t he1D l a t ticether e a re two p rimary li nk s whose c enters a reloc ateda t $\ vec{r} _{j}+\ve c{r}_ { 1,l}^{\,0}$ wi t h $l=\pm 1$.Pa r t it i onsand $g$-primar y pa r titi ons- -- --- - ----- ----- -- - -- - --------------
The b uildin g blo cks of the $N _{s1}^h=0$ c o nfigurat ions ta t e are singletpairs of spinon s on site s $\v ec{r}_j^ {\,-}$ an d $\vec{r} _j^ {\, +}$ of t he spin effecti v e lat ti ce, $$\ beg in{alig ned }
\ ver t\v ec {r}_j^{\, -},\vec{ r} _j ^{ \, +}\ rangl e
& = & { 1\ ove r\sq rt{2} } \left( \vert \upa rr ow _ {\v ec{r}_j ^ {\ , - }}
\ do wn arro w_{ \v ec{r} _j^{ \ ,+} }\rangl e -
\vert \do w narr ow _{ \vec{r} _j^{\,-}}
\up ar row_{\vec{ r} _j^ {\,+}} \ r angle\ri ght) \,,
\nonumber \\
\ v ec{r}_j ^{\ ,\mp} & = & \vec{r }_j + \ve c{r } _{d,l} ^{\,0}
\mp\v ec{ r } _{d,l } ^ {\ ,g} \, ; \hspa c e {0. 5cm}d= d(j) | primary_link vector_${\vec{r}_{d,l}}^{\,0}$ is given in_Eq. (\[r-r0-T\]).
For_the_model on_the_square lattice there_are four primary_links, one per family._Their link vectors_${\vec{r}_{d,l}}^{\,0}$_have components such that $N_1=N_2=0$ in Eqs. (\[r-r0-T\]) and (\[xd-xd\]). Therefore, the primary links_have_minimum length_$\xi_{{\vec{r}}_{d,l}^{\,0}}=a_s$._Alike_the remaining links of its_family, the centre of a_primary link_is located at $\vec{r}=\vec{r}_{j}+{\vec{r}_{d,l}}^{\,0}$. For the square lattice_there_are two horizontal_primary links whose centers are located at $\vec{r}_{j}+\vec{r}_{1,l}^{\,0}$ with_$l=\pm 1$ and two vertical primary_links whose centers_are_located_at $\vec{r}_{j}+\vec{r}_{2,l}^{\,0}$ with $l=\pm_1$. In the case of the_1D lattice there are two primary_links whose centers are located at $\vec{r}_{j}+\vec{r}_{1,l}^{\,0}$_with $l=\pm 1$.
Partitions and $g$-primary partitions
-------------------------------------
The_building blocks of the $N_{s1}^h=0$_configuration state_are singlet pairs of spinons_on sites $\vec{r}_j^{\,-}$_and $\vec{r}_j^{\,+}$_of the spin_effective lattice, $$\begin{aligned}
\vert\vec{r}_j^{\,-},\vec{r}_j^{\,+}\rangle
& = & {1\over_\sqrt{2}}\left(\vert\uparrow_{\vec{r}_j^{\,-}}
\downarrow_{\vec{r}_j^{\,+}}\rangle -
\vert\downarrow_{\vec{r}_j^{\,-}}
\uparrow_{\vec{r}_j^{\,+}}\rangle\right) \,,
\nonumber_\\
\vec{r}_j^{\,\mp} & = & \vec{r}_j +_\vec{r}_{d,l}^{\,0}
\mp_\vec{r}_{d,l}^{\,g}
\, ;_\hspace{0.5cm}
d_=_d (j) |
\]) below. This result is very much in the spirit of [@SS Theorem 1], in that it assumes a priori smallness of the (genuine) singular set (hypothesis (ii)).
In the following, $h$ denotes the generalized mean curvature of $V$, which is assumed to be in $L^p(\|V\|)$ for $p>n$; the excess $E_{\rho}$ that appears in (vi) is the one used in [@BW]. (See [@BW Section 1.4] for a discussion regarding the exponent (namely $1/p$ as opposed to $2/p$), that appears in the term involving the mean curvature).
\[thm:SSsheeting\] Given $q$ a positive integer, $p > n$ and numbers $\mu$, $\mu_{1}$, $\rho_{0}$, there exists $\eps = \eps(n, p, q, \mu, \mu_{1}\rho_{0}) \in (0, 1/2)$ such that the following holds: Let $\widetilde{F} \, : \, B_{\rho_{0}}^{n+1}(0) \times {\mathbb R}^{n+1} \setminus \{0\} \to {\mathbb R}$ be such that $\widetilde{F}( \cdot, p) \in C^{2, \alpha}(B_{\rho_{0}}^{n+1}(0))$ and $\widetilde{F}(x, \cdot) \in C^{3}({\mathbb R}^{n+1} \setminus \{0\})$. For constant vector $\vec{c}$, let $\widetilde{F}_{\vec{c}}(x, p) = \widetilde{F}(x, p) + \vec{c} \cdot p$. Suppose that for each $y \in B_{\rho_{0}/2}^{n+1}(0)$, there is a vector $\vec{c}(y) \in {\mathbb R}^{n+1}$ and a $C^{2}$ diffeomorphism $\eta_{y} \, : \, B^{n+1}_{\rho_{0}}(0) \to B^{n+1}_{\rho_{0}}(0)$ with $\eta_{y}(0) = y$ and $\eta_{0} = identity$ such that $\eta_{y}^{\#} \widetilde{F}_{\vec{c}(y)}$ satisfies [@SS (1.2)-(1.5)] taken with $\eta_{y}^{\ | \ ]) below. This result is very much in the spirit of [ @SS Theorem 1 ], in that it assume a priori pettiness of the (genuine) singular specify (guess (ii) ).
In the following, $ h$ denotes the generalized beggarly curvature of $ V$, which is assumed to be in $ L^p(\|V\|)$ for $ p > n$; the excess $ E_{\rho}$ that look in (vi) is the one used in [ @BW ]. (See [ @BW incision 1.4 ] for a discussion regarding the exponent (namely $ 1 / p$ as pit to $ 2 / p$), that appears in the condition involving the mean curvature).
\[thm: SSsheeting\ ] contribute $ q$ a positive integer, $ p > n$ and numbers $ \mu$, $ \mu_{1}$, $ \rho_{0}$, there exist $ \eps = \eps(n, p, q, \mu, \mu_{1}\rho_{0 }) \in (0, 1/2)$ such that the following holds: Let $ \widetilde{F } \, : \, B_{\rho_{0}}^{n+1}(0) \times { \mathbb R}^{n+1 } \setminus \{0\ } \to { \mathbb R}$ be such that $ \widetilde{F } (\cdot, p) \in C^{2, \alpha}(B_{\rho_{0}}^{n+1}(0))$ and $ \widetilde{F}(x, \cdot) \in C^{3}({\mathbb R}^{n+1 } \setminus \{0\})$. For constant vector $ \vec{c}$, permit $ \widetilde{F}_{\vec{c}}(x, p) = \widetilde{F}(x, p) + \vec{c } \cdot p$. Suppose that for each $ y \in B_{\rho_{0}/2}^{n+1}(0)$, there is a vector $ \vec{c}(y) \in { \mathbb R}^{n+1}$ and a $ C^{2}$ diffeomorphism $ \eta_{y } \, : \, B^{n+1}_{\rho_{0}}(0) \to B^{n+1}_{\rho_{0}}(0)$ with $ \eta_{y}(0) = y$ and $ \eta_{0 } = identity$ such that $ \eta_{y}^{\ # } \widetilde{F}_{\vec{c}(y)}$ satisfies [ @SS (1.2)-(1.5) ] taken with $ \eta_{y}^{\ | \]) bepow. This result is very ouch in the spitir of [@SV Theodem 1], in ghat it assumes a priori smaplbess if the (genuine) singulaf set (hyplthesis (ui)).
In rhe followmhg, $h$ dekjtes bhe gznxralized mean corvature of $E$, which is assgmdd to be in $L^p(\|V\|)$ for $p>n$; the excess $E_{\rhj}$ that spoears in (vi) is the jne hsed in [@BW]. (See [@BW Section 1.4] for a djscussimn regarding yhe exponent (namely $1/p$ as ooposfd to $2/p$), that appeags in the tgdm ybvolving the mean curvature).
\[thm:SSshgeting\] Given $q$ a positive integef, $p > u$ and numbets $\mk$, $\mu_{1}$, $\rho_{0}$, thece exifts $\eps = \eps(k, p, q, \mg, \mu_{1}\rho_{0}) \in (0, 1/2)$ such thab the foolowing holds: Let $\widxtilde{F} \, : \, B_{\rho_{0}}^{n+1}(0) \timgs {\mathbb S}^{n+1} \setminus \{0\} \to {\matybv R}$ bg sucv thxr $\wkdeuilve{F}( \cdot, o) \ii C^{2, \alpha}(B_{\rgo_{0}}^{n+1}(0))$ and $\widwtilde{F}(x, \cdot) \in C^{3}({\mstrvb R}^{n+1} \setminua \{0\})$. For cjnstant vector $\vec{c}$, let $\widetilde{F}_{\vec{c}}(x, p) = \widetilde{F}(x, p) + \vec{c} \cdit p$. Suppose that for each $y \ig B_{\rho_{0}/2}^{n+1}(0)$, there is a vector $\vec{c}(y) \in {\mathbb R}^{n+1}$ and d $C^{2}$ dmfweonovphirn $\fta_{y} \, : \, B^{n+1}_{\rho_{0}}(0) \to B^{n+1}_{\rho_{0}}(0)$ with $\eta_{y}(0) = y$ and $\eta_{0} = ysemtpty$ such that $\eta_{n}^{\#} \widetilde{F}_{\vec{c}(y)}$ xahixsies [@SS (1.2)-(1.5)] takev with $\etz_{y}^{\ | \]) below. This result is very much spirit [@SS Theorem in that it the singular set (hypothesis In the following, denotes the generalized mean curvature of which is assumed to be in $L^p(\|V\|)$ for $p>n$; the excess $E_{\rho}$ that in (vi) is the one used in [@BW]. (See [@BW Section 1.4] for discussion the (namely as opposed to $2/p$), that appears in the term involving the mean curvature). \[thm:SSsheeting\] Given $q$ positive integer, $p > n$ and numbers $\mu$, $\rho_{0}$, there exists $\eps \eps(n, p, q, \mu, \mu_{1}\rho_{0}) (0, such that following Let \, : \, \times {\mathbb R}^{n+1} \setminus \{0\} \to {\mathbb R}$ be such that $\widetilde{F}( \cdot, p) \in C^{2, \alpha}(B_{\rho_{0}}^{n+1}(0))$ $\widetilde{F}(x, \cdot) R}^{n+1} \setminus For vector let $\widetilde{F}_{\vec{c}}(x, p) p) + \vec{c} \cdot p$. Suppose $y \in B_{\rho_{0}/2}^{n+1}(0)$, there is a vector $\vec{c}(y) {\mathbb R}^{n+1}$ a $C^{2}$ diffeomorphism $\eta_{y} \, : B^{n+1}_{\rho_{0}}(0) \to B^{n+1}_{\rho_{0}}(0)$ with $\eta_{y}(0) = y$ and = identity$ such that $\eta_{y}^{\#} \widetilde{F}_{\vec{c}(y)}$ satisfies [@SS (1.2)-(1.5)] taken with $\eta_{y}^{\ | \]) below. This result is very much In the spiriT of [@SS theOreM 1], iN thaT it aSsumes a priori sMAllnEss of the (genuine) singulaR set (hYpOThesIS (iI)).
In thE followINg, $H$ DEnoTeS tHe gEnERaLized MeaN curvatUre of $V$, whicH is AsSumed to be in $L^P(\|v\|)$ fOr $p>n$; the excEss $e_{\rho}$ that appeArs In (vi) is ThE onE Used iN [@BW]. (see [@BW sectioN 1.4] For a diScussion rEgARding tHE exponeNT (NaMely $1/P$ as opposed to $2/p$), that APpEArs in the term inVolvinG tHE mEAN cuRvaTure).
\[thm:SSsHeEting\] gIven $q$ a pOSiTIVE inTEger, $p > n$ and numbErs $\mu$, $\mu_{1}$, $\rho_{0}$, tHEre Exists $\EpS = \epS(N, p, q, \mu, \mU_{1}\rho_{0}) \iN (0, 1/2)$ sUCh tHat the folloWing Holds: Let $\wIdetilDE{F} \, : \, B_{\rho_{0}}^{n+1}(0) \TImes {\matHbb R}^{n+1} \sEtmInuS \{0\} \to {\mAThBb r}$ be SuCH thAT $\wIdeTIldE{F}( \cdot, p) \iN C^{2, \AlPha}(B_{\rHo_{0}}^{n+1}(0))$ aND $\WIDetiLde{f}(x, \cdOt) \in C^{3}({\Mathbb R}^{n+1} \setmiNus \{0\})$. for cONstAnt veCtor $\vEc{c}$, lEt $\WidetIlde{F}_{\vEc{c}}(x, p) = \WiDetilde{F}(x, p) + \vec{c} \cDot p$. suppose thAt fOr EacH $y \In B_{\rhO_{0}/2}^{N+1}(0)$, there Is a VecTor $\vec{c}(Y) \in {\mathBB R}^{n+1}$ AnD A $c^{2}$ DiFfeomorphism $\eta_{y} \, : \, B^{n+1}_{\RhO_{0}}(0) \TO B^{N+1}_{\rho_{0}}(0)$ with $\Eta_{y}(0) = y$ aND $\eTa_{0} = IDentity$ sUcH thAt $\etA_{Y}^{\#} \WidetIlde{f}_{\VeC{c}(y)}$ satisFies [@SS (1.2)-(1.5)] TAkEn With $\eta_{Y}^{\ | \]) below. This result isvery muchin th e s pir it of[@SS Theorem 1], i n tha t it assumes a priorismall ne s s of th e (ge nuine)s in g u lar s et (h yp o th esis(ii )).
In the follo win g, $h$ denotes th e generali zed mean curvat ure of $V $, wh i ch is as sumed to be in $L^ p(\|V\|)$ f o r $p>n $ ; the e x c es s $E _{\rho}$ that app e ar s in (vi) is th e oneus e di n [@ BW] . (See [@B WSecti o n 1.4]f or a dis c ussion regard ing the exp o nen t (nam el y $ 1 /p$ as oppo se d to $2/p$), th at a ppears in the t e rm invo l ving th e mean cu rva ture ) .
\ [th m: S Ssh e et ing \ ] G iven $q$ a p ositi ve i n t e g er,$p> n$ andnumbers $\mu$ , $ \mu_ { 1}$ , $\r ho_{0 }$,th ere e xists$\eps = \eps(n, p, q,\mu, \mu_{1}\ rho _{ 0}) \ in (0 , 1/2)$ su chthat th e follo w ing h o l d s: Let $\widetilde{F }\ , : \, B_{\ rho_{0 } }^ {n + 1}(0) \t im es{\ma t h bb R} ^{n+ 1 }\setminu s \{0\ } \ to {\math bb R}$ b esuc h t hat $ \ wide tilde{ F}( \cdo t, p) \in C^{2, \alp h a}(B_{\rho_{0 } }^ { n +1 } (0)) $ a nd $\wideti lde{ F }(x, \cd o t) \i n C^{3 }({\m at h bb R}^{n+1} \setminus\{ 0\})$. Forconstant vect or $\vec{c } $ , let $\w idet i ld e {F}_{\vec{c}}( x, p) = \wideti l de{F}(x, p) + \vec{c} \cdot p$ . Supposetha t f oreac h $y \in B_{\rho_ { 0 }/2} ^{ n+1}(0) $,there i s a ve cto r $ \v ec{c}(y)\in {\ma th bb R }^ {n+ 1}$ a n d a $C^{ 2} $ d if feo morph i sm $\e ta_{y } \, : \ , B^ {n+1}_{ \ rh o _ {0}} (0 )\toB^{ n+ 1}_{\ rho_ { 0}} (0)$ wi th $\eta_ {y} ( 0) = y $and $\e ta_{0} = iden ti ty$ such t ha t $ \eta_{ y } ^{\#} \w idetilde{F}_{\vec{c}(y) } $ satis fie s [@S S (1 .2)-(1.5) ] t aken w ith $\eta_ {y}^{\ | \]) below._This result_is very much in_the spirit_of_[@SS Theorem_1],_in that it_assumes a priori_smallness of the (genuine)_singular set (hypothesis_(ii)).
In_the following, $h$ denotes the generalized mean curvature of $V$, which is assumed to_be_in $L^p(\|V\|)$_for_$p>n$;_the excess $E_{\rho}$ that appears_in (vi) is the one_used in_[@BW]. (See [@BW Section 1.4] for a discussion_regarding_the exponent (namely_$1/p$ as opposed to $2/p$), that appears in the_term involving the mean curvature).
\[thm:SSsheeting\] Given_$q$ a positive_integer,_$p_> n$ and numbers_$\mu$, $\mu_{1}$, $\rho_{0}$, there exists $\eps_= \eps(n, p, q, \mu, \mu_{1}\rho_{0})_\in (0, 1/2)$ such that the following_holds: Let $\widetilde{F} \, : \,_B_{\rho_{0}}^{n+1}(0) \times {\mathbb R}^{n+1} \setminus_\{0\} \to_{\mathbb R}$ be such that_$\widetilde{F}( \cdot, p)_\in C^{2,_\alpha}(B_{\rho_{0}}^{n+1}(0))$ and $\widetilde{F}(x,_\cdot) \in C^{3}({\mathbb R}^{n+1} \setminus \{0\})$._For constant vector_$\vec{c}$, let $\widetilde{F}_{\vec{c}}(x, p) = \widetilde{F}(x,_p)_+ \vec{c} \cdot_p$._Suppose_that for_each $y \in_B_{\rho_{0}/2}^{n+1}(0)$,_there is_a_vector $\vec{c}(y) \in {\mathbb R}^{n+1}$ and_a_$C^{2}$ diffeomorphism $\eta_{y} \, : \, B^{n+1}_{\rho_{0}}(0)_\to B^{n+1}_{\rho_{0}}(0)$ with $\eta_{y}(0)_=_y$ and $\eta_{0} =_identity$ such that $\eta_{y}^{\#} \widetilde{F}_{\vec{c}(y)}$_satisfies [@SS (1.2)-(1.5)] taken with $\eta_{y}^{\ |
cdot z}\ar[d]^{z} & Z'\ar@{=}[d] \\
\Omega Z'\ar[r]^{x'} & Y'\ar[r]^{y'} & Z\ar[r]^{z'} & Z', }$$ where the third row and the second column are left triangles.
A triangulated category is a left triangulated category.
Let $\mathcal{B}$ be an exact category with enough projectives. Denote by $\mathcal{P}$ the subcategory of $\mathcal{B}$ consisting of projectives. Then the quotient category $\underline{\mathcal{B}}=\mathcal{B}/\mathcal{P}$ is a left triangulated category.
By (LTR0) and (LTR2), we have the following easy lemma.
Let $\Omega Z\xrightarrow{x}X\xrightarrow{y}Y\xrightarrow{z}Z$ be a left triangle, then so is $\Omega Y\xrightarrow{\Omega z}\Omega
Z\xrightarrow{x}X\xrightarrow{-y}Y$.
(cf. \[8\]) Let $\mathcal{C}$ be a left triangulated category. Then for any left triangle $\Omega Z\xrightarrow{x}X\xrightarrow{y}Y\xrightarrow{z}Z$ and any object $U$ of $\mathcal{C}$, there exists an exact sequence
$\cdots\rightarrow$Hom$_{\mathcal{C}}(U,\Omega
Z)\xrightarrow{x_{\ast}}$Hom$_{\mathcal{C}}(U,X)\xrightarrow{y_{\ast}}$ Hom$_{\mathcal{C}}(U,Y)\xrightarrow{z_{\ast}}$Hom$_{\mathcal{C}}(U,Z)$.
Let $\mathcal{C}$ be a left triangulated category. A subcategory $\mathcal{M}$ of $\mathcal{C}$ is called a rigid subcategory if Hom$_{\mathcal{C}}(\Omega\mathcal{M},\mathcal{M})=0$.
Let $\mathcal{M}$ be a rigid subcategory of $\mathcal{C}$. Denote by $\Omega\mathcal{M}\ast\mathcal{M}$ the subcategory of objects $X$ in $\mathcal{C}$ such that there exist left triangles $\Omega
M_{1}\rightarrow X\rightarrow M_{0}\rightarrow M_{1}$, where $M_{0},M_{1}\in\mathcal{M}$. Now we consider the functor $H:\Omega\mathcal{M | cdot z}\ar[d]^{z } & Z'\ar@{=}[d ] \\
\Omega Z'\ar[r]^{x' } & Y'\ar[r]^{y' } & Z\ar[r]^{z' } & Z', } $ $ where the third row and the second column are left triangles.
A triangulated class is a leftover triangulate category.
Let $ \mathcal{B}$ be an accurate category with enough projectives. Denote by $ \mathcal{P}$ the subcategory of $ \mathcal{B}$ dwell of projectives. Then the quotient category $ \underline{\mathcal{B}}=\mathcal{B}/\mathcal{P}$ is a left triangulate category.
By (LTR0) and (LTR2), we have the following comfortable lemma.
Let $ \Omega Z\xrightarrow{x}X\xrightarrow{y}Y\xrightarrow{z}Z$ be a left triangle, then therefore is $ \Omega Y\xrightarrow{\Omega z}\Omega
Z\xrightarrow{x}X\xrightarrow{-y}Y$.
(cf. \[8\ ]) Let $ \mathcal{C}$ be a left triangulated category. Then for any leftover triangle $ \Omega Z\xrightarrow{x}X\xrightarrow{y}Y\xrightarrow{z}Z$ and any object $ U$ of $ \mathcal{C}$, there exists an accurate sequence
$ \cdots\rightarrow$Hom$_{\mathcal{C}}(U,\Omega
Z)\xrightarrow{x_{\ast}}$Hom$_{\mathcal{C}}(U, X)\xrightarrow{y_{\ast}}$ Hom$_{\mathcal{C}}(U, Y)\xrightarrow{z_{\ast}}$Hom$_{\mathcal{C}}(U, Z)$.
Let $ \mathcal{C}$ be a left triangulated category. A subcategory $ \mathcal{M}$ of $ \mathcal{C}$ is call a rigid subcategory if Hom$_{\mathcal{C}}(\Omega\mathcal{M},\mathcal{M})=0$.
Let $ \mathcal{M}$ be a rigid subcategory of $ \mathcal{C}$. Denote by $ \Omega\mathcal{M}\ast\mathcal{M}$ the subcategory of objects $ X$ in $ \mathcal{C}$ such that there exist left triangles $ \Omega
M_{1}\rightarrow X\rightarrow M_{0}\rightarrow M_{1}$, where $ M_{0},M_{1}\in\mathcal{M}$. nowadays we consider the functor $ H:\Omega\mathcal{M | cdoh z}\ar[d]^{z} & Z'\ar@{=}[d] \\
\Omega E'\ae[r]^{x'} & Y'\ar[r]^{y'} & Z\ar[r]^{z'} & Z', }$$ where the third rlw and uke second column are left triwngles.
A rriaigulated categorb is a lcyt trjwnguneted category.
Lej $\mathcal{B}$ ba an exact catagurv with enough projectives. Denote by $\iathcal{L}$ hhe subcategori of $\kwthczl{B}$ consisting of projectives. Then the qumtient categoty $\underline{\mathcal{B}}=\mathcap{B}/\mahhcal{P}$ is a left tgiangulated catqtory.
By (LTR0) avd (LTR2), we have the folmowing easy lemma.
Let $\Omega Z\xriehtarxow{x}X\xrightqreow{j}I\xrightarrow{v}Z$ be w left triannke, thet so is $\Omega Y\xrightsrrmw{\Onega z}\Omega
Z\xrightarrmw{x}X\xrightarrow{-y}Y$.
(cs. \[8\]) Let $\madheal{C}$ be a left triangylqted watecory. Rhev fkr ahy lefh tciangle $\Omefa Z\xrightaerow{x}X\xrightarrow{y}Y\crythtarrow{z}Z$ ans any jbtect $U$ of $\mathcal{C}$, there exists an exacu seqhence
$\cdots\rightarrow$Hom$_{\nathcal{C}}(U,\Omega
Z)\xrightwrrow{x_{\ast}}$Rom$_{\mathcal{C}}(U,X)\xrightarrow{y_{\ast}}$ Hom$_{\mathcal{C}}(U,Y)\xrightasrow{z_{\esg}}$Hon$_{\mwgycwl{C}}(U,Z)$.
Let $\mathcal{C}$ be a left triangulated catedkru. S subcategory $\iathcal{M}$ of $\mwtnsal{C}$ is callea a riyjd subcategory if Hol$_{\mathcaj{C}}(\Ometa\mathcal{I},\matncal{M})=0$.
Let $\mathcal{M}$ be a rigie subcategorj of $\mathcal{C}$. Denote bv $\Omega\mathccl{M}\ast\kathcsl{M}$ the subcategory of ubjedts $X$ in $\mahhcal{C}$ sudf that there exirt kext triangles $\Omega
M_{1}\rightarwow X\righvarroc M_{0}\rightxrroe M_{1}$, whqre $M_{0},M_{1}\in\mahhcal{M}$. Now we consider tje fuuctor $H:\Omega\matjcal{M | cdot z}\ar[d]^{z} & Z'\ar@{=}[d] \\ \Omega Z'\ar[r]^{x'} & & Z', where the third are triangles. A triangulated is a left category. Let $\mathcal{B}$ be an exact with enough projectives. Denote by $\mathcal{P}$ the subcategory of $\mathcal{B}$ consisting of projectives. the quotient category $\underline{\mathcal{B}}=\mathcal{B}/\mathcal{P}$ is a left triangulated category. By (LTR0) and (LTR2), have following lemma. $\Omega Z\xrightarrow{x}X\xrightarrow{y}Y\xrightarrow{z}Z$ be a left triangle, then so is $\Omega Y\xrightarrow{\Omega z}\Omega Z\xrightarrow{x}X\xrightarrow{-y}Y$. (cf. \[8\]) Let be a left triangulated category. Then for any triangle $\Omega Z\xrightarrow{x}X\xrightarrow{y}Y\xrightarrow{z}Z$ and object $U$ of $\mathcal{C}$, there an sequence $\cdots\rightarrow$Hom$_{\mathcal{C}}(U,\Omega Hom$_{\mathcal{C}}(U,Y)\xrightarrow{z_{\ast}}$Hom$_{\mathcal{C}}(U,Z)$. $\mathcal{C}$ a left triangulated A subcategory $\mathcal{M}$ of $\mathcal{C}$ is called a rigid subcategory if Hom$_{\mathcal{C}}(\Omega\mathcal{M},\mathcal{M})=0$. Let $\mathcal{M}$ be a rigid of $\mathcal{C}$. $\Omega\mathcal{M}\ast\mathcal{M}$ the of $X$ $\mathcal{C}$ such that left triangles $\Omega M_{1}\rightarrow X\rightarrow M_{0}\rightarrow Now we consider the functor $H:\Omega\mathcal{M | cdot z}\ar[d]^{z} & Z'\ar@{=}[d] \\
\Omega Z'\ar[r]^{x'} & Y'\aR[r]^{y'} & Z\ar[r]^{z'} & Z', }$$ wHere tHe tHirD rOw anD the Second column arE Left Triangles.
A triangulated CategOrY Is a lEFt TrianGulated CAtEGOry.
leT $\mAthCaL{b}$ bE an exAct CategorY with enougH prOjEctives. DenotE By $\Mathcal{P}$ thE suBcategory of $\mAthCal{B}$ coNsIstINg of pRojEctivEs. Then THe quotIent categOrY $\UnderlINe{\mathcAL{b}}=\mAthcAl{B}/\mathcal{P}$ is a lefT TrIAngulated categOry.
By (LtR0) ANd (ltr2), we HavE the followInG easy LEmma.
Let $\oMeGA z\XriGHtarrow{x}X\xrigHtarrow{y}Y\xrIGhtArrow{z}z$ bE a lEFt triaNgle, tHeN So iS $\Omega Y\xrigHtarRow{\Omega z}\omega
Z\XRightarROw{x}X\xriGhtarrOw{-y}y$.
(cf. \[8\]) let $\mAThCaL{C}$ bE a LEft TRiAngULatEd categoRy. thEn for Any lEFT TRianGle $\omegA Z\xriGhtarrow{x}X\xriGhtArroW{Y}Y\xRightArrow{Z}Z$ anD aNy objEct $U$ of $\MathcAl{c}$, there exists an eXact Sequence
$\cDotS\rIghTaRrow$HOM$_{\mathcAl{C}}(u,\OmEga
Z)\xriGhtarroW{X_{\asT}}$HOM$_{\MAtHcal{C}}(U,X)\xrightarrow{Y_{\aST}}$ hoM$_{\mathcal{c}}(U,Y)\xriGHtArROw{z_{\ast}}$HoM$_{\mAthCal{C}}(u,z)$.
let $\maThcaL{c}$ bE a left trIangulATeD cAtegory. a sUbcateGoRy $\mAthCal{M}$ oF $\MathCal{C}$ is Called a rIgid sUBcategory if Hom$_{\MAthcal{C}}(\Omega\mAThCAL{M},\MAthcAl{M})=0$.
let $\mathcal{M}$ Be a rIGid sUbcaTEgOry OF $\mathCal{C}$. DEnOTe BY $\Omega\mathcal{M}\ast\maThCal{M}$ thE subcAtegory of objeCts $X$ in $\mathCAL{c}$ such thaT theRE eXIst left trianglEs $\OmeGa
M_{1}\rightarROw X\rightArrow m_{0}\rightarRow M_{1}$, where $m_{0},m_{1}\In\mathcaL{M}$. NOw wE coNsiDER tHe functor $H:\OmeGA\MathCaL{M | cdot z}\ar[d]^{z} & Z'\ar@ {=}[d] \ \
\O mega Z'\ar[r]^ { x'} & Y'\ar[r]^{y '} &Z\ a r[r] ^ {z '} &Z', }$$ wh e r e t he t hir dr ow andthe second column ar e l ef t triangles.
A triangula ted category is aleft t ri ang u lated ca tegor y.
Le t $\mat hcal{B}$be an exa c t categ o r ywith enough projectiv e s. Denote by $\ma thcal{ P} $ t h e su bca tegory of$\ mathc a l{B}$ c o ns i s t ing of projective s. Then the quo tientca teg o ry $\u nderl in e {\m athcal{B}}= \mat hcal{B}/\ mathca l {P}$ is a lefttriang ula ted cat e go ry .
By (LT R 0) an d (L TR2), we h av e the fol l o w i ng e asy lem ma.
Let $\Omega Z \xr ight a rro w{x}X \xrig htar ro w{y}Y \xrigh tarro w{ z}Z$ be a lefttria ngle, the n s ois$\ Omega Y\xrig hta rro w{\Omeg a z}\Om e gaZ\ x r i gh tarrow{x}X\xrighta rr o w {- y}Y$.
( cf. \[ 8 \] )L et $\mat hc al{ C}$b e a le ft t r ia ngulated categ o ry .Then fo rany le ft tr ian gle $ \ Omeg a Z\xr ightarro w{x}X \ xrightarrow{y} Y \xrightarrow{ z }Z $ an d any ob ject $U$ of $\m a thca l{C} $ ,the r e exi sts a ne xa c t sequence
$\cdots \r ightar row$H om$_{\mathcal {C}}(U,\Om e g a
Z)\xrig htar r ow { x_{\ast}}$Hom$ _{\ma thcal{C}}( U ,X)\xrig htarr ow{y_{\a st}}$ Hom $ _ {\mathca l{C }}( U,Y )\x r i gh tarrow{z_{\as t } }$Ho m$ _{\math cal {C}}(U, Z)$ .
Let $\ ma thcal{C}$ be a le ft t ri an gul atedc ategory. A su bc ate gory$ \mathc al{M} $ of $ \m a thc al{C}$i sc a lled a r igid su bc atego ry i f Ho m$_{\ma thcal{C}} (\O m ega\ ma th cal{M}, \mathcal{M})= 0$ .
Let $\m at hca l{M}$b e a rigid subcategory of $\mathc a l{C}$.Den ote b y $\ Omega\mat hca l{M}\a st\ m athcal {M}$ t he su bc ate g o ry of o bj ect s$X$ in $\m a t hca l{C}$ s uchthat th ere exist left tri a ngl es $\Omega
M_ {1} \rig h t ar row X\ r igh ta r row M _{0}\rightarrow M_{1}$, w he r e$M_{0},M_{ 1 }\i n\ mathcal {M}$. N ow we conside r the fun ctor $H:\ Om ega\ m a thc al{M | cdot z}\ar[d]^{z}_& Z'\ar@{=}[d]_ _ __ __ _ _ \\
\Omega_Z'\ar[r]^{x'} __ & Y'\ar[r]^{y'} & Z\ar[r]^{z'} & Z', }$$ where_the_third row_and_the_second column are left triangles.
A_triangulated category is a left_triangulated category.
Let_$\mathcal{B}$ be an exact category with enough projectives._Denote_by $\mathcal{P}$ the_subcategory of $\mathcal{B}$ consisting of projectives. Then the quotient_category $\underline{\mathcal{B}}=\mathcal{B}/\mathcal{P}$ is a left triangulated_category.
By (LTR0) and_(LTR2),_we_have the following easy_lemma.
Let $\Omega Z\xrightarrow{x}X\xrightarrow{y}Y\xrightarrow{z}Z$ be a left_triangle, then so is $\Omega Y\xrightarrow{\Omega_z}\Omega
Z\xrightarrow{x}X\xrightarrow{-y}Y$.
(cf. \[8\]) Let $\mathcal{C}$ be a left_triangulated category. Then for any left_triangle $\Omega Z\xrightarrow{x}X\xrightarrow{y}Y\xrightarrow{z}Z$ and any_object $U$_of $\mathcal{C}$, there exists an_exact sequence
$\cdots\rightarrow$Hom$_{\mathcal{C}}(U,\Omega
Z)\xrightarrow{x_{\ast}}$Hom$_{\mathcal{C}}(U,X)\xrightarrow{y_{\ast}}$ Hom$_{\mathcal{C}}(U,Y)\xrightarrow{z_{\ast}}$Hom$_{\mathcal{C}}(U,Z)$.
Let_$\mathcal{C}$ be_a left triangulated_category. A subcategory $\mathcal{M}$ of $\mathcal{C}$_is called a_rigid subcategory if Hom$_{\mathcal{C}}(\Omega\mathcal{M},\mathcal{M})=0$.
Let $\mathcal{M}$ be_a_rigid subcategory of_$\mathcal{C}$._Denote_by $\Omega\mathcal{M}\ast\mathcal{M}$_the subcategory of_objects_$X$ in_$\mathcal{C}$_such that there exist left triangles_$\Omega
M_{1}\rightarrow_X\rightarrow M_{0}\rightarrow M_{1}$, where $M_{0},M_{1}\in\mathcal{M}$. Now we_consider the functor $H:\Omega\mathcal{M |
} }}{{r^{4}}}\sin4\theta -....$$
The vortex components of the right pair are arranged symmetrically about the axis *Ox* in the points (*x; y*) and (*x;-y*), and the intrinsic (due to the interaction of vortices and not related to the presence of the background flow) velocity of the vortex in the 1-st quadrant is
$$\overrightarrow {V} _{sing} = \frac{\Gamma }{4\pi }\left\{
\frac{1}{y} - \frac{y}{x^2 + y^2};\,\,\frac{x}{x^2 + y^2} -
\frac{1}{x} \right\}.$$
In the case of only one vortex pair without a background flow that corresponds to the known expression
$$\label{eq4} \overrightarrow {V} _{s} = \left( {\Gamma /4\pi y;0}
\right),$$
where $\Gamma $ is the intensity of the vortex, *x* and *y* are its abscissa and ordinate.
The dynamics of two pairs of vortices in the flow is described by the equations
$$\label{eq5} \dot {x} = \frac{\Gamma }{4\pi }\left( \frac{1}{y} -
\frac{y}{x^2 + y^2} \right) + \frac{\partial \psi _{reg}
}{\partial y};\,\,\dot {y} = \left( \frac{x}{x^2 + y^2} -
\frac{1}{x} \right) - \frac{\partial \psi _{reg}} {\partial x}$$
and the Hamiltonian of the system of 4 vortices is reduced to the form
$$\label{eq6} H = \frac{\Gamma }{4\pi }\ln{\frac{xy}{\sqrt {x^2 +
y^2}}} + C_0 \arctan\frac{y}{x} - C_2 \frac{2xy}{\left( x^2 + y^2
\right)^2} -....$$
Accordingly, the equations (5) can be represented as
$$\dot{x}= \frac{\partial H}{\partial y}, \quad \dot {y} = - \frac
{\partial H}{\partial x | } } } { { r^{4}}}\sin4\theta -.... $$
The vortex components of the right pair are stage symmetrically about the bloc * Ox * in the points (* x; y *) and (* x;-y *), and the intrinsic (due to the interaction of whirl and not refer to the presence of the background stream) velocity of the vortex in the 1 - st quadrant is
$ $ \overrightarrow { volt } _ { sing } = \frac{\Gamma } { 4\pi } \left\ {
\frac{1}{y } - \frac{y}{x^2 + y^2};\,\,\frac{x}{x^2 + y^2 } -
\frac{1}{x } \right\}.$$
In the case of entirely one vortex pair without a backdrop menstruation that corresponds to the known expression
$ $ \label{eq4 } \overrightarrow { V } _ { s } = \left ({ \Gamma /4\pi y;0 }
\right),$$
where $ \Gamma $ is the intensity of the vortex, * x * and * y * are its abscissa and ordinate.
The moral force of two pairs of vortices in the flow is report by the equations
$ $ \label{eq5 } \dot { x } = \frac{\Gamma } { 4\pi } \left (\frac{1}{y } -
\frac{y}{x^2 + y^2 } \right) + \frac{\partial \psi _ { reg }
} { \partial y};\,\,\dot { y } = \left (\frac{x}{x^2 + y^2 } -
\frac{1}{x } \right) - \frac{\partial \psi _ { reg } } { \partial x}$$
and the Hamiltonian of the system of 4 whirl is reduced to the kind
$ $ \label{eq6 } H = \frac{\Gamma } { 4\pi } \ln{\frac{xy}{\sqrt { x^2 +
y^2 } } } + C_0 \arctan\frac{y}{x } - C_2 \frac{2xy}{\left (x^2 + y^2
\right)^2 } -.... $$
Accordingly, the equations (5) can be represented as
$ $ \dot{x}= \frac{\partial H}{\partial y }, \quad \dot { y } = - \frac
{ \partial H}{\partial x | } }}{{r^{4}}}\sln4\theta -....$$
The vortex compokents of the rigkr pair are adranged rymmetrically about the axis *Oz* in uke points (*x; y*) and (*x;-y*), and the pntrinsic (due ro the intxdaction of vodbices end not related to the prasence of the tazkyround flow) velocity of the vortex ig the 1-sy euadrant is
$$\ovetrighuarwow {B} _{sing} = \frac{\Gamma }{4\pi }\left\{
\frac{1}{y} - \frzc{y}{x^2 + y^2};\,\,\hrac{x}{x^2 + y^2} -
\frac{1}{x} \right\}.$$
In the case of only lne gortex pair withouh a backgroohd soow that corfesponds to the known gxpression
$$\label{eq4} \overrightarrow {V} _{s} = \left( {\Gammq /4\pi j;0}
\tight),$$
where $\Gemma $ ps the intenslny of tve vortrx, *x* and *y* are itv avscissa and ordinate.
Tie dynamics of two pwirs of vmrcices in the flow is eewcribgd by the wquxtiknx
$$\lzbel{eq5} \dov {x} = \frac{\Gajma }{4\pi }\left( \frac{1}{y} -
\frac{y}{x^2 + y^2} \ribhe) + \frac{\partial \psi _{rqg}
}{\[artial y};\,\,\dot {y} = \left( \frac{x}{x^2 + y^2} -
\frac{1}{x} \richt) - \frac{\partial \psi _{reg}} {\pqrtial x}$$
and the Hamiljonian of ehe system of 4 vortices is reduced to the form
$$\labal{eq6} I = \frcg{\Nammx }{4\pl }\ln{\frac{xy}{\sqrt {x^2 +
y^2}}} + C_0 \arctan\frac{y}{x} - C_2 \frac{2xy}{\jsfu( x^2 + y^2
\right)^2} -....$$
Accorbingly, the equatoojs (5) can be reprerented as
$$\sot{x}= \frac{\partial H}{\oartial y}, \quqd \dot {y} = - \frsc
{\partial H}{\partial x | } }}{{r^{4}}}\sin4\theta -....$$ The vortex components of pair arranged symmetrically the axis *Ox* and and the intrinsic to the interaction vortices and not related to the of the background flow) velocity of the vortex in the 1-st quadrant is {V} _{sing} = \frac{\Gamma }{4\pi }\left\{ \frac{1}{y} - \frac{y}{x^2 + y^2};\,\,\frac{x}{x^2 + y^2} \frac{1}{x} In case only one vortex pair without a background flow that corresponds to the known expression $$\label{eq4} \overrightarrow _{s} = \left( {\Gamma /4\pi y;0} \right),$$ where $ is the intensity the vortex, *x* and *y* its and ordinate. dynamics two of vortices in flow is described by the equations $$\label{eq5} \dot {x} = \frac{\Gamma }{4\pi }\left( \frac{1}{y} - \frac{y}{x^2 + \right) + _{reg} }{\partial {y} \left( + y^2} - - \frac{\partial \psi _{reg}} {\partial x}$$ of the system of 4 vortices is reduced the form H = \frac{\Gamma }{4\pi }\ln{\frac{xy}{\sqrt {x^2 y^2}}} + C_0 \arctan\frac{y}{x} - C_2 \frac{2xy}{\left( x^2 y^2 \right)^2} -....$$ Accordingly, the equations (5) can be represented as $$\dot{x}= \frac{\partial H}{\partial y}, {y} = - \frac H}{\partial x | } }}{{r^{4}}}\sin4\theta -....$$
The vortex componeNts of the riGht paIr aRe aRrAngeD symMetrically abouT The aXis *Ox* in the points (*x; y*) and (*x;-Y*), and tHe INtriNSiC (due tO the intERaCTIon Of VoRtiCeS AnD not rElaTed to thE presence oF thE bAckground floW) VeLocity of thE voRtex in the 1-st qUadRant is
$$\OvErrIGhtarRow {v} _{sing} = \Frac{\GaMMa }{4\pi }\leFt\{
\frac{1}{y} - \frAc{Y}{X^2 + y^2};\,\,\frac{X}{X^2 + y^2} -
\frac{1}{x} \RIGhT\}.$$
In tHe case of only one voRTeX Pair without a baCkgrouNd FLoW THat CorResponds to ThE knowN ExpressIOn
$$\LABEl{eQ4} \OverrightarroW {V} _{s} = \left( {\GammA /4\Pi y;0}
\Right),$$
wHeRe $\GAMma $ is tHe intEnSIty Of the vortex, *X* and *Y* are its abScissa ANd ordinATe.
The dyNamics Of tWo pAirs OF vOrTicEs IN thE FlOw iS DesCribed by ThE eQuatiOns
$$\lABEL{Eq5} \doT {x} = \fRac{\GAmma }{4\pI }\left( \frac{1}{y} -
\fraC{y}{x^2 + Y^2} \rigHT) + \frAc{\parTial \pSi _{reG}
}{\pArtiaL y};\,\,\dot {y} = \Left( \fRaC{x}{x^2 + y^2} -
\frac{1}{x} \right) - \fRac{\pArtial \psi _{Reg}} {\PaRtiAl X}$$
and tHE HamilTonIan Of the syStem of 4 vORtiCeS IS ReDuced to the form
$$\labeL{eQ6} h = \FrAc{\Gamma }{4\pI }\ln{\fraC{Xy}{\SqRT {x^2 +
y^2}}} + C_0 \arctAn\FraC{y}{x} - C_2 \FRAc{2xy}{\lEft( x^2 + Y^2
\RiGht)^2} -....$$
AccorDingly, THe EqUations (5) CaN be repReSenTed As
$$\dot{X}= \Frac{\PartiaL H}{\partiaL y}, \quaD \Dot {y} = - \frac
{\partiaL h}{\partial x | } }}{{r^{4}}}\sin4\theta - ....$$
Th e vor tex co mp onen ts o f the right pa i r ar e arranged symmetrical ly ab ou t the ax is *O x* in t h ep o int s(* x;y* ) a nd (* x;- y*), an d the intr ins ic (due to the in teractionofvortices and no t rela te d t o thepre sence of th e backg round flo w) veloci t y of th e vo rtex in the 1-st quad r an t is
$$\overri ghtarr ow {V } _{s ing } = \frac{ \G amma} {4\pi } \ le f t \ {
\ f rac{1}{y} - \ frac{y}{x^2 + y ^2};\, \, \fr a c{x}{x ^2 +y^ 2 } -
\frac{1}{x } \r ight\}.$$
In t h e caseo f onlyone vo rte x p airw it ho utab ack g ro und flo w that c or re spond s to t h e kno wnexpr essio n
$$\label{e q4} \ov e rri ghtar row { V} _ {s } = \ left({\Gam ma /4\pi y;0}
\ri ght) ,$$
wher e $ \G amm a$ ist he int ens ity of the vortex , *x *a n d * y* are its absciss aa n dordinate .
The dy na m ics of t wo pa irso f vort ices in the flo w is d e sc ri bed byth e equa ti ons
$ $\lab e l{eq 5} \do t {x} =\frac { \Gamma }{4\pi} \left( \frac{ 1 }{ y } - \fra c{y }{x^2 + y^2 } \r i ght) + \ f ra c{\ p artia l \ps i_ {r e g}
}{\partial y};\, \, \dot { y} =\left( \frac{ x}{x^2 + y ^ 2 } -
\frac {1}{ x }\ right) - \frac {\par tial \psi_ {reg}} { \part ial x}$$
and the H amiltoni anofthe sy s t em of 4 vortice s is r ed uced to th e form
$$ \la bel {eq 6} H = \fra c{\Gamma } {4 \p i}\l n{\fr a c{xy}{\s qr t { x^ 2 +
y^2} } } + C _0 \a rcta n\ fr a c{y }{x} -C _2 \ frac {2 xy }{\l eft (x^2 + y^2 \ri ght)^2} -....$$
Ac c ordi ng ly , the e quations (5)ca n be repre se nte d as
$ $ \dot{x}= \frac{\partial H}{\par t ial y}, \q uad \ dot{y} = - \ fra c
{\pa rti a l H}{\ partia l x | } }}{{r^{4}}}\sin4\theta_-....$$
The vortex_components of the right_pair are_arranged_symmetrically about_the_axis *Ox* in_the points (*x;_y*) and (*x;-y*), and_the intrinsic (due_to_the interaction of vortices and not related to the presence of the background flow)_velocity_of the_vortex_in_the 1-st quadrant is
$$\overrightarrow {V}__{sing} = \frac{\Gamma }{4\pi }\left\{
\frac{1}{y}_- \frac{y}{x^2_+ y^2};\,\,\frac{x}{x^2 + y^2} -
\frac{1}{x} \right\}.$$
In the case_of_only one vortex_pair without a background flow that corresponds to the_known expression
$$\label{eq4} \overrightarrow {V} _{s} =_\left( {\Gamma /4\pi_y;0}
\right),$$
where_$\Gamma_$ is the intensity_of the vortex, *x* and *y*_are its abscissa and ordinate.
The dynamics_of two pairs of vortices in the_flow is described by the equations
$$\label{eq5}_\dot {x} = \frac{\Gamma }{4\pi_}\left( \frac{1}{y}_-
\frac{y}{x^2 + y^2} \right) +_\frac{\partial \psi _{reg}
}{\partial_y};\,\,\dot {y}_= \left( \frac{x}{x^2_+ y^2} -
\frac{1}{x} \right) - \frac{\partial_\psi _{reg}} {\partial_x}$$
and the Hamiltonian of the system_of_4 vortices is_reduced_to_the form
$$\label{eq6}_H = \frac{\Gamma_}{4\pi_}\ln{\frac{xy}{\sqrt {x^2_+
y^2}}}_ + C_0 \arctan\frac{y}{x} - C_2_\frac{2xy}{\left(_x^2 + y^2
\right)^2} -....$$
Accordingly, the equations (5)_can be represented as
$$\dot{x}=_\frac{\partial_H}{\partial y}, \quad \dot_{y} = - \frac
{\partial H}{\partial_x |
southern galactic cap ($b \leq
-40^o$), within the declination range $-40^o \leq \delta \leq -2.5^o$. The SSRS2 north contains 1939 galaxies distributed over a solid angle of 0.57 steradians with $\delta \leq 0^o$ with $b \geq +35^o$. As described by da Costa (1997) the galaxy morphologies in this sample used classifications based on other works (Lauberts & Valentijn 1989) as well as those made by the authors.
In this work we also considered a sub-sample of SSRS2 galaxies which have measured colors. For this we used the Lauberts & Valentijn (1989) catalog, so that the subcatalog of galaxies with colors only contains objects south of $\delta$ $\approx$ -17.5$^o$, in both galactic hemispheres. We also imposed a cut at $m_B$ = 14.5 because beyond this magnitude the Lauberts & Valentijn (1989) catalog becomes incomplete, in particular for early type galaxies (Pellegrini 1990$a$). This selection effect is caused by the fact that the Lauberts & Valentijn (1989) catalog is derived from the diameter-limited Lauberts (1982) catalog. The sample limited at $m_B$=14.5 contains 780 galaxies in both galactic hemispheres, of which 694 (89%) have colors. As pointed out by Marzke & da Costa (1997), this sample presents no systematic dependence of the color completeness with magnitude down to the 14.5 limit of this sub-sample.
All galaxy heliocentric velocities, $v_\odot$, have been corrected for the Solar motion with respect to the centroid of the Local Group using $v =v_\odot + 300 sin(l) cos(b)$, where ($l$, $b$) are the galactic coordinates of the galaxy. We also remove from the analyses all galaxies with radial velocities less that 500, as the redshifts for these objects are likely to be dominated by their peculiar velocities. We use throughout $H_o = 100 h$Mpc$^{-1}$.
The Two-Point Correlation Function
==================================
Method
------
The two-point correlation function $\xi (r)$ can be computed from the data using the estimator suggested by Hamilton (1993): $$\xi(r) = {DD(r) RR(r) \over [DR(r | southern galactic cap ($ b \leq
-40^o$), within the decline compass $ -40^o \leq \delta \leq -2.5^o$. The SSRS2 north contains 1939 galaxies spread over a solid angle of 0.57 steradian with $ \delta \leq 0^o$ with $ b \geq +35^o$. As described by da Costa (1997) the galaxy morphologies in this sample use classifications based on early works (Lauberts & Valentijn 1989) as well as those made by the authors.
In this workplace we also considered a sub - sample of SSRS2 galaxies which have measure colors. For this we used the Lauberts & Valentijn (1989) catalog, so that the subcatalog of galaxies with color only contains objects south of $ \delta$ $ \approx$ -17.5$^o$, in both galactic hemisphere. We also imposed a cut at $ m_B$ = 14.5 because beyond this magnitude the Lauberts & Valentijn (1989) catalog becomes incomplete, in especial for early type galaxies (Pellegrini 1990$a$). This selection effect is caused by the fact that the Lauberts & Valentijn (1989) catalog is derived from the diameter - limited Lauberts (1982) catalog. The sample limited at $ m_B$=14.5 contains 780 galaxies in both galactic hemispheres, of which 694 (89 %) have colors. As pointed out by Marzke & da Costa (1997), this sample introduce no systematic dependence of the coloring material completeness with order of magnitude down to the 14.5 terminus ad quem of this sub - sample.
All galaxy heliocentric velocities, $ v_\odot$, have been corrected for the Solar motion with respect to the centroid of the Local Group using $ five = v_\odot + 300 sin(l) cos(b)$, where ($ l$, $ b$) are the galactic coordinates of the galaxy. We also remove from the analyses all galax with radial velocities less that 500, as the redshifts for these objects are likely to be dominate by their curious velocities. We use throughout $ H_o = 100 h$Mpc$^{-1}$.
The Two - Point Correlation Function
= = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = =
Method
------
The two - point correlation function $ \xi (r)$ can be calculate from the data using the estimator suggested by Hamilton (1993 ): $ $ \xi(r) = { DD(r) RR(r) \over [ DR(r | sokthern galactic cap ($b \led
-40^o$), within the dgcoinatimn ranfe $-40^o \leq \delta \leq -2.5^o$. The SSRS2 north rontqins 1939 galaxies distributed uver a sopid anglw of 0.57 steradians with $\dcjta \mcq 0^o$ civh $b \geq +35^o$. As dgscribed by ga Costa (1997) the cauary morphologies in this sample used slassifocwtions based og otnqr wkgkw (Lauberts & Valentijn 1989) as welm as thmse made by tne authors.
In this work we wlso considered a sub-swmple of SSTA2 gwoaxies which have measlxed colors. Ror this we used the Lauberts & Xalencijn (1989) catalig, so jhat the subratalod of galaxies with cmlors omly contains onjectv siuth of $\delta$ $\approx$ -17.5$^m$, in both galactic hemisphesea. We also imposed a cut dt $m_T$ = 14.5 cwcajse bxyohd thid megnitude ths Lauberts & Valentijn (1989) catalog bqbpmes incomplste, in pwrticular for early type galaxies (Pellecrihi 1990$a$). This selection effwct is caused by the vact that the Lauberts & Valentijn (1989) catalog is derived from the vixmeuev-jknihed Lauberts (1982) catalog. The sample limited at $m_F$=14.5 cpnnains 780 galaxies ik both galactic hekidpngres, of which 694 (89%) have comors. As pointed ouh by Matzke & ea Costa (1997), thix sample presents no systemqtic dependeuce of the color comppeteness wich magmitudr down to the 14.5 limit of thia sub-sample.
Wll galaxg heliocentric veuocptiev, $v_\odot$, have been correcteq for the Solcr motiov wijh respqct to the centvmid of the Local Ggoup osing $e =v_\odot + 300 din(l) cos(b)$, where ($l$, $b$) are the galertic coordinajes of the galcxy. We also remove srom the analyxes all galabies with gadial venocities lefs that 500, as tvg redshifts fmr these objwcts are liydly to be domimated by nhtir peculiqr velocities. We uxe jhdoughout $H_o = 100 h$Npc$^{-1}$.
Uhw Two-Point Cortelxtijn Fnnctijt
==================================
Method
------
The tfo-poknt vorreuation funcbiov $\xi (r)$ can be computed fsom fhe data using the eftimator suggestqd by Hamiltom (1993): $$\xi(r) = {DD(r) RR(r) \ovtr [DR(r | southern galactic cap ($b \leq -40^o$), within range \leq \delta -2.5^o$. The SSRS2 over solid angle of steradians with $\delta 0^o$ with $b \geq +35^o$. As by da Costa (1997) the galaxy morphologies in this sample used classifications based other works (Lauberts & Valentijn 1989) as well as those made by the In work also a sub-sample of SSRS2 galaxies which have measured colors. For this we used the Lauberts & (1989) catalog, so that the subcatalog of galaxies colors only contains objects of $\delta$ $\approx$ -17.5$^o$, in galactic We also a at = 14.5 because this magnitude the Lauberts & Valentijn (1989) catalog becomes incomplete, in particular for early type galaxies (Pellegrini This selection caused by fact the & Valentijn (1989) derived from the diameter-limited Lauberts (1982) limited at $m_B$=14.5 contains 780 galaxies in both hemispheres, of 694 (89%) have colors. As pointed by Marzke & da Costa (1997), this sample no systematic dependence of the color completeness with magnitude down to the 14.5 limit of All galaxy heliocentric velocities, have been corrected the motion respect the centroid the Local Group using $v =v_\odot + 300 sin(l) cos(b)$, where $b$) are the galactic coordinates of the galaxy. We also the all galaxies with velocities less that 500, the for these objects are be by We throughout = 100 h$Mpc$^{-1}$. The Correlation Function ================================== Method ------ two-point correlation function $\xi the data using the estimator suggested by Hamilton $$\xi(r) = {DD(r) RR(r) \over [DR(r | southern galactic cap ($b \leq
-40^o$), wIthin the deClinaTioN raNgE $-40^o \leQ \delTa \leq -2.5^o$. The SSRS2 nORth cOntains 1939 galaxies distribUted oVeR A solID aNgle oF 0.57 steradIAnS WIth $\DeLtA \leQ 0^o$ WItH $b \geq +35^O$. As DescribEd by da CostA (1997) thE gAlaxy morpholOGiEs in this saMplE used classifIcaTions bAsEd oN Other WorKs (LauBerts & VALentijN 1989) as well as ThOSe made BY the autHORs.
in thIs work we also consiDErED a sub-sample of SsRS2 galAxIEs WHIch HavE measured cOlOrs. FoR This we uSEd THE lauBErts & Valentijn (1989) Catalog, so thAT thE subcaTaLog OF galaxIes wiTh COloRs only contaIns oBjects souTh of $\deLTa$ $\approX$ -17.5$^O$, in both GalactIc hEmiSpheREs. we AlsO iMPosED a Cut AT $m_B$ = 14.5 Because bEyOnD this MagnITUDE the lauBertS & ValeNtijn (1989) catalog bEcoMes iNComPlete, In parTicuLaR for eArly tyPe galAxIes (Pellegrini 1990$a$). THis sElection eFfeCt Is cAuSed by THe fact ThaT thE LauberTs & ValenTIjn (1989) CaTALOg Is derived from the diAmETEr-Limited LAubertS (1982) CaTaLOg. The samPlE liMiteD AT $m_B$=14.5 coNtaiNS 780 gAlaxies iN both gALaCtIc hemisPhEres, of WhIch 694 (89%) HavE coloRS. As pOinted Out by MarZke & da cOsta (1997), this sample PResents no systEMaTIC dEPendEncE of the color CompLEtenEss wITh MagNItude Down tO tHE 14.5 lIMit of this sub-sample.
ALl Galaxy HelioCentric velociTies, $v_\odot$, hAVE Been corrEcteD FoR The Solar motion With rEspect to thE Centroid Of the local GroUp using $v =v_\ODOt + 300 sin(l) coS(b)$, wHerE ($l$, $b$) Are THE gAlactic coordiNATes oF tHe galaxY. We Also remOve FroM thE anAlYses all gaLaxies wiTh RaDiAl VelOcitiES less thaT 500, aS thE rEdsHifts FOr thesE objeCts aRe LiKEly To be domINaTED by tHeIr PecuLiaR vElociTies. wE usE througHout $H_o = 100 h$MpC$^{-1}$.
ThE two-POiNt correlaTion Function
==================================
MEtHod
------
The two-pOiNt cOrrelaTIOn functiOn $\xi (r)$ can be computed from tHE data usIng The esTimaTor suggesTed By HamiLtoN (1993): $$\Xi(r) = {DD(r) rR(r) \oveR [DR(r | southern galactic cap ($b \leq
-40^ o$),wit hin t he d ecli nation range $ - 40^o \leq \delta \leq -2.5 ^o$.Th e SSR S 2north contai n s1 9 39ga la xie sd is tribu ted over a solid ang leof 0.57 sterad i an s with $\d elt a \leq 0^o$wit h $b \ ge q + 3 5^o$. As desc ribedb y da C osta (199 7) the ga l axy mor p h ol ogie s in this sampleu se d classificatio ns bas ed on o the r w orks (Laub er ts &V alentij n 1 9 8 9 ) a s well as thos e made by t h e a uthors .
In this w ork w ea lso considered a s ub-sample of SS R S2 gala x ies whi ch hav e m eas ured co lo rs. F o r t h is we use d the La ub er ts &Vale n t i j n (1 989 ) ca talog , so that the su bcat a log of g alaxi es w it h col ors on ly co nt ains objects so uthof $\delt a$$\ app ro x$ -1 7 .5$^o$ , i n b oth gal actic h e mis ph e r e s. We also imposed a c u t a t $m_B$= 14.5 be ca u se beyon dthi s ma g n itude the La uberts & Valen t ij n(1989)ca talogbe com esincom p lete , in p articula r for early type gal a xies (Pellegr i ni 1 99 0 $a$) . T his selecti on e f fect isc au sed by th e fac tt ha t the Lauberts & Val en tijn ( 1989) catalog is d erived fro m t he diame ter- l im i ted Lauberts ( 1982) catalog.T he sampl e lim ited at$m_B$=14. 5 contains 78 0 g ala xie s in both galacti c hemi sp heres,ofwhich 6 94(89 %)hav ecolors. A s pointe dou tby Ma rzke& da Cost a(19 97 ),thiss ampleprese ntsno s y ste matic d e pe n d ence o fthecol or comp lete n ess with m agnitudedow n toth e14.5 li mit of this s ub -sample.
Al l g alaxyh e liocentr ic velocities, $v_\odot $ , havebee n cor rect ed for th e S olar m oti o n with respe ct to t hec e ntroi d of th eLocal Grou p usi ng $v = v_\o dot + 3 00 sin(l) cos(b)$, whe re ($l$, $b$) ar e th e ga lac t ic coo rd i nat e s of the galaxy. We also r em o ve from thea nal ys es allgalaxie s wit h radial velociti es less t ha t 50 0 , as the redsh ifts for these ob j ectsa re like lyto bedo min atedby the i r p eculi ar vel oc ities. We u se through out $H_o = 100 h$Mpc$^{ -1}$.
TheTwo -Point Co rre l ati on Functi on
= ========== === === ===== === = ===== ===Me tho d
---- --
T he two-po i nt co r r el ation funct i o n $\ xi (r )$c an becomp uted from the dat a using the est imat o r su gge s tedby Hamilton (199 3): $ $ \ xi(r) ={D D(r) RR(r)\over [D R( r | southern_galactic cap_($b \leq
-40^o$), within the_declination range_$-40^o_\leq \delta_\leq_-2.5^o$. The SSRS2_north contains 1939_galaxies distributed over a_solid angle of_0.57_steradians with $\delta \leq 0^o$ with $b \geq +35^o$. As described by da Costa_(1997)_the galaxy_morphologies_in_this sample used classifications based_on other works (Lauberts &_Valentijn 1989)_as well as those made by the authors.
In_this_work we also_considered a sub-sample of SSRS2 galaxies which have measured_colors. For this we used the_Lauberts & Valentijn_(1989)_catalog,_so that the subcatalog_of galaxies with colors only contains_objects south of $\delta$ $\approx$ -17.5$^o$,_in both galactic hemispheres. We also imposed_a cut at $m_B$ = 14.5_because beyond this magnitude the_Lauberts &_Valentijn (1989) catalog becomes incomplete,_in particular for_early type_galaxies (Pellegrini 1990$a$)._This selection effect is caused by_the fact that_the Lauberts & Valentijn (1989) catalog_is_derived from the_diameter-limited_Lauberts_(1982) catalog._The sample limited_at_$m_B$=14.5 contains_780_galaxies in both galactic hemispheres, of_which_694 (89%) have colors. As pointed out_by Marzke & da_Costa_(1997), this sample presents_no systematic dependence of the_color completeness with magnitude down to_the 14.5_limit of_this sub-sample.
All galaxy heliocentric velocities, $v_\odot$, have been corrected for the_Solar motion with respect to the_centroid of the Local_Group using_$v_=v_\odot + 300_sin(l)_cos(b)$, where_($l$, $b$) are the galactic coordinates of_the galaxy._We also remove from the analyses_all galaxies with radial_velocities_less that 500, as the redshifts_for these objects are likely to_be dominated by their peculiar_velocities._We_use throughout $H_o = 100_h$Mpc$^{-1}$.
The Two-Point Correlation Function
==================================
Method
------
The two-point correlation_function $\xi (r)$_can be computed from the data using_the_estimator suggested by Hamilton (1993): $$\xi(r)_=_{DD(r) RR(r) \over [DR(r |
Cu2O2] suggests $\langle \hat{c} \rangle \approx 0.3$ which is consistent with the estimated value $\langle \hat{c} \rangle = 0.44$ based on the ordered moment, $0.56 \mu_B$ per magnetic copper site [@NaCu2O2].
For magnet RMnO$_3$, the helical spin ordering occurs, corresponding to the condensation of the spin bosons. By using the standard linear-spin-wave approximation, a dynamical magon-phonon interaction reads, $$\begin{aligned}
\tilde{H}_{DM} &=& -\lambda S \cos Q \sum_q \delta u_q^x \tilde{S}_q^x ( \cos q -1) \nonumber \\ &~~~& - \lambda S \sum_q \delta u_q^y \tilde{S}_{q\pm Q}^y ( e^{\mp i Q} - e^{i(q \pm Q)}) /2\end{aligned}$$ $\delta u_q^y$ is hybridized with the spin at $q \pm Q$ (optical magnons), but $\delta u_q^x$ is coupled to $\tilde{S}^x$ at $q$ (acoustical magnons). The polarization correlation functions are given as $$\begin{aligned}
&& \ll \delta u_q^x | \delta u_{\bar{q}}^x \gg = \frac{\omega^2 - \omega_s^2}{M[\omega^4 - \omega^2(\omega_p^2 + \omega_s^2) + \omega_p^2 (\omega_{s}^2-\omega_{sp}^2)]}, \\
&& \ll \delta u_q^y | \delta u_{\bar{q}}^y \gg = \frac{1}{M[\omega^2 - \omega _p^2 + \frac{\lambda^2 S^3}{2M}\sum_{q'=q \pm Q}G_{s}(q')]}.\end{aligned}$$ where $\omega_p$ is the frequency for the transverse phonon, $\omega_s (q)$ is the energy dispersion of the spin-excitation, $\omega'_{sp}(q)= [2(A(q)-2B(q))(\lambda ^2 S^3 \cos^2 Q (1- \ | Cu2O2 ] suggests $ \langle \hat{c } \rangle \approx 0.3 $ which is consistent with the estimated value $ \langle \hat{c } \rangle = 0.44 $ based on the arranged here and now, $ 0.56 \mu_B$ per magnetic copper site [ @NaCu2O2 ].
For attraction RMnO$_3 $, the helical spin ordering happen, corresponding to the condensation of the tailspin boson. By using the standard linear - tailspin - wave approximation, a dynamical magon - phonon interaction read, $ $ \begin{aligned }
\tilde{H}_{DM } & = & -\lambda S \cos Q \sum_q \delta u_q^x \tilde{S}_q^x (\cos q -1) \nonumber \\ & ~~~ & - \lambda S \sum_q \delta u_q^y \tilde{S}_{q\pm Q}^y (e^{\mp i Q } - e^{i(q \pm Q) }) /2\end{aligned}$$ $ \delta u_q^y$ is hybridized with the spin at $ q \pm Q$ (optical magnons), but $ \delta u_q^x$ is coupled to $ \tilde{S}^x$ at $ q$ (acoustical magnons). The polarization correlation function are given as $ $ \begin{aligned }
& & \ll \delta u_q^x | \delta u_{\bar{q}}^x \gg = \frac{\omega^2 - \omega_s^2}{M[\omega^4 - \omega^2(\omega_p^2 + \omega_s^2) + \omega_p^2 (\omega_{s}^2-\omega_{sp}^2) ] }, \\
& & \ll \delta u_q^y | \delta u_{\bar{q}}^y \gg = \frac{1}{M[\omega^2 - \omega _ p^2 + \frac{\lambda^2 S^3}{2M}\sum_{q'=q \pm Q}G_{s}(q')]}.\end{aligned}$$ where $ \omega_p$ is the frequency for the transverse phonon, $ \omega_s (q)$ is the energy distribution of the spin - excitation, $ \omega'_{sp}(q)= [ 2(A(q)-2B(q))(\lambda ^2 S^3 \cos^2 Q (1- \ | Cu2O2] suggests $\langle \hat{c} \rakgle \approx 0.3$ whiey is cmnsistsnt with the estimated value $\langle \iat{c} \rangoe = 0.44$ based on the ordefed momenn, $0.56 \mu_B$ pee majnetic copper sivs [@NaCu2O2].
Njr mznnet XMiO$_3$, the helical xpin orderhng occurs, corserplnding to the condensation of the s[in bospnd. By using the stamqard linear-spin-wave approximation, a dyhamical magon-phonon onteraction reads, $$\begin{alihned}
\hilde{H}_{DM} &=& -\lambda S \cos Q \sum_q \deleq u_q^x \tilde{S}_d^x ( \cos q -1) \nonumber \\ &~~~& - \lambda S \sum_q \delta u_q^y \tilde{S}_{d\pm Q}^v ( e^{\mp i Q} - e^{u(q \ok Q)}) /2\end{aligied}$$ $\dejta u_q^y$ is hnnridizad with the spin at $q \pm Q$ (iptical magnons), but $\dxlta u_q^x$ is coupled jo $\tilde{S}^x$ ac $q$ (acoustical magnonw). Rhe pmlarhzatkin zordekafion fkncvions are gjven as $$\begun{aligned}
&& \ll \delta i_q^v | \delta u_{\bar{q}}^s \gg = \srwc{\omega^2 - \omega_s^2}{M[\omega^4 - \omega^2(\omega_p^2 + \omeca_s^2) + \omega_p^2 (\omega_{s}^2-\omega_{sp}^2)]}, \\
&& \ll \delta u_q^y | \delta o_{\bar{q}}^y \gg = \frac{1}{M[\omega^2 - \omega _p^2 + \frac{\lambda^2 S^3}{2M}\sum_{q'=q \pm Q}G_{s}(q')]}.\and{almgved}$$ wmere $\imfga_p$ is the frequency for the transverse phonjh, $\pmvga_s (q)$ is the enevgy dispersion of yhf xkin-excitation, $\umega'_{s'}(s)= [2(Z(q)-2B(q))(\lambda ^2 S^3 \cos^2 E (1- \ | Cu2O2] suggests $\langle \hat{c} \rangle \approx 0.3$ consistent the estimated $\langle \hat{c} \rangle ordered $0.56 \mu_B$ per copper site [@NaCu2O2]. magnet RMnO$_3$, the helical spin ordering corresponding to the condensation of the spin bosons. By using the standard linear-spin-wave a dynamical magon-phonon interaction reads, $$\begin{aligned} \tilde{H}_{DM} &=& -\lambda S \cos Q \sum_q u_q^x ( q \nonumber \\ &~~~& - \lambda S \sum_q \delta u_q^y \tilde{S}_{q\pm Q}^y ( e^{\mp i Q} - \pm Q)}) /2\end{aligned}$$ $\delta u_q^y$ is hybridized with spin at $q \pm (optical magnons), but $\delta u_q^x$ coupled $\tilde{S}^x$ at (acoustical The correlation functions are as $$\begin{aligned} && \ll \delta u_q^x | \delta u_{\bar{q}}^x \gg = \frac{\omega^2 - \omega_s^2}{M[\omega^4 - \omega^2(\omega_p^2 + + \omega_p^2 && \ll u_q^y \delta \gg = \frac{1}{M[\omega^2 _p^2 + \frac{\lambda^2 S^3}{2M}\sum_{q'=q \pm Q}G_{s}(q')]}.\end{aligned}$$ the frequency for the transverse phonon, $\omega_s (q)$ the energy of the spin-excitation, $\omega'_{sp}(q)= [2(A(q)-2B(q))(\lambda ^2 \cos^2 Q (1- \ | Cu2O2] suggests $\langle \hat{c} \rangLe \approx 0.3$ whIch is ConSisTeNt wiTh thE estimated valuE $\LangLe \hat{c} \rangle = 0.44$ based on the OrderEd MOmenT, $0.56 \Mu_b$ per mAgnetic COpPER siTe [@nacu2O2].
foR MaGnet RmnO$_3$, The heliCal spin ordEriNg Occurs, corresPOnDing to the cOndEnsation of thE spIn bosoNs. by uSIng thE stAndarD lineaR-Spin-waVe approxiMaTIon, a dyNAmical mAGOn-PhonOn interaction readS, $$\BeGIn{aligned}
\tilde{h}_{DM} &=& -\lamBdA s \cOS q \suM_q \dElta u_q^x \tilDe{s}_q^x ( \coS Q -1) \nonumbER \\ &~~~& - \lAMBDa S \SUm_q \delta u_q^y \tiLde{S}_{q\pm Q}^y ( e^{\mP I Q} - e^{I(q \pm Q)}) /2\eNd{AliGNed}$$ $\delTa u_q^y$ Is HYbrIdized with tHe spIn at $q \pm Q$ (oPtical MAgnons), bUT $\delta u_Q^x$ is coUplEd tO $\tilDE{S}^X$ aT $q$ (aCoUStiCAl MagNOns). the polarIzAtIon coRrelATION funCtiOns aRe givEn as $$\begin{aligNed}
&& \Ll \deLTa u_Q^x | \delTa u_{\baR{q}}^x \gG = \fRac{\omEga^2 - \omeGa_s^2}{M[\oMeGa^4 - \omega^2(\omega_p^2 + \omEga_s^2) + \Omega_p^2 (\omeGa_{s}^2-\OmEga_{Sp}^2)]}, \\
&& \Ll \delTA u_q^y | \deLta U_{\baR{q}}^y \gg = \frAc{1}{M[\omegA^2 - \OmeGa _P^2 + \FRAc{\Lambda^2 S^3}{2M}\sum_{q'=q \pm Q}G_{s}(Q')]}.\eND{AlIgned}$$ wheRe $\omegA_P$ iS tHE frequenCy For The tRANsverSe phONoN, $\omega_s (q)$ Is the eNErGy DispersIoN of the SpIn-eXciTatioN, $\OmegA'_{sp}(q)= [2(A(q)-2b(q))(\lambda ^2 s^3 \cos^2 Q (1- \ | Cu2O2] suggests $\langle \ hat{c} \ra ngle\ap pro x0.3$ whi ch is consiste n t wi th the estimated value $\la ng l e \h a t{ c} \r angle = 0. 4 4 $ b as ed on t h eorder edmoment, $0.56 \mu _B$ p er magneticc op per site [ @Na Cu2O2].
For ma gnet R Mn O$_ 3 $, th e h elica l spin orderi ng occurs ,c orresp o nding t o th e co ndensation of the sp i n bosons. By u sing t he st a n dar d l inear-spin -w ave a p proxima t io n , a d y namical magon -phonon int e rac tion r ea ds, $$\beg in{al ig n ed}
\tilde{H}_ {DM} &=& -\la mbda S \cos Q\ sum_q \ deltau_q ^x\til d e{ S} _q^ x( \c o sq - 1 ) \ nonumber \ \&~~~& - \ l a m b da S \s um_q \del ta u_q^y \til de{ S}_{ q \pm Q}^y ( e^ {\mp i Q} - e^{i( q \pm Q )}) /2\end{alig ned} $$ $\delt a u _q ^y$ i s hyb r idized wi ththe spi n at $q \pm Q $ ( op tical magnons), bu t$ \ de lta u_q^ x$ isc ou pl e d to $\t il de{ S}^x $ at $q $ (a c ou stical m agnons ) .Th e polar iz ationco rre lat ion f u ncti ons ar e givenas $$ \ begin{aligned} && \ll \delta u_ q ^ x| \de lta u_{\bar{q} }^x\ gg = \fr a c{ \om e ga^2- \om eg a _s ^ 2}{M[\omega^4 - \om eg a^2(\o mega_ p^2 + \omega_ s^2) + \om e g a _p^2 (\o mega _ {s } ^2-\omega_{sp} ^2)]} , \\
&& \l l \deltau_q^y | \delt a u_{\bar { q }}^y \gg =\fr ac{ 1}{ M [ \o mega^2 - \ome g a _p^ 2+ \fra c{\ lambda^ 2 S ^3} {2M }\s um _{q'=q \p m Q}G_{s }( q' )] }. \en d{ali g ned}$$ w he re$\ ome ga_p$ is the freq uenc yfo r th e trans v er s e pho no n, $\o meg a_ s (q) $ is the energy dispersi ono f th esp in-exci tation, $\ome ga '_{sp}(q)= [ 2(A (q)-2B ( q ))(\lamb da ^2 S^3 \cos^2 Q (1-\ | Cu2O2] suggests_$\langle \hat{c}_\rangle \approx 0.3$ which_is consistent_with_the estimated_value_$\langle \hat{c} \rangle_= 0.44$ based_on the ordered moment,_$0.56 \mu_B$ per_magnetic_copper site [@NaCu2O2].
For magnet RMnO$_3$, the helical spin ordering occurs, corresponding to the condensation_of_the spin_bosons._By_using the standard linear-spin-wave approximation,_a dynamical magon-phonon interaction reads,_$$\begin{aligned}
\tilde{H}_{DM} &=&_-\lambda S \cos Q \sum_q \delta u_q^x \tilde{S}_q^x_(_\cos q -1)_\nonumber \\ &~~~& - \lambda S \sum_q \delta u_q^y_\tilde{S}_{q\pm Q}^y ( e^{\mp i Q}_- e^{i(q \pm_Q)})_/2\end{aligned}$$_$\delta u_q^y$ is hybridized_with the spin at $q \pm_Q$ (optical magnons), but $\delta u_q^x$_is coupled to $\tilde{S}^x$ at $q$ (acoustical_magnons). The polarization correlation functions are_given as $$\begin{aligned}
&& \ll \delta_u_q^x |_\delta u_{\bar{q}}^x \gg = \frac{\omega^2_- \omega_s^2}{M[\omega^4 -_\omega^2(\omega_p^2 +_\omega_s^2) + \omega_p^2_(\omega_{s}^2-\omega_{sp}^2)]}, \\
&& \ll \delta u_q^y |_\delta u_{\bar{q}}^y \gg_= \frac{1}{M[\omega^2 - \omega _p^2 +__\frac{\lambda^2 S^3}{2M}\sum_{q'=q \pm_Q}G_{s}(q')]}.\end{aligned}$$_where_$\omega_p$ is_the frequency for_the_transverse phonon,_$\omega_s_(q)$ is the energy dispersion of_the_spin-excitation, $\omega'_{sp}(q)= [2(A(q)-2B(q))(\lambda ^2 S^3 \cos^2 Q_(1- \ |
Big)\Big|\\
& \leq & c\,\frac{|u_1|^3+|u_2|^3}{\sqrt {n^3h^k}}+\frac{(u_1u_2)^2}{n^2}+c\,\frac{|u_2|^3(1+u_1^2)+|u_1|^3(1+u_2^2)}{\sqrt {n^3h^k}}+c\,\frac{|u_1u_2|h^k}{n}.\end{aligned}$$ By assumption, $nh^{3k}\to 0$ so that for $n$ large enough : $h^k\leq 1/\sqrt {nh^k}$. Consequently, $$\Big|E\exp\Big(i\Big(M_1+M_2\Big)\Big)-E\exp\Big(iM_1\Big)E\exp\Big(iM_2\Big)\Big|\leq \frac{Q(|u_1|,|u_2|)}{\sqrt {nh^k}},$$ where $Q$ is defined for all $u_1,u_2\in{I\!\!R}$ by : $$Q(u_1,u_2)=c\big(u_1^3+u_2^3+(u_1u_2)^2+u_1u_2+u_2^2u_1^3+u_1^3u_2^2\big).$$ Consequently, for all $u_1,u_2\in{I\!\!R}$ and $x,y\in{{\cal V}}_n^t\cup{{\overline {\cal V}}}_n^t$ such that $\|x-y\|\geq 2h$ : $$\begin{aligned}
& & \Big|E\exp\Big(i\Big(u_1Z_n(x)+u_2Z_n(y)\Big)\Big)-E\exp\Big(iu_1Z_n(x)\Big)E\exp\Big(iu_2Z_n(y)\Big)\Big|\\
& = & \Big|\Big(E\exp\Big(i\Big(M_1+M_2\Big)\Big)\Big)^n-\Big(E\exp\Big(iM_1\Big)E\exp\Big(i | Big)\Big|\\
& \leq & c\,\frac{|u_1|^3+|u_2|^3}{\sqrt { n^3h^k}}+\frac{(u_1u_2)^2}{n^2}+c\,\frac{|u_2|^3(1+u_1 ^ 2)+|u_1|^3(1+u_2 ^ 2)}{\sqrt { n^3h^k}}+c\,\frac{|u_1u_2|h^k}{n}.\end{aligned}$$ By assumption, $ nh^{3k}\to 0 $ so that for $ n$ large enough: $ h^k\leq 1/\sqrt { nh^k}$. Consequently, $ $ \Big|E\exp\Big(i\Big(M_1+M_2\Big)\Big)-E\exp\Big(iM_1\Big)E\exp\Big(iM_2\Big)\Big|\leq \frac{Q(|u_1|,|u_2|)}{\sqrt { nh^k}},$$ where $ Q$ is defined for all $ u_1,u_2\in{I\!\!R}$ by: $ $ Q(u_1,u_2)=c\big(u_1 ^ 3+u_2 ^ 3+(u_1u_2)^2+u_1u_2+u_2 ^ 2u_1 ^ 3+u_1 ^ 3u_2 ^ 2\big).$$ Consequently, for all $ u_1,u_2\in{I\!\!R}$ and $ x, y\in{{\cal V}}_n^t\cup{{\overline { \cal V}}}_n^t$ such that $ \|x - y\|\geq 2h$: $ $ \begin{aligned }
& & \Big|E\exp\Big(i\Big(u_1Z_n(x)+u_2Z_n(y)\Big)\Big)-E\exp\Big(iu_1Z_n(x)\Big)E\exp\Big(iu_2Z_n(y)\Big)\Big|\\
& = & \Big|\Big(E\exp\Big(i\Big(M_1+M_2\Big)\Big)\Big)^n-\Big(E\exp\Big(iM_1\Big)E\exp\Big(i | Big)\Hig|\\
& \leq & c\,\frac{|u_1|^3+|u_2|^3}{\sqrt {n^3h^k}}+\nrac{(u_1u_2)^2}{n^2}+c\,\frac{|u_2|^3(1+u_1^2)+|u_1|^3(1+u_2^2)}{\sqrt {n^3h^n}}+c\,\frac{|h_1u_2|h^k}{n}.\end{xligned}$$ By assumption, $nh^{3k}\to 0$ si thau for $n$ large enougf : $h^k\leq 1/\dqrt {nh^k}$. Conwwquently, $$\Bmf|E\exp\Bin(n\Big(M_1+J_2\Nig)\Biy)-E\xxp\Big(iM_1\Big)E\exp\Nig(iM_2\Big)\Big|\neq \frac{Q(|u_1|,|u_2|)}{\sqrd {vh^n}},$$ where $Q$ is defined for all $u_1,u_2\in{I\!\!R}$ by : $$Q(u_1,i_2)=c\hig(u_1^3+u_2^3+(u_1u_2)^2+u_1u_2+u_2^2u_1^3+u_1^3u_2^2\byg).$$ Cpgseqhvnuly, for all $u_1,u_2\in{I\!\!R}$ and $x,y\in{{\cal V}}_n^f\cup{{\ovegline {\cal V}}}_n^t$ sucn that $\|x-y\|\geq 2h$ : $$\begin{alignfd}
& & \Hig|E\exp\Big(i\Big(u_1Z_n(x)+k_2Z_n(y)\Big)\Big)-E\gsp\Byt(iu_1Z_n(x)\Big)E\exo\Big(iu_2Z_n(y)\Bpy)\Big|\\
& = & \Big|\Bjg(E\exp\Big(i\Big(M_1+M_2\Big)\Big)\Big)^n-\Big(E\exo\Big(iK_1\Big)E\exp\Bit(i | Big)\Big|\\ & \leq & c\,\frac{|u_1|^3+|u_2|^3}{\sqrt {n^3h^k}}+\frac{(u_1u_2)^2}{n^2}+c\,\frac{|u_2|^3(1+u_1^2)+|u_1|^3(1+u_2^2)}{\sqrt {n^3h^k}}+c\,\frac{|u_1u_2|h^k}{n}.\end{aligned}$$ $nh^{3k}\to so that $n$ large enough $$\Big|E\exp\Big(i\Big(M_1+M_2\Big)\Big)-E\exp\Big(iM_1\Big)E\exp\Big(iM_2\Big)\Big|\leq {nh^k}},$$ where $Q$ defined for all by : $$Q(u_1,u_2)=c\big(u_1^3+u_2^3+(u_1u_2)^2+u_1u_2+u_2^2u_1^3+u_1^3u_2^2\big).$$ Consequently, for all and $x,y\in{{\cal V}}_n^t\cup{{\overline {\cal V}}}_n^t$ such that $\|x-y\|\geq 2h$ : $$\begin{aligned} & & & = & \Big|\Big(E\exp\Big(i\Big(M_1+M_2\Big)\Big)\Big)^n-\Big(E\exp\Big(iM_1\Big)E\exp\Big(i | Big)\Big|\\
& \leq & c\,\frac{|u_1|^3+|u_2|^3}{\sqrt {n^3h^k}}+\frAc{(u_1u_2)^2}{n^2}+c\,\frac{|U_2|^3(1+u_1^2)+|u_1|^3(1+u_2^2)}{\sQrt {N^3h^k}}+C\,\fRac{|u_1U_2|h^k}{n}.\End{aligned}$$ By asSUmptIon, $nh^{3k}\to 0$ so that for $n$ largE enouGh : $H^K\leq 1/\SQrT {nh^k}$. COnsequeNTlY, $$\bIg|E\ExP\BIg(i\biG(m_1+M_2\big)\BiG)-E\eXp\Big(iM_1\big)E\exp\Big(IM_2\BIg)\big|\leq \frac{Q(|u_1|,|U_2|)}{\SqRt {nh^k}},$$ where $q$ is Defined for alL $u_1,u_2\In{I\!\!R}$ by : $$q(u_1,U_2)=c\bIG(u_1^3+u_2^3+(u_1u_2)^2+U_1u_2+u_2^2U_1^3+u_1^3u_2^2\biG).$$ ConseQUently, For all $u_1,u_2\iN{I\!\!r}$ And $x,y\iN{{\Cal V}}_n^t\cUP{{\OvErliNe {\cal V}}}_n^t$ such that $\|x-Y\|\GeQ 2H$ : $$\begin{aligned}
& & \BIg|E\exp\biG(I\BIG(U_1Z_n(X)+u_2Z_N(y)\Big)\Big)-E\eXp\big(iu_1z_N(x)\Big)E\eXP\BIG(IU_2Z_n(Y)\big)\Big|\\
& = & \Big|\Big(E\Exp\Big(i\Big(M_1+m_2\big)\big)\Big)^N-\BIg(E\EXp\Big(im_1\Big)E\ExP\big(I | Big)\Big|\\
& \leq & c\,\f rac{|u_1|^ 3+|u_ 2|^ 3}{ \s qrt{n^3 h^k}}+\frac{(u _ 1u_2 )^2}{n^2}+c\,\frac{|u_ 2|^3( 1+ u _1^2 ) +| u_1|^ 3(1+u_2 ^ 2) } { \sq rt { n^3 h^ k }} +c\,\ fra c{|u_1u _2|h^k}{n} .\e nd {aligned}$$B yassumption , $ nh^{3k}\to 0 $ s o that f or$ n$ la rge enou gh : $ h ^k\leq 1/\sqrt{n h ^k}$.C onseque n t ly , $$ \Big|E\exp\Big(i\ B ig ( M_1+M_2\Big)\B ig)-E\ ex p \B i g (iM _1\ Big)E\exp\ Bi g(iM_ 2 \Big)\B i g| \ l e q \ f rac{Q(|u_1|,| u_2|)}{\sqr t {n h^k}}, $$ wh e re $Q$ is d ef i ned for all $u _1,u _2\in{I\! \!R}$b y : $$Q ( u_1,u_2 )=c\bi g(u _1^ 3+u_ 2 ^3 +( u_1 u_ 2 )^2 + u_ 1u_ 2 +u_ 2^2u_1^3 +u _1 ^3u_2 ^2\b i g ) . $$ C ons eque ntly, for all $u_1 ,u_ 2\in { I\! \!R}$ and$x,y \i n{{\c al V}} _n^t\ cu p{{\overline {\ calV}}}_n^t$ su ch th at $\|x - y\|\ge q 2 h$: $$\be gin{ali g ned }& & \ Big|E\exp\Big(i\Bi g( u _ 1Z _n(x)+u_ 2Z_n(y ) \B ig ) \Big)-E\ ex p\B ig(i u _ 1Z_n( x)\B i g) E\exp\Bi g(iu_2 Z _n (y )\Big)\ Bi g|\\
& = &\Bi g|\Bi g (E\e xp\Big (i\Big(M _1+M_ 2 \Big)\Big)\Big ) ^n-\Big(E\exp \ Bi g ( iM _ 1\Bi g)E \exp\Big(i | Big)\Big|\\
& \leq_& c\,\frac{|u_1|^3+|u_2|^3}{\sqrt_{n^3h^k}}+\frac{(u_1u_2)^2}{n^2}+c\,\frac{|u_2|^3(1+u_1^2)+|u_1|^3(1+u_2^2)}{\sqrt {n^3h^k}}+c\,\frac{|u_1u_2|h^k}{n}.\end{aligned}$$ By assumption,_$nh^{3k}\to 0$_so_that for_$n$_large enough :_$h^k\leq 1/\sqrt {nh^k}$._Consequently, $$\Big|E\exp\Big(i\Big(M_1+M_2\Big)\Big)-E\exp\Big(iM_1\Big)E\exp\Big(iM_2\Big)\Big|\leq \frac{Q(|u_1|,|u_2|)}{\sqrt {nh^k}},$$_where $Q$ is_defined_for all $u_1,u_2\in{I\!\!R}$ by : $$Q(u_1,u_2)=c\big(u_1^3+u_2^3+(u_1u_2)^2+u_1u_2+u_2^2u_1^3+u_1^3u_2^2\big).$$ Consequently, for all $u_1,u_2\in{I\!\!R}$ and $x,y\in{{\cal V}}_n^t\cup{{\overline {\cal_V}}}_n^t$_such that_$\|x-y\|\geq_2h$_: $$\begin{aligned}
& & \Big|E\exp\Big(i\Big(u_1Z_n(x)+u_2Z_n(y)\Big)\Big)-E\exp\Big(iu_1Z_n(x)\Big)E\exp\Big(iu_2Z_n(y)\Big)\Big|\\
& =_& \Big|\Big(E\exp\Big(i\Big(M_1+M_2\Big)\Big)\Big)^n-\Big(E\exp\Big(iM_1\Big)E\exp\Big(i |
ensuremath{\mathrm{ns}}}}}$ and no additional dephasing, while the dashed lines correspond to $T_1 = 17\ {\ensuremath{{\ensuremath{\mathrm{ns}}}}}$ and pure dephasing with $T_\phi = 16\ {\ensuremath{{\ensuremath{\mathrm{ns}}}}}$ due to low frequency noise (see text).](010206H1linesv3.eps){width="3.0in"}
As with all of the simulations discussed in this section so far, the solid lines in Fig. \[F010206H1lines\] were calculated with $T_1 = 17\ {\ensuremath{{\ensuremath{\mathrm{ns}}}}}$ and $T_\phi = 16\ {\ensuremath{{\ensuremath{\mathrm{ns}}}}}$. This gives $T_2 = 10.9\ {\ensuremath{{\ensuremath{\mathrm{ns}}}}}$ and a Rabi decay time roughly equal to the two-level value[@Torrey49a; @Smith78a] of $T^\prime = {\ensuremath{\left[ 1/{\ensuremath{\left( 2 T_1 \right)}} + 1/{\ensuremath{\left( 2 T_2 \right)}} \right]}}^{-1} = 13.3\ {\ensuremath{{\ensuremath{\mathrm{ns}}}}}$, consistent with measurements of the decay envelope of the escape rate. If pure dephasing were not present then $T^\prime$ would be 22.7 ns, which is significantly longer than what is observed. Although unlikely, our thermal measurement of $T_1 = 17\ {\ensuremath{{\ensuremath{\mathrm{ns}}}}}$ could be incorrect. To examine this possibility, simulations with $T_1 = 10\ {\ensuremath{{\ensuremath{\mathrm{ns}}}}}$ and $T_2 = 2 T_1$ (which also give $T^\prime = 13.3\ {\ensuremath{{\ensuremath{\mathrm{ns}}}}}$) are shown with dotted lines in Fig. \[F010206H1lines\]. They are nearly identical to the curves calculated with dephasing, although [$\Gamma_{\infty}$]{} is somewhat smaller. Thus this data set alone cannot rule out dissipation-limited decoherence. However, the shorter $T_1$ reduces the prediction for [$\Gamma_{\infty}$]{} by roughly 15% over the full range of measured powers in Fig. \[F031806DGeq\], suggesting that additional dephasing is instead affecting the Rabi oscillations.
While dephasing is needed to faithfully reproduce features of the | ensuremath{\mathrm{ns}}}}}$ and no additional dephasing, while the dashed lines correspond to $ T_1 = 17\ { \ensuremath{{\ensuremath{\mathrm{ns}}}}}$ and arrant dephasing with $ T_\phi = 16\ { \ensuremath{{\ensuremath{\mathrm{ns}}}}}$ due to depleted frequency noise (see text).](010206H1linesv3.eps){width="3.0 in " }
As with all of the simulations hash out in this section so far, the hearty lines in Fig. \[F010206H1lines\ ] were calculated with $ T_1 = 17\ { \ensuremath{{\ensuremath{\mathrm{ns}}}}}$ and $ T_\phi = 16\ { \ensuremath{{\ensuremath{\mathrm{ns}}}}}$. This contribute $ T_2 = 10.9\ { \ensuremath{{\ensuremath{\mathrm{ns}}}}}$ and a Rabi decay time roughly adequate to the two - level value[@Torrey49a; @Smith78a ] of $ T^\prime = { \ensuremath{\left [ 1/{\ensuremath{\left (2 T_1 \right) } } + 1/{\ensuremath{\left (2 T_2 \right) } } \right]}}^{-1 } = 13.3\ { \ensuremath{{\ensuremath{\mathrm{ns}}}}}$, consistent with measurements of the decay envelope of the escape pace. If pure dephasing were not present then $ T^\prime$ would be 22.7 ns, which is importantly longer than what is watch. Although improbable, our thermal measurement of $ T_1 = 17\ { \ensuremath{{\ensuremath{\mathrm{ns}}}}}$ could be incorrect. To examine this possibility, pretense with $ T_1 = 10\ { \ensuremath{{\ensuremath{\mathrm{ns}}}}}$ and $ T_2 = 2 T_1 $ (which also give $ T^\prime = 13.3\ { \ensuremath{{\ensuremath{\mathrm{ns}}}}}$) are shown with dotted lineage in Fig. \[F010206H1lines\ ]. They are nearly identical to the curves calculated with dephasing, although [ $ \Gamma_{\infty}$ ] { } is somewhat smaller. therefore this data set alone cannot rule out dissipation - limited decoherence. However, the short $ T_1 $ reduces the prediction for [ $ \Gamma_{\infty}$ ] { } by roughly 15% over the full range of measured powers in Fig. \[F031806DGeq\ ], suggesting that additional dephasing is instead affecting the Rabi oscillation.
While dephasing is needed to faithfully reproduce feature of the | enskremath{\mathrm{ns}}}}}$ and no aaditional dephasing, whmle the dashed uines correspond to $T_1 = 17\ {\ensucemarh{{\ensyremath{\mathrm{ns}}}}}$ and pufe dephaspng with $R_\phi = 16\ {\ensuremavg{{\ensuremath{\mafmrm{ns}}}}}$ vue to low freqoency noise (vee text).](010206H1linese3.eos){cidth="3.0in"}
As with all of the simulationf discuxsfd in this secjion xj fad, the solid lines in Fig. \[F010206H1lines\] wede calcllated with $T_1 = 17\ {\emsuremath{{\ensuremath{\mathrm{nd}}}}}$ anf $T_\phi = 16\ {\ensurematj{{\ensuremath{\nathwn{ns}}}}}$. This givds $T_2 = 10.9\ {\enslxemath{{\ensurgmath{\mathrm{ns}}}}}$ and a Rabi decay tkme rpughly equql to jhe two-level valuv[@Torrey49a; @Smitm78s] of $T^\[rime = {\rnsuremath{\left[ 1/{\envurwmath{\left( 2 T_1 \right)}} + 1/{\eisuremath{\left( 2 T_2 \rigrt)}} \right]}}^{-1} = 13.3\ {\znsuremath{{\ensuremath{\mqtyrm{ns}}}}}$, convistdbt ditg kezsuremfnta of the dscay envelope of the escape raue. Yd pure dephasjng wewe not present then $T^\prime$ would be 22.7 ns, wvicg is significantly longwr than what is obserged. Althotgh unlikely, our thermal measurement of $T_1 = 17\ {\ensuramath{{\xnrurtmqth{\magyrl{ns}}}}}$ could be incorrect. To examine this possibymiuy, ximulations wibh $T_1 = 10\ {\ensuremath{{\emskrriath{\mathrm{ns}}}}}$ xnd $T_2 = 2 T_1$ (which also give $T^\orime = 13.3\ {\ensueemath{{\enstremsth{\mathrm{ns}}}}}$) are shown with eotted lines un Fig. \[F010206H1lines\]. They are nearly idemticak to the curves calculaced wifh dephasinh, althougg [$\Gamma_{\infty}$]{} is sooewmat smaller. Thus this data see alone cennot rule ojt dossipaeion-limitef decoherence. However, thf shotter $T_1$ reduces tje prediction for [$\Gamma_{\infty}$]{} by cpughly 15% over tve xull ranye of keasured powews in Fig. \[F031806DGeq\], suggescing tfat additiknal de'hasing is igstead affecthjg the Rabi mscillatyons.
Qhilw depharkng is needed yo faithflljy rwproduce features pf jhs | ensuremath{\mathrm{ns}}}}}$ and no additional dephasing, while the correspond $T_1 = {\ensuremath{{\ensuremath{\mathrm{ns}}}}}$ and pure {\ensuremath{{\ensuremath{\mathrm{ns}}}}}$ to low frequency (see text).](010206H1linesv3.eps){width="3.0in"} As all of the simulations discussed in section so far, the solid lines in Fig. \[F010206H1lines\] were calculated with $T_1 17\ {\ensuremath{{\ensuremath{\mathrm{ns}}}}}$ and $T_\phi = 16\ {\ensuremath{{\ensuremath{\mathrm{ns}}}}}$. This gives $T_2 = 10.9\ {\ensuremath{{\ensuremath{\mathrm{ns}}}}}$ a decay roughly to the two-level value[@Torrey49a; @Smith78a] of $T^\prime = {\ensuremath{\left[ 1/{\ensuremath{\left( 2 T_1 \right)}} + 1/{\ensuremath{\left( 2 \right)}} \right]}}^{-1} = 13.3\ {\ensuremath{{\ensuremath{\mathrm{ns}}}}}$, consistent with measurements the decay envelope of escape rate. If pure dephasing not then $T^\prime$ be ns, is significantly longer what is observed. Although unlikely, our thermal measurement of $T_1 = 17\ {\ensuremath{{\ensuremath{\mathrm{ns}}}}}$ could be incorrect. To this possibility, $T_1 = {\ensuremath{{\ensuremath{\mathrm{ns}}}}}$ $T_2 2 T_1$ (which $T^\prime = 13.3\ {\ensuremath{{\ensuremath{\mathrm{ns}}}}}$) are shown in Fig. \[F010206H1lines\]. They are nearly identical to curves calculated dephasing, although [$\Gamma_{\infty}$]{} is somewhat smaller. this data set alone cannot rule out dissipation-limited However, the shorter $T_1$ reduces the prediction for [$\Gamma_{\infty}$]{} by roughly 15% over the full measured powers in Fig. suggesting that additional is affecting Rabi While dephasing needed to faithfully reproduce features of the | ensuremath{\mathrm{ns}}}}}$ and no adDitional dePhasiNg, wHilE tHe daShed Lines corresponD To $T_1 = 17\ {\eNsuremath{{\ensuremath{\matHrm{ns}}}}}$ AnD Pure DEpHasinG with $T_\pHI = 16\ {\eNSUreMaTh{{\EnsUrEMaTh{\matHrm{Ns}}}}}$ due to Low frequenCy nOiSe (see text).](010206H1liNEsV3.eps){width="3.0iN"}
As With all of the SimUlatioNs DisCUssed In tHis seCtion sO Far, the Solid lineS iN fig. \[F010206H1lINes\] were CALcUlatEd with $T_1 = 17\ {\ensuremath{{\ENsURemath{\mathrm{ns}}}}}$ And $T_\phI = 16\ {\eNSuREMatH{{\enSuremath{\maThRm{ns}}}}}$. THIs gives $t_2 = 10.9\ {\EnSUREmaTH{{\ensuremath{\maThrm{ns}}}}}$ and a RABi dEcay tiMe RouGHly equAl to tHe TWo-lEvel value[@ToRrey49A; @Smith78a] of $t^\prime = {\ENsuremaTH{\left[ 1/{\enSuremaTh{\lEft( 2 t_1 \rigHT)}} + 1/{\eNsUreMaTH{\leFT( 2 T_2 \RigHT)}} \riGht]}}^{-1} = 13.3\ {\ensurEmAtH{{\ensuRemaTH{\MAThrm{Ns}}}}}$, cOnsiStent With measuremeNts Of thE DecAy envElope Of thE eScape Rate. If Pure dEpHasing were not prEsenT then $T^\priMe$ wOuLd bE 22.7 nS, whicH Is signIfiCanTly longEr than wHAt iS oBSERvEd. Although unlikely, OuR THeRmal measUremenT Of $t_1 = 17\ {\eNSuremath{{\EnSurEmatH{\MAthrm{Ns}}}}}$ coULd Be incorrEct. To eXAmInE this poSsIbilitY, sImuLatIons wITh $T_1 = 10\ {\eNsuremAth{{\ensurEmath{\MAthrm{ns}}}}}$ and $T_2 = 2 T_1$ (whICh also give $T^\prIMe = 13.3\ {\ENSuREmatH{{\enSuremath{\matHrm{nS}}}}}$) Are sHown WItH doTTed liNes in fiG. \[f010206H1LInes\]. They are nearly idEnTical tO the cUrves calculatEd with dephASINg, althouGh [$\GaMMa_{\INfty}$]{} is somewhat SmallEr. Thus this DAta set alOne caNnot rule Out dissipATIon-limitEd dEcoHerEncE. hOwEver, the shorteR $t_1$ ReduCeS the preDicTion for [$\gamMa_{\iNftY}$]{} by RoUghly 15% over The full rAnGe Of MeAsuRed poWErs in Fig. \[f031806Dgeq\], SuGgeSting THat addItionAl dePhAsINg iS insteaD AfFECtinG tHe rabi OscIlLatioNs.
WhILe dEphasinG is needed To fAIthfUlLy ReproduCe features of tHe | ensuremath{\mathrm{ns}}}}} $ and no a dditi ona l d ep hasi ng,while the dash e d li nes correspond to $T_1 = 17 \{ \ens u re math{ {\ensur e ma t h {\m at hr m{n s} } }} }$ an d p ure dep hasing wit h $ T_ \phi = 16\ { \ en suremath{{ \en suremath{\ma thr m{ns}} }} }$d ue to lo w fre quency noise(see text ). ] (01020 6 H1lines v 3 .e ps){ width="3.0in"}
A s w i th all of thesimula ti o ns d isc uss ed in this s ectio n so far , t h e sol i d lines in Fi g. \[F01020 6 H1l ines\] w ere calcul atedwi t h $ T_1 = 17\ { \ens uremath{{ \ensur e math{\m a thrm{ns }}}}}$ an d $ T_\p h i=16\ { \ ens u re mat h {{\ ensurema th {\ mathr m{ns } } } } }$.Thi s gi ves $ T_2 = 10.9\ { \en sure m ath {{\en surem ath{ \m athrm {ns}}} }}$ a nd a Rabi decay t imeroughly e qua ltoth e two - levelval ue[ @Torrey 49a; @S m ith 78 a ] of $T^\prime = {\ens ur e m at h{\left[ 1/{\e n su re m ath{\lef t( 2T_1\ r ight) }} + 1/ {\ensure math{\ l ef t( 2 T_2\r ight)} }\ri ght ]}}^{ - 1} = 13.3\ {\ensur emath { {\ensuremath{\ m athrm{ns}}}}} $ ,c o ns i sten t w ith measure ment s ofthed ec aye nvelo pe of t h ee scape rate. If pure d ephasi ng we re not presen t then $T^ \ p r ime$ wou ld b e 2 2 .7 ns, which i s sig nificantly longer t han w hat is o bserved.A l though u nli kel y,our t he rmal measurem e n t of $ T_1 = 1 7\{\ensur ema th{ {\e nsu re math{\mat hrm{ns}} }} }$ c ou ldbe in c orrect.To ex am ine this possib ility , si mu la t ion s with$ T_ 1 = 10 \{\ ensu rem at h{{\e nsur e mat h{\math rm{ns}}}} }$a nd $ T_ 2= 2 T_1 $ (which also g ive $T^\pr im e = 13.3\ { \ensurem ath{{\ensuremath{\mathr m {ns}}}} }$) areshow n with do tte d line s i n Fig.\[F010 206H1 li nes \ ] . The y ar e n ea rly identi c a l t o the c urve s calcu lated with dephasi n g,although [$\G amm a_{\ i n ft y}$ ] {} isso m ewh a t smaller. Thusthis datase t a lone canno t ru le out di ssipati on-li m ited de coherence . However ,thes h ort er $T_1$ r educes t he predic t ion f o r[$\Ga mma _{\inf ty }$] {} by rough l y 1 5% ov er the f ull ra nge o fmeasured powers in Fig. \[F0318 06DGeq \], s ugg esting th ata ddi tional de phas ing is ins tea d a ffect ing the R abio sc ill a tions .
W h ile depha s in g i s ne eded to fai t h f ull y rep rod u ce fea ture s of the | ensuremath{\mathrm{ns}}}}}$ and_no additional_dephasing, while the dashed_lines correspond_to_$T_1 =_17\_{\ensuremath{{\ensuremath{\mathrm{ns}}}}}$ and pure_dephasing with $T_\phi_= 16\ {\ensuremath{{\ensuremath{\mathrm{ns}}}}}$ due_to low frequency_noise_(see text).](010206H1linesv3.eps){width="3.0in"}
As with all of the simulations discussed in this section so far, the_solid_lines in_Fig. \[F010206H1lines\]_were_calculated with $T_1 = 17\_{\ensuremath{{\ensuremath{\mathrm{ns}}}}}$ and $T_\phi = 16\_{\ensuremath{{\ensuremath{\mathrm{ns}}}}}$. This_gives $T_2 = 10.9\ {\ensuremath{{\ensuremath{\mathrm{ns}}}}}$ and a Rabi_decay_time roughly equal_to the two-level value[@Torrey49a; @Smith78a] of $T^\prime = {\ensuremath{\left[_1/{\ensuremath{\left( 2 T_1 \right)}} + 1/{\ensuremath{\left(_2 T_2 \right)}}_\right]}}^{-1}_=_13.3\ {\ensuremath{{\ensuremath{\mathrm{ns}}}}}$, consistent with_measurements of the decay envelope of_the escape rate. If pure dephasing_were not present then $T^\prime$ would be_22.7 ns, which is significantly longer_than what is observed. Although_unlikely, our_thermal measurement of $T_1 =_17\ {\ensuremath{{\ensuremath{\mathrm{ns}}}}}$ could_be incorrect._To examine this_possibility, simulations with $T_1 = 10\_{\ensuremath{{\ensuremath{\mathrm{ns}}}}}$ and $T_2_= 2 T_1$ (which also give_$T^\prime_= 13.3\ {\ensuremath{{\ensuremath{\mathrm{ns}}}}}$)_are_shown_with dotted_lines in Fig. \[F010206H1lines\]._They_are nearly_identical_to the curves calculated with dephasing,_although_[$\Gamma_{\infty}$]{} is somewhat smaller. Thus this data set_alone cannot rule out_dissipation-limited_decoherence. However, the shorter_$T_1$ reduces the prediction for_[$\Gamma_{\infty}$]{} by roughly 15% over the full_range of_measured powers_in Fig. \[F031806DGeq\], suggesting that additional dephasing is instead affecting the Rabi_oscillations.
While dephasing is needed to faithfully_reproduce features of the |
time of integration is given by $T=\ln\left( \frac{\eta_j}{v_j} \right)^{1/\beta_j}$. Such a time of integration is obtained from the equation $v_j'=\beta_jv_j$. The we have [ $$\begin{split}
\tilde u &= u \left( \frac{v_j}{\eta_j} \right)^{1/\beta}\\
\tilde v_i &= v_i \left( \frac{\eta_j}{v_j} \right)^{\beta_i/\beta_j}\\
\tilde w &= w \left( \frac{\eta_j}{v_j} \right)^{\gamma/\beta_j}.
\end{split}$$ ]{}
It then only rests to compute $\tilde Z$. Following similar arguments as for the transition $\Pi^w$ we get in this case [ $$\begin{split}
\tilde Z=Z\exp\left[ -\frac{\Lambda}{\gamma w} \left( 1+\tilde\alpha' w\ln(v_j)+w\tilde G'\right)\right],
\end{split}$$ ]{}
where now [ $$\begin{split}
\tilde\alpha' &=\tilde\alpha'(uv_i^{1/\beta_i},uw^{1/\gamma})\\
\tilde G' &=\tilde G'(uv_i^{1/\beta_i},uw^{1/\gamma},\mu_w,\mu_i)
\end{split}$$ ]{}
are $C^\ell$ functions with $\mu_w=wv_j^{-\gamma/\beta_j}$ and $\mu_i=v_iv_j^{-\beta_i/\beta_j}$.
Saddle 2 {#sub:case2_.unnumbered}
--------
In this case we investigate the transitions of a vector field of the form [ $$\begin{split}
\label{eq:semihyp2}
Y:\begin{cases}
u' &= w u \\
v_j' &= -\beta_j w v_j \\
w' &= -\gamma w^2\\
Z' &=-g Z,
\end{cases}
\end{split}$$ ]{}
where the coefficients $\beta_j$, $\gamma$ are positive. | time of integration is given by $ T=\ln\left (\frac{\eta_j}{v_j } \right)^{1/\beta_j}$. Such a time of consolidation is prevail from the equation $ v_j'=\beta_jv_j$. The we have [ $ $ \begin{split }
\tilde u & = uracil \left (\frac{v_j}{\eta_j } \right)^{1/\beta}\\
\tilde v_i & = v_i \left (\frac{\eta_j}{v_j } \right)^{\beta_i/\beta_j}\\
\tilde west & = w \left (\frac{\eta_j}{v_j } \right)^{\gamma/\beta_j }.
\end{split}$$ ] { }
It then only rests to calculate $ \tilde Z$. Following similar arguments as for the passage $ \Pi^w$ we get in this case [ $ $ \begin{split }
\tilde Z = Z\exp\left [ -\frac{\Lambda}{\gamma watt } \left (1+\tilde\alpha' w\ln(v_j)+w\tilde G'\right)\right ],
\end{split}$$ ] { }
where now [ $ $ \begin{split }
\tilde\alpha' & = \tilde\alpha'(uv_i^{1/\beta_i},uw^{1/\gamma})\\
\tilde G' & = \tilde G'(uv_i^{1/\beta_i},uw^{1/\gamma},\mu_w,\mu_i)
\end{split}$$ ] { }
are $ C^\ell$ functions with $ \mu_w = wv_j^{-\gamma/\beta_j}$ and $ \mu_i = v_iv_j^{-\beta_i/\beta_j}$.
Saddle 2 { # bomber: case2_.unnumbered }
--------
In this case we investigate the transitions of a vector field of the form [ $ $ \begin{split }
\label{eq: semihyp2 }
Y:\begin{cases }
u' & = w u \\
v_j' & = -\beta_j w v_j \\
w' & = -\gamma w^2\\
Z' & = -g Z,
\end{cases }
\end{split}$$ ] { }
where the coefficients $ \beta_j$, $ \gamma$ are convinced. | tile of integration is givtn by $T=\ln\left( \frae{\wta_j}{v_j} \right)^{1/\geta_j}$. Suzh a time of integration is lbrainee from the equation $v_j'=\ceta_jv_j$. Tje we hace [ $$\uegin{split}
\tilde u &= u \left( \rvac{v_j}{\zte_j} \right)^{1/\beta}\\
\tilce v_i &= v_i \neft( \frac{\eta_j}{v_b} \fiyht)^{\beta_i/\beta_j}\\
\tilde w &= w \left( \frac{\eta_t}{v_j} \rignt)^{\hamma/\beta_j}.
\egd{spkyt}$$ ]{}
If then only rests to compute $\tilde Z$. Follmwing similar arguments as for the tranditiln $\Pi^w$ we get in tjis case [ $$\bgfin{fplit}
\tildd Z=Z\exp\lefu[ -\yrac{\Lambda}{\gzmma w} \left( 1+\tilde\alpha' w\ln(v_j)+w\tiude G'\xight)\right],
\wnd{dklit}$$ ]{}
where niw [ $$\bvgin{split}
\bplde\alpva' &=\tildr\alpha'(uv_i^{1/\beta_i},mw^{1/\gamka})\\
\tulde G' &=\tilde G'(uv_i^{1/\beta_m},uw^{1/\gamma},\mu_w,\mu_i)
\end{fplit}$$ ]{}
are $C^\zll$ functions with $\mu_q=wc_j^{-\gamka/\beda_j}$ xbd $\ou_i=b_it_j^{-\bsta_i/\beha_j}$.
Aaddle 2 {#sug:case2_.unnumbwred}
--------
In this case we ignrstigate the transytyons of a vector field of the form [ $$\begpn{spmit}
\label{eq:semihyp2}
Y:\bwgin{cases}
u' &= w u \\
v_j' &= -\bgta_j w v_j \\
r' &= -\gamma w^2\\
Z' &=-g Z,
\end{cases}
\end{split}$$ ]{}
where the coefxicieitr $\bttq_j$, $\gaona$ are positive. | time of integration is given by $T=\ln\left( Such time of is obtained from have $$\begin{split} \tilde u u \left( \frac{v_j}{\eta_j} \tilde v_i &= v_i \left( \frac{\eta_j}{v_j} \tilde w &= w \left( \frac{\eta_j}{v_j} \right)^{\gamma/\beta_j}. \end{split}$$ ]{} It then only rests compute $\tilde Z$. Following similar arguments as for the transition $\Pi^w$ we get this [ \tilde -\frac{\Lambda}{\gamma w} \left( 1+\tilde\alpha' w\ln(v_j)+w\tilde G'\right)\right], \end{split}$$ ]{} where now [ $$\begin{split} \tilde\alpha' &=\tilde\alpha'(uv_i^{1/\beta_i},uw^{1/\gamma})\\ \tilde G' G'(uv_i^{1/\beta_i},uw^{1/\gamma},\mu_w,\mu_i) \end{split}$$ ]{} are $C^\ell$ functions with $\mu_w=wv_j^{-\gamma/\beta_j}$ $\mu_i=v_iv_j^{-\beta_i/\beta_j}$. Saddle 2 {#sub:case2_.unnumbered} In this case we investigate transitions a vector of form $$\begin{split} \label{eq:semihyp2} Y:\begin{cases} &= w u \\ v_j' &= -\beta_j w v_j \\ w' &= -\gamma w^2\\ Z' &=-g Z, \end{split}$$ ]{} coefficients $\beta_j$, are | time of integration is given bY $T=\ln\left( \frAc{\eta_J}{v_j} \RigHt)^{1/\Beta_J}$. SucH a time of integrATion Is obtained from the equatIon $v_j'=\BeTA_jv_j$. tHe We havE [ $$\begin{sPLiT}
\TIldE u &= U \lEft( \FrAC{v_J}{\eta_j} \RigHt)^{1/\beta}\\
\tIlde v_i &= v_i \leFt( \fRaC{\eta_j}{v_j} \right)^{\BEtA_i/\beta_j}\\
\tilDe w &= W \left( \frac{\eta_J}{v_j} \Right)^{\gAmMa/\bETa_j}.
\enD{spLit}$$ ]{}
It Then onLY rests To compute $\TiLDe Z$. FolLOwing siMILaR argUments as for the traNSiTIon $\Pi^w$ we get in tHis casE [ $$\bEGiN{SPliT}
\tiLde Z=Z\exp\leFt[ -\Frac{\LAMbda}{\gamMA w} \LEFT( 1+\tiLDe\alpha' w\ln(v_j)+w\Tilde G'\right)\RIghT],
\end{spLiT}$$ ]{}
whERe now [ $$\bEgin{sPlIT}
\tiLde\alpha' &=\tilDe\alPha'(uv_i^{1/\betA_i},uw^{1/\gaMMa})\\
\tilde g' &=\Tilde G'(uV_i^{1/\beta_I},uw^{1/\GamMa},\mu_W,\Mu_I)
\eNd{sPlIT}$$ ]{}
arE $c^\eLl$ fUNctIons with $\Mu_W=wV_j^{-\gamMa/\beTA_J}$ ANd $\mu_I=v_iV_j^{-\beTa_i/\beTa_j}$.
Saddle 2 {#sub:cAse2_.UnnuMBerEd}
--------
In tHis caSe we InVestiGate thE tranSiTions of a vector fIeld Of the form [ $$\BegIn{SplIt}
\Label{EQ:semihYp2}
Y:\BegIn{cases}
U' &= w u \\
v_j' &= -\beTA_j w V_j \\
W' &= -\GAMmA w^2\\
Z' &=-g Z,
\end{cases}
\end{spLiT}$$ ]{}
WHeRe the coeFficieNTs $\BeTA_j$, $\gamma$ aRe PosItivE. | time of integration is gi ven by $T= \ln\l eft ( \ fr ac{\ eta_ j}{v_j} \right ) ^{1/ \beta_j}$. Such a time of i nt e grat i on is o btained fr o m th eeq uat io n $ v_j'= \be ta_jv_j $. The wehav e[ $$\begin{s p li t}
\t ild e u &= u \le ft( \frac {v _j} { \eta_ j}\righ t)^{1/ \ beta}\ \
\tildev_ i &= v_ i \left( \ fr ac{\ eta_j}{v_j} \righ t )^ { \beta_i/\beta_ j}\\
\ ti l de w &= w\left( \fr ac {\eta _ j}{v_j} \r i g h t)^ { \gamma/\beta_ j}.
\en d {sp lit}$$ ] {}It the n onl yr est s to comput e $\ tilde Z$. Follo w ing sim i lar arg uments as fo r th e t ra nsi ti o n $ \ Pi ^w$ weget in t hi scase[ $$ \ b e g in{s pli t}
\ tilde Z=Z\exp \le ft[- \fr ac{\L ambda }{\g am ma w} \left ( 1+\ ti lde\alpha' w\ln (v_j )+w\tilde G' \r igh t) \righ t ],
\e nd{ split}$ $ ]{}
w her en o w [ $$\begin{split}
\t ilde\alp ha' &= \ ti ld e \alpha'( uv _i^ {1/\ b e ta_i} ,uw^ { 1/ \gamma}) \\
\ti l de G ' &=\ti ld e G'(u v_ i^{ 1/\ beta_ i },uw ^{1/\g amma},\m u_w,\ m u_i)
\end{ s plit}$$ ]{}
a re $ C^ \ ell$ fu nctions wit h $\ m u_w= wv_j ^ {- \ga m ma/\b eta_j }$ an d $\mu_i=v_iv_j^{-\b et a_i/\b eta_j }$.
Saddle 2 {#sub:cas e 2 _ .unnumbe red} -- - -----
In this case we invest i gate the tran sitionsof a vect o r field o f t hefor m [ $ $\ begin{split} \l ab el{eq:s emi hyp2}
Y :\b egi n{c ase s}
u' &= wu \\
v_j '&= - \b eta _j wv _j \\
w' & = - \g amm a w^2 \ \
Z' & =-g Z ,
\e nd {c a ses }
\ e nd { s plit }$ $]{}
wh er e the coe f fic ients $ \beta_j$, $\ g amma $ar e posit ive. | time_of integration_is given by $T=\ln\left(_\frac{\eta_j}{v_j} \right)^{1/\beta_j}$._Such_a time_of_integration is obtained_from the equation_$v_j'=\beta_jv_j$. The we have_[ $$\begin{split}
__
\tilde u &= u \left( \frac{v_j}{\eta_j} \right)^{1/\beta}\\
\tilde v_i &= v_i \left( \frac{\eta_j}{v_j} \right)^{\beta_i/\beta_j}\\
\tilde_w_&= w_\left(_\frac{\eta_j}{v_j}_\right)^{\gamma/\beta_j}.
\end{split}$$ ]{}
It_then only rests to compute_$\tilde Z$._Following similar arguments as for the transition $\Pi^w$_we_get in this_case [ $$\begin{split}
\tilde Z=Z\exp\left[ -\frac{\Lambda}{\gamma_w} \left( 1+\tilde\alpha' w\ln(v_j)+w\tilde G'\right)\right],
_ \end{split}$$ ]{}
where_now_[_$$\begin{split}
_
\tilde\alpha' &=\tilde\alpha'(uv_i^{1/\beta_i},uw^{1/\gamma})\\
\tilde G' &=\tilde G'(uv_i^{1/\beta_i},uw^{1/\gamma},\mu_w,\mu_i)
_ \end{split}$$ ]{}
are $C^\ell$ functions with_$\mu_w=wv_j^{-\gamma/\beta_j}$ and $\mu_i=v_iv_j^{-\beta_i/\beta_j}$.
Saddle 2 {#sub:case2_.unnumbered}
--------
In this case_we investigate the transitions of a_vector field of the form_[ $$\begin{split}
_ \label{eq:semihyp2}
Y:\begin{cases}
u' &=_w u \\
v_j'_&= -\beta_j_w v_j \\
w'_&= -\gamma w^2\\
Z' &=-g Z,
\end{cases}
_ \end{split}$$ ]{}
where_the coefficients $\beta_j$, $\gamma$ are positive. |
0^{+}$, when $S_n$ is sampled from the manifold ${\mathcal{M}}$.
Diffusion $K$-means {#sec:diffusion_Kmeans}
===================
Recall that in our clustering model, $S_n=\{X_{1},X_2,\ldots,X_n\}$ is a sample of independent random variables taking values in $S$, where $S$ is the union of $K$ disjoint Riemannian submanifolds ${\mathcal{D}}_{1},\dots,{\mathcal{D}}_{K}$ embedded in the ambient space ${\mathbb{R}}^p$. The clustering problem is to divide these $n$ data points into $K$ clusters, so that points in the same cluster belongs to the same connected component in $S$, based on certain similarity measures between the points. In particular, the (classical) $K$-means clustering method minimizes the total intra-cluster squared Euclidean distances in ${\mathbb{R}}^p$ $$\min_{G_{1},\dots,G_{K}} \sum_{k=1}^{K} {1 \over |G_{k}|} \sum_{i,j \in G_{k}} \|X_{i}-X_{j}\|^{2}$$ over all possible partitions on $[n]$, where $|G_{k}|$ is the cardinality of $G_{k}$. Dropping the sum of squared norms $\sum_{i=1}^{n} \|X_{i}\|^{2}$, we see that the $K$-means clustering is equivalent to the maximization of the total within-cluster covariances $$\max_{G_{1},\dots,G_{K}} \sum_{k=1}^{K} {1 \over |G_{k}|} \sum_{i,j \in G_{k}} a_{ij}, \quad\mbox{with }a_{ij} = X_{i}^{T} X_{j}.$$ Here, $a_{ij}=X_{i}^{T} X_{j}$ can be viewed as a similarity measure specified by the Euclidean space inner product $\langle X_{i}, X_{j}\rangle_{{\mathbb{R}}^p}$. In general, we can replace the Euclidean inner product with any other inner product over $S_n$ [@ChenYang2018]. For manifold clustering, we replace it with the inner product induced from the empirical diffusion distance, that is, $$\begin{aligned}
\langle x,\, y\rangle_{D_{n,t}} = \ | 0^{+}$, when $ S_n$ is sampled from the manifold $ { \mathcal{M}}$.
Diffusion $ K$-means { # sec: diffusion_Kmeans }
= = = = = = = = = = = = = = = = = = =
Recall that in our clustering model, $ S_n=\{X_{1},X_2,\ldots, X_n\}$ is a sample of autonomous random variable star taking values in $ S$, where $ S$ is the coupling of $ K$ disjoint Riemannian submanifolds $ { \mathcal{D}}_{1},\dots,{\mathcal{D}}_{K}$ embed in the ambient space $ { \mathbb{R}}^p$. The clustering problem is to separate these $ n$ datum points into $ K$ clusters, so that points in the same cluster belong to the same connected part in $ S$, based on certain similarity measures between the points. In particular, the (classical) $ K$-means clustering method acting minimizes the total intra - cluster squared euclidian distances in $ { \mathbb{R}}^p$ $ $ \min_{G_{1},\dots, G_{K } } \sum_{k=1}^{K } { 1 \over |G_{k}| } \sum_{i, j \in G_{k } } \|X_{i}-X_{j}\|^{2}$$ over all possible partitions on $ [ n]$, where $ |G_{k}|$ is the cardinality of $ G_{k}$. Dropping the sum of squared norm $ \sum_{i=1}^{n } \|X_{i}\|^{2}$, we see that the $ K$-means clustering is equivalent to the maximization of the total within - cluster covariances $ $ \max_{G_{1},\dots, G_{K } } \sum_{k=1}^{K } { 1 \over |G_{k}| } \sum_{i, j \in G_{k } } a_{ij }, \quad\mbox{with } a_{ij } = X_{i}^{T } X_{j}.$$ Here, $ a_{ij}=X_{i}^{T } X_{j}$ can be watch as a similarity measure specified by the Euclidean space inner product $ \langle X_{i }, X_{j}\rangle_{{\mathbb{R}}^p}$. In general, we can replace the Euclidean inner product with any other inner product over $ S_n$ [ @ChenYang2018 ]. For manifold clustering, we replace it with the inner merchandise induced from the empirical dispersion distance, that is, $ $ \begin{aligned }
\langle x,\, y\rangle_{D_{n, t } } = \ | 0^{+}$, wjen $S_n$ is sampled from tme manifold ${\matheql{M}}$.
Difhusion $I$-means {#sdc:diffusion_Kmeans}
===================
Recall that ib our clustering model, $S_n=\{X_{1},X_2,\udots,X_n\}$ id a sampoe oh independent raisom varlcbles bakiny talues in $S$, whete $S$ is the gnion of $K$ disboknc Riemannian submanifolds ${\mathcal{D}}_{1},\does,{\mathcsl{F}}_{K}$ embedded in the wmbisnt space ${\mathbb{R}}^p$. The clustering lroblem is to divide these $n$ data points into $N$ clksters, so that poijts in the wame xluster belovgs to the same connecjed component in $S$, based on certxin snmilarity mgcwurfv between tie poigts. In partigllar, tha (classocal) $K$-means clmstermng nethod minimizes the votal intra-cluster szuared Euwlndean distances in ${\marhvb{R}}^p$ $$\kin_{G_{1},\gots,E_{J}} \sjm_{k=1}^{I} {1 \ober |G_{k}|} \suj_{i,j \in G_{k}} \|S_{i}-X_{j}\|^{2}$$ over aol possible partitipnf on $[n]$, where $|G_{i}|$ is tre cardinality of $G_{k}$. Dropping the sum of vquzred norms $\sum_{i=1}^{n} \|X_{i}\|^{2}$, we wee that the $K$-means cpustering is equivalent to the maximization of the total whthin-rljsttr covafuajces $$\max_{G_{1},\dots,G_{K}} \sum_{k=1}^{K} {1 \over |G_{k}|} \sum_{i,j \in G_{k}} a_{it}, \qiac\mbox{with }a_{ij} = X_{i}^{T} X_{j}.$$ Herr, $w_{ik}=V_{i}^{T} X_{j}$ can be vieweb aa a similarity meadure spgcifiee by the Tuclicean space inner product $\labgle X_{i}, X_{j}\rauglw_{{\mathbb{R}}^p}$. In genercl, we can re'lace jhe Euvlidean inner product wnth ang other innfr producf over $S_n$ [@ChenYane2018]. Fpr manifold clustering, we re[lace it xith che innef prpduct ynduced frlm thc empirical diffusiln didtdnce, that ls, $$\begin{aligned}
\langle x,\, y\rangle_{V_{i,t}} = \ | 0^{+}$, when $S_n$ is sampled from the Diffusion {#sec:diffusion_Kmeans} =================== that in our sample independent random variables values in $S$, $S$ is the union of $K$ Riemannian submanifolds ${\mathcal{D}}_{1},\dots,{\mathcal{D}}_{K}$ embedded in the ambient space ${\mathbb{R}}^p$. The clustering problem is divide these $n$ data points into $K$ clusters, so that points in the cluster to same component in $S$, based on certain similarity measures between the points. In particular, the (classical) $K$-means method minimizes the total intra-cluster squared Euclidean distances ${\mathbb{R}}^p$ $$\min_{G_{1},\dots,G_{K}} \sum_{k=1}^{K} {1 |G_{k}|} \sum_{i,j \in G_{k}} \|X_{i}-X_{j}\|^{2}$$ all partitions on where is cardinality of $G_{k}$. the sum of squared norms $\sum_{i=1}^{n} \|X_{i}\|^{2}$, we see that the $K$-means clustering is equivalent to the of the covariances $$\max_{G_{1},\dots,G_{K}} {1 |G_{k}|} \in G_{k}} a_{ij}, = X_{i}^{T} X_{j}.$$ Here, $a_{ij}=X_{i}^{T} X_{j}$ as a similarity measure specified by the Euclidean inner product X_{i}, X_{j}\rangle_{{\mathbb{R}}^p}$. In general, we can the Euclidean inner product with any other inner over $S_n$ [@ChenYang2018]. For manifold clustering, we replace it with the inner product induced from diffusion distance, that is, \langle x,\, y\rangle_{D_{n,t}} \ | 0^{+}$, when $S_n$ is sampled from the manIfold ${\mathcAl{M}}$.
DiFfuSioN $K$-MeanS {#sec:Diffusion_KmeanS}
===================
recaLl that in our clustering mOdel, $S_N=\{X_{1},x_2,\LdotS,x_n\}$ Is a saMple of iNDePENdeNt RaNdoM vARiAbles TakIng valuEs in $S$, where $s$ is ThE union of $K$ disJOiNt RiemanniAn sUbmanifolds ${\mAthCal{D}}_{1},\doTs,{\MatHCal{D}}_{K}$ EmbEdded In the aMBient sPace ${\mathbB{R}}^P$. the cluSTering pROBlEm is To divide these $n$ datA PoINts into $K$ clusteRs, so thAt POiNTS in The Same clusteR bElongS To the saME cONNEctED component in $S$, Based on certAIn sImilarItY meASures bEtweeN tHE poInts. In partiCulaR, the (classIcal) $K$-mEAns clusTEring meThod miNimIzeS the TOtAl IntRa-CLusTEr SquARed euclideaN dIsTanceS in ${\mATHBB{R}}^p$ $$\mIn_{G_{1},\Dots,g_{K}} \sum_{K=1}^{K} {1 \over |G_{k}|} \sum_{i,j \In G_{K}} \|X_{i}-X_{J}\|^{2}$$ OveR all pOssibLe paRtItionS on $[n]$, whEre $|G_{k}|$ Is The cardinality oF $G_{k}$. DRopping thE suM oF sqUaRed noRMs $\sum_{i=1}^{N} \|X_{i}\|^{2}$, We sEe that tHe $K$-meanS CluStERINg Is equivalent to the mAxIMIzAtion of tHe totaL WiThIN-cluster CoVarIancES $$\Max_{G_{1},\dOts,G_{k}} \SuM_{k=1}^{K} {1 \over |G_{K}|} \sum_{i,j \IN G_{K}} a_{Ij}, \quad\mBoX{with }a_{Ij} = x_{i}^{T} x_{j}.$$ HEre, $a_{iJ}=x_{i}^{T} X_{J}$ can be Viewed as A simiLArity measure spECified by the EuCLiDEAn SPace InnEr product $\laNgle x_{I}, X_{j}\rAnglE_{{\MaThbB{r}}^p}$. In gEneraL, wE CaN Replace the Euclidean InNer proDuct wIth any other inNer product OVER $S_n$ [@ChenYAng2018]. FOR mANifold clusteriNg, we rEplace it wiTH the inneR prodUct inducEd from the EMPirical dIffUsiOn dIstANCe, That is, $$\begin{alIGNed}
\lAnGle x,\, y\raNglE_{D_{n,t}} = \ | 0^{+}$, when $S_n$ is sam pled fromthe m ani fol d${\m athc al{M}}$.
Diff u sion $K$-means {#sec:diffu sion_ Km e ans} == ===== ======= = == = =
R ec al l t ha t i n our cl usterin g model, $ S_n =\ {X_{1},X_2,\ l do ts,X_n\}$isa sample ofind epende nt ra n dom v ari ables takin g value s in $S$, w h ere $S $ is the u ni on o f $K$ disjoint Ri e ma n nian submanifo lds ${ \m a th c a l{D }}_ {1},\dots, {\ mathc a l{D}}_{ K }$ e m bed d ed in the amb ient space$ {\m athbb{ R} }^p $ . Theclust er i ngproblem isto d ivide the se $n$ data po i nts int o $K$clu ste rs,s oth atpo i nts in th e sa me clust er b elong s to t h e sam e c onne ctedcomponent in$S$ , ba s edon ce rtain sim il arity measu res b et ween the points . In particul ar, t he(c lassi c al) $K $-m ean s clust ering m e tho dm i n im izes the total int ra - c lu ster squ ared E u cl id e an dista nc esin $ { \ mathb b{R} } ^p $ $$\min _{G_{1 } ,\ do ts,G_{K }} \sum_ {k =1} ^{K } {1\ over |G_{k }|} \sum _{i,j \in G_{k}} \|X _ {i}-X_{j}\|^{ 2 }$ $ ov e r al l p ossible par titi o ns o n $[ n ]$ , w h ere $ |G_{k }| $ i s the cardinality of $ G_{k}$ . Dro pping the sum of square d n orms $\s um_{ i =1 } ^{n} \|X_{i}\| ^{2}$ , we see t h at the $ K$-me ans clus tering is e quivalen t t o t hemax i m iz ation of thet o talwi thin-cl ust er cova ria nce s $ $\m ax _{G_{1},\ dots,G_{ K} }\s um _{k =1}^{ K } {1 \ov er |G _{ k}| } \su m _{i,j\in G _{k} }a_ { ij} , \quad \ mb o x {wit h}a _{ij } = X _{i}^ {T}X _{j }.$$ He re, $a_{i j}= X _{i} ^{ T} X_{j}$ can be viewe das a simil ar ity measu r e specifi ed by the Euclidean spa c e inner pr oduct $\l angle X_{ i}, X_{j} \ra n gle_{{ \mathb b{R}} ^p }$. I n gen e r al , w ecan replac e the Eucl id eaninner p roduct with any ot h erinner product ov er $ S _ n$ [@ C he n Yan g2 0 18] . For manifold cl ustering,we re place it w i thth e inner produc t ind u ced fro m the emp irical di ff usio n dis tance, tha t is, $$ \begin{al i gned} \l angle x, \, y\r an gle _{D_{ n,t}}= \ | 0^{+}$,_when $S_n$_is sampled from the_manifold ${\mathcal{M}}$.
Diffusion_$K$-means_{#sec:diffusion_Kmeans}
===================
Recall that_in_our clustering model,_$S_n=\{X_{1},X_2,\ldots,X_n\}$ is a_sample of independent random_variables taking values_in_$S$, where $S$ is the union of $K$ disjoint Riemannian submanifolds ${\mathcal{D}}_{1},\dots,{\mathcal{D}}_{K}$ embedded in_the_ambient space_${\mathbb{R}}^p$._The_clustering problem is to divide_these $n$ data points into_$K$ clusters,_so that points in the same cluster belongs_to_the same connected_component in $S$, based on certain similarity measures between_the points. In particular, the (classical)_$K$-means clustering method_minimizes_the_total intra-cluster squared Euclidean_distances in ${\mathbb{R}}^p$ $$\min_{G_{1},\dots,G_{K}} \sum_{k=1}^{K} {1_\over |G_{k}|} \sum_{i,j \in G_{k}} \|X_{i}-X_{j}\|^{2}$$_over all possible partitions on $[n]$, where_$|G_{k}|$ is the cardinality of $G_{k}$._Dropping the sum of squared_norms $\sum_{i=1}^{n}_\|X_{i}\|^{2}$, we see that the_$K$-means clustering is_equivalent to_the maximization of_the total within-cluster covariances $$\max_{G_{1},\dots,G_{K}} \sum_{k=1}^{K}_{1 \over |G_{k}|}_\sum_{i,j \in G_{k}} a_{ij}, \quad\mbox{with }a_{ij}_=_X_{i}^{T} X_{j}.$$ Here,_$a_{ij}=X_{i}^{T}_X_{j}$_can be_viewed as a_similarity_measure specified_by_the Euclidean space inner product $\langle_X_{i},_X_{j}\rangle_{{\mathbb{R}}^p}$. In general, we can replace the_Euclidean inner product with_any_other inner product over_$S_n$ [@ChenYang2018]. For manifold clustering,_we replace it with the inner_product induced_from the_empirical diffusion distance, that is, $$\begin{aligned}
\langle x,\, y\rangle_{D_{n,t}} = \ |
_1 \right\rangle}-{\left | 1 1_01_1 \right\rangle}
\end{array}\right\},\end{aligned}$$
$$\begin{aligned}
B_6&=&\frac{1}{2\sqrt{2}}\left(\begin{array}{cccccccc}
1&1&1&1&1&1&1&1\\
i&-i&i&-i&i&-i&i&-i\\
1&1&-1&-1&1&1&-1&-1\\
-i&i&i&-i&-i&i&i&-i\\
1&1&1&1&-1&-1&-1&-1\\
-i&i&-i&i&i&-i&i&-i\\
-1&-1&1&1&1&1&-1&-1\\
i&-i&-i&i&-i&i&i&-i
\end{array}\right)
\\
& = &
\frac{1}{\sqrt{2}}\left\{\begin{array}{c}
|0 0_00_1\rangle+i|1 1_01_1\rangle\\
|0 0_01_1\rangle+i|1 1_00_1\rangle\\
|0 1_00_1\rangle+i|1 0_01_1\rangle\\
|0 1_01_1\rangle+i|1 0_00_1\rangle\\
|0 0_00_1\rangle-i|1 1_01_1\rangle\\
|0 0_01_1\rangle-i|1 1_00_1\rangle\\
|0 1_00_1\rangle-i|1 0_01_1\rangle\\
|0 1_01_1\rangle-i|1 0_00_1\rangle\end{array}\right\},\end{aligned}$$
$$B_7=\frac{1}{2\sqrt{2}}\left(\begin{array}{cccccccc}
1&1&1&1&1&1&1&1\\
1&-1&1&-1&1&-1&1&-1\\
1&1&-1&-1&1&1&-1&-1\\
1&-1&-1&1&1&-1&-1 | _ 1 \right\rangle}-{\left | 1 1_01_1 \right\rangle }
\end{array}\right\},\end{aligned}$$
$ $ \begin{aligned }
B_6&=&\frac{1}{2\sqrt{2}}\left(\begin{array}{cccccccc }
1&1&1&1&1&1&1&1\\
i&-i&i&-i&i&-i&i&-i\\
1&1&-1&-1&1&1&-1&-1\\
-i&i&i&-i&-i&i&i&-i\\
1&1&1&1&-1&-1&-1&-1\\
-i&i&-i&i&i&-i&i&-i\\
-1&-1&1&1&1&1&-1&-1\\
i&-i&-i&i&-i&i&i&-i
\end{array}\right)
\\
& = &
\frac{1}{\sqrt{2}}\left\{\begin{array}{c }
|0 0_00_1\rangle+i|1 1_01_1\rangle\\
|0 0_01_1\rangle+i|1 1_00_1\rangle\\
|0 1_00_1\rangle+i|1 0_01_1\rangle\\
|0 1_01_1\rangle+i|1 0_00_1\rangle\\
|0 0_00_1\rangle - i|1 1_01_1\rangle\\
|0 0_01_1\rangle - i|1 1_00_1\rangle\\
|0 1_00_1\rangle - i|1 0_01_1\rangle\\
|0 1_01_1\rangle - i|1 0_00_1\rangle\end{array}\right\},\end{aligned}$$
$ $ B_7=\frac{1}{2\sqrt{2}}\left(\begin{array}{cccccccc }
1&1&1&1&1&1&1&1\\
1&-1&1&-1&1&-1&1&-1\\
1&1&-1&-1&1&1&-1&-1\\
1&-1&-1&1&1&-1&-1 | _1 \rihht\rangle}-{\left | 1 1_01_1 \right\rakgle}
\end{array}\righj\},\ebd{aligied}$$
$$\begih{aligned}
C_6&=&\frac{1}{2\sqrt{2}}\left(\begin{array}{cccccrcc}
1&1&1&1&1&1&1&1\\
i&-u&i&-i&i&-i&u&-i\\
1&1&-1&-1&1&1&-1&-1\\
-i&i&i&-i&-i&i&i&-i\\
1&1&1&1&-1&-1&-1&-1\\
-i&i&-i&i&i&-i&i&-i\\
-1&-1&1&1&1&1&-1&-1\\
i&-i&-i&i&-k&i&i&-i
\end{argay}\right)
\\
& = &
\frar{1}{\sqrt{2}}\left\{\begin{arczy}{c}
|0 0_00_1\rannje+i|1 1_01_1\dwnglz\\
|0 0_01_1\cangle+i|1 1_00_1\rangle\\
|0 1_00_1\tangle+i|1 0_01_1\rangne\\
|0 1_01_1\rangle+i|1 0_00_1\rancld\\
|0 0_00_1\xangle-i|1 1_01_1\rangle\\
|0 0_01_1\rangle-i|1 1_00_1\rangle\\
|0 1_00_1\rangle-y|1 0_01_1\ranglr\\
|0 1_01_1\gangle-i|1 0_00_1\rangle\gnd{argar}\riggn\},\ekd{aligned}$$
$$B_7=\frac{1}{2\sqrt{2}}\left(\begin{array}{dccccccb}
1&1&1&1&1&1&1&1\\
1&-1&1&-1&1&-1&1&-1\\
1&1&-1&-1&1&1&-1&-1\\
1&-1&-1&1&1&-1&-1 | _1 \right\rangle}-{\left | 1 1_01_1 \right\rangle} \end{array}\right\},\end{aligned}$$ 1&1&1&1&1&1&1&1\\ 1&1&-1&-1&1&1&-1&-1\\ -i&i&i&-i&-i&i&i&-i\\ -i&i&-i&i&i&-i&i&-i\\ -1&-1&1&1&1&1&-1&-1\\ i&-i&-i&i&-i&i&i&-i \frac{1}{\sqrt{2}}\left\{\begin{array}{c} 0_00_1\rangle+i|1 1_01_1\rangle\\ |0 1_00_1\rangle\\ |0 1_00_1\rangle+i|1 |0 1_01_1\rangle+i|1 0_00_1\rangle\\ |0 0_00_1\rangle-i|1 1_01_1\rangle\\ 0_01_1\rangle-i|1 1_00_1\rangle\\ |0 1_00_1\rangle-i|1 0_01_1\rangle\\ |0 1_01_1\rangle-i|1 0_00_1\rangle\end{array}\right\},\end{aligned}$$ $$B_7=\frac{1}{2\sqrt{2}}\left(\begin{array}{cccccccc} 1&1&1&1&1&1&1&1\\ 1&-1&1&-1&1&-1&1&-1\\ 1&1&-1&-1&1&1&-1&-1\\ 1&-1&-1&1&1&-1&-1 | _1 \right\rangle}-{\left | 1 1_01_1 \right\ranglE}
\end{array}\rIght\},\eNd{aLigNeD}$$
$$\begIn{alIgned}
B_6&=&\frac{1}{2\sqrt{2}}\LEft(\bEgin{array}{cccccccc}
1&1&1&1&1&1&1&1\\
i&-i&i&-i&i&-I&i&-i\\
1&1&-1&-1&1&1&-1&-1\\
-i&i&I&-i&-I&I&i&-i\\
1&1&1&1&-1&-1&-1&-1\\
-i&I&-I&i&I&-i&i&-i\\
-1&-1&1&1&1&1&-1&-1\\
i&-I&-i&i&-i&i&i&-i
\ENd{ARRay}\RiGhT)
\\
& = &
\frAc{1}{\SQrT{2}}\left\{\BegIn{array}{C}
|0 0_00_1\rangle+i|1 1_01_1\raNglE\\
|0 0_01_1\rAngle+i|1 1_00_1\rangle\\
|0 1_00_1\RAnGle+i|1 0_01_1\rangle\\
|0 1_01_1\RanGle+i|1 0_00_1\rangle\\
|0 0_00_1\raNglE-i|1 1_01_1\rangLe\\
|0 0_01_1\RanGLe-i|1 1_00_1\raNglE\\
|0 1_00_1\rangLe-i|1 0_01_1\ranGLe\\
|0 1_01_1\rangLe-i|1 0_00_1\rangle\EnD{Array}\rIGht\},\end{aLIGnEd}$$
$$B_7=\fRac{1}{2\sqrt{2}}\left(\begin{aRRaY}{Cccccccc}
1&1&1&1&1&1&1&1\\
1&-1&1&-1&1&-1&1&-1\\
1&1&-1&-1&1&1&-1&-1\\
1&-1&-1&1&1&-1&-1 | _1 \right\rangle}-{\left | 1 1_01_1\righ t\r ang le }
\e nd{a rray}\right\}, \ end{ aligned}$$
$$\begin{a ligne d} B_6& = &\ frac{ 1}{2\sq r t{ 2 } }\l ef t( \be gi n {a rray} {cc cccccc}
1&1&1&1&1 &1& 1& 1\\
i&-i&i&- i &i &-i&i&-i\\
1& 1&-1&-1&1&1& -1& -1\\
- i& i&i & -i&-i &i& i&-i\ \
1&1& 1 &1&-1& -1&-1&-1\ \- i&i&-i & i&i&-i& i & -i \\
- 1&-1&1&1&1&1&-1&- 1 \\ i&-i&-i&i&-i&i &i&-i\e n d{ a r ray }\r ight)
\\
& = &
\f r ac{1}{\ s qr t { 2 }}\ l eft\{\begin{a rray}{c}
|0 0_0 0_1\ra ng le+ i |1 1_0 1_1\r an g le\ \
|0 0_01_1 \ran gle+i|1 1 _00_1\ r angle\\ |0 1_00 _1\ran gle +i| 1 0_ 0 1_ 1\ ran gl e \\| 01_0 1 _1\ rangle+i |1 0 _00_1 \ran g l e \ \
|0 0_ 00_1 \rang le-i|1 1_01_1 \ra ngle \ \
| 0 0_0 1_1\r angl e- i|1 1 _00_1\ rangl e\ \
|0 1_00_1\ran gle- i|1 0_01_ 1\r an gle \\
|0 1 _ 01_1\r ang le- i|1 0_0 0_1\ran g le\ en d { a rr ay}\right\},\end{a li g n ed }$$
$$B _7=\fr a c{ 1} { 2\sqrt{2 }} \le ft(\ b e gin{a rray } {c ccccccc}
1&1&1 & 1& 1& 1&1&1\\
1 &-1&1& -1 &1& -1& 1&-1\ \
1&1 &-1&-1 &1&1&-1& -1\\1 &-1&-1&1&1&-1& - 1 | _1 \right\rangle}-{\left_| 1_1_01_1 \right\rangle}
\end{array}\right\},\end{aligned}$$
$$\begin{aligned}
B_6&=&\frac{1}{2\sqrt{2}}\left(\begin{array}{cccccccc}
1&1&1&1&1&1&1&1\\
i&-i&i&-i&i&-i&i&-i\\
1&1&-1&-1&1&1&-1&-1\\
-i&i&i&-i&-i&i&i&-i\\
1&1&1&1&-1&-1&-1&-1\\
-i&i&-i&i&i&-i&i&-i\\
-1&-1&1&1&1&1&-1&-1\\
i&-i&-i&i&-i&i&i&-i
\end{array}\right)
\\
& = &
\frac{1}{\sqrt{2}}\left\{\begin{array}{c}
|0_0_00_1\rangle+i|1 1_01_1\rangle\\
|0_0_01_1\rangle+i|1_1_00_1\rangle\\
|0 1_00_1\rangle+i|1_0_01_1\rangle\\
|0_1_01_1\rangle+i|1 0_00_1\rangle\\
|0 0_00_1\rangle-i|1_1_01_1\rangle\\
|0 0_01_1\rangle-i|1 1_00_1\rangle\\
|0_1_00_1\rangle-i|1 0_01_1\rangle\\
|0 1_01_1\rangle-i|1 0_00_1\rangle\end{array}\right\},\end{aligned}$$
$$B_7=\frac{1}{2\sqrt{2}}\left(\begin{array}{cccccccc}
1&1&1&1&1&1&1&1\\
1&-1&1&-1&1&-1&1&-1\\
1&1&-1&-1&1&1&-1&-1\\
1&-1&-1&1&1&-1&-1 |
}(F(\omega_1,\omega_2)): (\omega_1,\omega_2) \in {{\mathcal M}}\right\},\end{aligned}$$ where ${{\mathcal M}}$ is the uniform sampling grids defined in.
Second, we show that the accumulated difference of the singular values between ${{{\uppercase{{\bm{c}}}}}}$ and ${{{\uppercase{{\bm{t}}}}}}$ is upper-bounded.
By inspecting ${{{\uppercase{{\bm{t}}}}}}$ and ${{{\uppercase{{\bm{c}}}}}}$, we find from Lemma \[lemma:proof-thm1\] that ${\hbox{\boldmath$\Delta$}}_{k,l}=0$ if and only if $(k,l) \in {{\mathcal B}}_{11}$. The number of rows and columns with indices outside ${{\mathcal B}}_{11}$ scales as $n$. As such, invoking Theorem 3.1 in [@Zizler2002], we conclude that $$\begin{aligned}
\sum_{j=1}^{\min\{r,s\}n^2} \abs{\sigma_j({{{\uppercase{{\bm{t}}}}}})-\sigma_j({{{\uppercase{{\bm{c}}}}}})} \le O(n).\end{aligned}$$ Thus, we have $$\begin{aligned}
\lim_{n \to \infty} \frac{1}{n}\sum_{j=1}^{\min\{r,s\}n^2} \abs{\sigma_j({{{\uppercase{{\bm{t}}}}}})-\sigma_j({{{\uppercase{{\bm{c}}}}}})}=O(1)\end{aligned}$$ This completes the proof.
The intuition behind is that the number of different elements between two matrices scales as $n$ but not $n^2$ because of the banded structure of ${{{\uppercase{{\bm{c}}}}}}$ and ${{{\uppercase{{\bm{t}}}}}}$. Although not rigorously proved, it looks the equality holds with the term $O(1)$ strictly larger than 0, meaning that the circular approximation can be arbitrarily loose as $n$ tends to infinity.
Proof of Theorem 3
------------------
Let $\phi_j:[-\pi,\pi]^2 \mapsto {\mbox{\bb R}}_+$ be the $j$-th singular value function of $F({\hbox{\boldmath$\omega$}})$ and $\sigma_k^{(j)}({{{\uppercase{{\bm{t}}}}}})$ be $k$-th singular | } (F(\omega_1,\omega_2) ): (\omega_1,\omega_2) \in { { \mathcal M}}\right\},\end{aligned}$$ where $ { { \mathcal M}}$ is the uniform sampling grids defined in.
Second, we picture that the roll up difference of the singular values between $ { { { \uppercase{{\bm{c}}}}}}$ and $ { { { \uppercase{{\bm{t}}}}}}$ is upper - bound.
By inspecting $ { { { \uppercase{{\bm{t}}}}}}$ and $ { { { \uppercase{{\bm{c}}}}}}$, we find from Lemma \[lemma: proof - thm1\ ] that $ { \hbox{\boldmath$\Delta$}}_{k, l}=0 $ if and entirely if $ (k, l) \in { { \mathcal B}}_{11}$. The issue of rows and column with indices outside $ { { \mathcal B}}_{11}$ scales as $ n$. As such, appeal Theorem 3.1 in [ @Zizler2002 ], we conclude that $ $ \begin{aligned }
\sum_{j=1}^{\min\{r, s\}n^2 } \abs{\sigma_j({{{\uppercase{{\bm{t}}}}}})-\sigma_j({{{\uppercase{{\bm{c } } } } } }) } \le O(n).\end{aligned}$$ Thus, we accept $ $ \begin{aligned }
\lim_{n \to \infty } \frac{1}{n}\sum_{j=1}^{\min\{r, s\}n^2 } \abs{\sigma_j({{{\uppercase{{\bm{t}}}}}})-\sigma_j({{{\uppercase{{\bm{c}}}}}})}=O(1)\end{aligned}$$ This completes the proof.
The intuition behind is that the number of different elements between two matrix scales as $ n$ but not $ n^2 $ because of the banded structure of $ { { { \uppercase{{\bm{c}}}}}}$ and $ { { { \uppercase{{\bm{t}}}}}}$. Although not rigorously prove, it looks the equality holds with the term $ O(1)$ strictly bigger than 0, meaning that the circular approximation can be arbitrarily loose as $ n$ tend to infinity.
Proof of Theorem 3
------------------
Let $ \phi_j:[-\pi,\pi]^2 \mapsto { \mbox{\bb R}}_+$ be the $ j$-th singular value function of $ F({\hbox{\boldmath$\omega$}})$ and $ \sigma_k^{(j)}({{{\uppercase{{\bm{t}}}}}})$ be $ k$-th singular | }(F(\omfga_1,\omega_2)): (\omega_1,\omega_2) \in {{\mxthcal M}}\right\},\enb{qlignev}$$ where ${{\mathcal M}}$ is the uniform sampling gcids defibed in.
Second, we show tfat the abcumulatee dihference of the singular valuea betcexn ${{{\uppercase{{\bm{c}}}}}}$ and ${{{\upperwase{{\bm{t}}}}}}$ is uppar-cobnded.
By inspecting ${{{\uppercase{{\bm{t}}}}}}$ and ${{{\tppercaxe{{\hm{c}}}}}}$, we find frjm Ltmmw \[lejma:proof-thm1\] that ${\hbox{\boldmath$\Delta$}}_{i,l}=0$ if aid only if $(k,l) \im {{\mathcal B}}_{11}$. The number of gows and columns with lndices outwide ${{\nathcal B}}_{11}$ scxles as $n$. As such, invoiing Theorem 3.1 in [@Zizler2002], we concuude chat $$\begin{aoitnef}
\sum_{j=1}^{\min\{c,s\}n^2} \abf{\sigma_j({{{\uppergsse{{\bm{t}}}}}})-\vigma_j({{{\ulpercase{{\bm{c}}}}}})} \le O(n).\xnd{aoigned}$$ Thus, we have $$\bxgin{aligned}
\lim_{n \jo \infty} \fxac{1}{n}\sum_{j=1}^{\min\{r,s\}n^2} \abs{\signa_h({{{\uppetcase{{\tm{t}}}}}})-\sktma_g({{{\upkerrass{{\bm{c}}}}}})}=O(1)\ejd{amigned}$$ Thia completes the proof.
The intuiuiog behind is thzt the ntmber of different elements between two mafrices scales as $n$ but bot $n^2$ because of the handed stwucture of ${{{\uppercase{{\bm{c}}}}}}$ and ${{{\uppercase{{\bm{t}}}}}}$. Although not cieoriufut oroved, it looks the equality holds with the tqdm $O(1)$ strictly largcr than 0, meaning tnah yre circular akproximcfikn can be arbitrarlly loofe as $n$ tends uo inginity.
Proof of Theorem 3
------------------
Let $\phi_j:[-\pi,\pi]^2 \mapfro {\mbox{\bb R}}_+$ be the $j$-th singulcr valoe funvtion of $F({\hbox{\boldmath$\ooega$}})$ and $\sigma_k^{(u)}({{{\uppercass{{\cm{t}}}}}})$ be $k$-th singuuar | }(F(\omega_1,\omega_2)): (\omega_1,\omega_2) \in {{\mathcal M}}\right\},\end{aligned}$$ where ${{\mathcal the sampling grids in. Second, we of singular values between and ${{{\uppercase{{\bm{t}}}}}}$ is By inspecting ${{{\uppercase{{\bm{t}}}}}}$ and ${{{\uppercase{{\bm{c}}}}}}$, we from Lemma \[lemma:proof-thm1\] that ${\hbox{\boldmath$\Delta$}}_{k,l}=0$ if and only if $(k,l) \in {{\mathcal B}}_{11}$. number of rows and columns with indices outside ${{\mathcal B}}_{11}$ scales as $n$. such, Theorem in we conclude that $$\begin{aligned} \sum_{j=1}^{\min\{r,s\}n^2} \abs{\sigma_j({{{\uppercase{{\bm{t}}}}}})-\sigma_j({{{\uppercase{{\bm{c}}}}}})} \le O(n).\end{aligned}$$ Thus, we have $$\begin{aligned} \lim_{n \to \infty} \frac{1}{n}\sum_{j=1}^{\min\{r,s\}n^2} This completes the proof. The intuition behind is the number of different between two matrices scales as but $n^2$ because the structure ${{{\uppercase{{\bm{c}}}}}}$ and ${{{\uppercase{{\bm{t}}}}}}$. not rigorously proved, it looks the equality holds with the term $O(1)$ strictly larger than 0, meaning the circular be arbitrarily as tends infinity. Proof of ------------------ Let $\phi_j:[-\pi,\pi]^2 \mapsto {\mbox{\bb R}}_+$ singular value function of $F({\hbox{\boldmath$\omega$}})$ and $\sigma_k^{(j)}({{{\uppercase{{\bm{t}}}}}})$ be singular | }(F(\omega_1,\omega_2)): (\omega_1,\omega_2) \in {{\maThcal M}}\righT\},\end{aLigNed}$$ WhEre ${{\mAthcAl M}}$ is the uniforM SampLing grids defined in.
SecoNd, we sHoW That THe AccumUlated dIFfEREncE oF tHe sInGUlAr valUes Between ${{{\Uppercase{{\bM{c}}}}}}$ aNd ${{{\Uppercase{{\bm{t}}}}}}$ IS uPper-boundeD.
By Inspecting ${{{\upPerCase{{\bm{T}}}}}}$ aNd ${{{\uPPercaSe{{\bM{c}}}}}}$, we fInd froM lemma \[lEmma:proof-ThM1\] That ${\hbOX{\boldmaTH$\deLta$}}_{k,L}=0$ if and only if $(k,l) \in {{\mAThCAl B}}_{11}$. The number of Rows anD cOLuMNS wiTh iNdices outsIdE ${{\mathCAl B}}_{11}$ scalES aS $N$. aS suCH, invoking TheoRem 3.1 in [@Zizler2002], WE coNclude ThAt $$\bEGin{aliGned}
\sUm_{J=1}^{\Min\{R,s\}n^2} \abs{\sigma_J({{{\uppErcase{{\bm{t}}}}}})-\Sigma_j({{{\UPpercasE{{\Bm{c}}}}}})} \le O(n).\End{aliGneD}$$ ThUs, we HAvE $$\bEgiN{aLIgnED}
\lIm_{n \TO \inFty} \frac{1}{n}\SuM_{j=1}^{\Min\{r,s\}N^2} \abs{\SIGMA_j({{{\upPerCase{{\Bm{t}}}}}})-\siGma_j({{{\uppercase{{\Bm{c}}}}}})}=o(1)\end{ALigNed}$$ ThIs comPletEs The prOof.
The IntuiTiOn behind is that tHe nuMber of difFerEnT elEmEnts bETween tWo mAtrIces scaLes as $n$ bUT noT $n^2$ BECAuSe of the banded strucTuRE Of ${{{\UppercasE{{\bm{c}}}}}}$ anD ${{{\UpPeRCase{{\bm{t}}}}}}$. ALtHouGh noT RIgoroUsly PRoVed, it looKs the eQUaLiTy holds WiTh the tErM $O(1)$ sTriCtly lARger Than 0, meAning thaT the cIRcular approximATion can be arbiTRaRILy LOose As $n$ Tends to infiNity.
pRoof Of ThEOrEm 3
------------------
LET $\phi_j:[-\Pi,\pi]^2 \mApSTo {\MBox{\bb R}}_+$ be the $j$-th singuLaR value FunctIon of $F({\hbox{\bolDmath$\omega$}})$ AND $\Sigma_k^{(j)}({{{\uPperCAsE{{\Bm{t}}}}}})$ be $k$-th singulAr | }(F(\omega_1,\omega_2)): ( \omega_1,\ omega _2) \i n{{\m athc al M}}\right\} , \end {aligned}$$ where ${{\ mathc al M}}$ is theuniform sa m p lin ggr ids d e fi ned i n.
Second , we showtha tthe accumula t ed differenc e o f the singul arvalues b etw e en ${ {{\ upper case{{ \ bm{c}} }}}}$ and $ { {{\upp e rcase{{ \ b m{ t}}} }}}$ is upper-bou n de d .
By inspecti ng ${{ {\ u pp e r cas e{{ \bm{t}}}}} }$ and$ {{{\upp e rc a s e {{\ b m{c}}}}}}$, w e find from Lem ma \[l em ma: p roof-t hm1\] t h at${\hbox{\bo ldma th$\Delta $}}_{k , l}=0$ i f and on ly if$(k ,l) \in {{ \m ath ca l B} } _{ 11} $ . T he numbe rof rows and c o l umns wi th i ndice s outside ${{ \ma thca l B} }_{11 }$ sc ales a s $n$ . As s uch,in voking Theorem3.1in [@Zizl er2 00 2], w e con c lude t hat $$ \begin{ aligned }
\ s u m_ {j=1}^{\min\{r,s\} n^ 2 } \ abs{\sig ma_j({ { {\ up p ercase{{ \b m{t }}}} } } )-\si gma_ j ({ {{\upper case{{ \ bm {c }}}}}}) }\le O( n) .\e nd{ align e d}$$ Thus, we have $$\b e gin{aligned}
\lim_{n \to \i n f ty } \ fra c{1}{n}\sum _{j= 1 }^{\ min\ { r, s\} n ^2} \ abs{\ si g ma _ j({{{\uppercase{{\b m{ t}}}}} })-\s igma_j({{{\up percase{{\ b m { c}}}}}}) }=O( 1 )\ e nd{aligned}$$Thiscompletest he proof .
Th e intuit ion behin d is thatthe nu mbe r o f di fferent eleme n t s be tw een two ma tricessca les as $n $but not $ n^2$ bec au se o fthe band e d struct ur e o f${{ {\upp e rcase{ {\bm{ c}}} }} }$ and ${{{\u p pe r c ase{ {\ bm {t}} }}} }$ . Alt houg h no t rigor ously pro ved , itlo ok s the e quality holds w ith the te rm $O (1)$ s t r ictly la rger than 0, meaning th a t the c irc ularappr oximation ca n be a rbi t rarily loose as $ n$ te n d s toi n fi nit y.
Proof of T heo rem 3
- ---- ------- ------
Let $\phi_ j :[- \pi,\pi]^2 \m aps to { \ m bo x{\ b bR }}_ +$ bet h e $j$-th singul ar value f un c ti on of $F({ \ hbo x{ \boldma th$\ome ga$}} ) $ and $ \sigma_k^ {(j)}({{{ \u pper c a se{ {\bm{t}}}} }})$ be$k$-th si n gular | }(F(\omega_1,\omega_2)): (\omega_1,\omega_2)_\in {{\mathcal_M}}\right\},\end{aligned}$$ where ${{\mathcal M}}$_is the_uniform_sampling grids_defined_in.
Second, we show_that the accumulated_difference of the singular_values between ${{{\uppercase{{\bm{c}}}}}}$_and_${{{\uppercase{{\bm{t}}}}}}$ is upper-bounded.
By inspecting ${{{\uppercase{{\bm{t}}}}}}$ and ${{{\uppercase{{\bm{c}}}}}}$, we find from Lemma \[lemma:proof-thm1\] that ${\hbox{\boldmath$\Delta$}}_{k,l}=0$_if_and only_if_$(k,l)_\in {{\mathcal B}}_{11}$. The number_of rows and columns with_indices outside_${{\mathcal B}}_{11}$ scales as $n$. As such, invoking_Theorem_3.1 in [@Zizler2002],_we conclude that $$\begin{aligned}
\sum_{j=1}^{\min\{r,s\}n^2} \abs{\sigma_j({{{\uppercase{{\bm{t}}}}}})-\sigma_j({{{\uppercase{{\bm{c}}}}}})}_\le O(n).\end{aligned}$$ Thus, we have $$\begin{aligned}
_ _\lim_{n_\to_\infty} \frac{1}{n}\sum_{j=1}^{\min\{r,s\}n^2}_\abs{\sigma_j({{{\uppercase{{\bm{t}}}}}})-\sigma_j({{{\uppercase{{\bm{c}}}}}})}=O(1)\end{aligned}$$ This completes the proof.
The intuition_behind is that the number of_different elements between two matrices scales as_$n$ but not $n^2$ because of_the banded structure of ${{{\uppercase{{\bm{c}}}}}}$_and ${{{\uppercase{{\bm{t}}}}}}$._Although not rigorously proved, it_looks the equality_holds with_the term $O(1)$_strictly larger than 0, meaning that_the circular approximation_can be arbitrarily loose as $n$_tends_to infinity.
Proof of_Theorem_3
------------------
Let_$\phi_j:[-\pi,\pi]^2 \mapsto_{\mbox{\bb R}}_+$ be_the_$j$-th singular_value_function of $F({\hbox{\boldmath$\omega$}})$ and $\sigma_k^{(j)}({{{\uppercase{{\bm{t}}}}}})$ be_$k$-th_singular |
the gap probability [@BRD08]; see also [@BE; @E10; @E06] for the approach of operator theory. It seems unlikely that these methods are applicable in the present case. One rough idea to tackle this problem is based on the observation that, as the parameter $\rho$ tends to $-\infty$, the cusp singularity at the origin disappears and the origin becomes a regular point inside the bulk. Thus, we expect that $F(s;\rho)$ might be related to the determinant of (generalized) sine kernel under certain scaling limits when $\rho \to -\infty$, from which the constant term can be derived. We will leave this issue to a future publication.
Finally, we note that the asymptotics of Fredholm determinant associated with the Pearcey kernel is also investigated from the viewpoint of phase transition in [@ACV; @BC1], i.e., to show how the Pearcey process becomes an Airy process by sending both $s$ and the parameter $\rho$ to positive infinity. We emphasize the asymptotic results therein are essentially different from ours.
The rest of this paper is devoted to the proof of Theorem \[main-thm\]. We mainly follow the general strategy established in [@Bor:Dei2002; @DIZ97]. In Section \[sec:DiffIdentity\], we relate the partial derivatives of $F(s;\rho)$ to a $3 \times 3$ RH problem with constant jumps, which is essential in the proof. After introducing some auxiliary functions defined on a Riemann surface with a specified sheet structure in Section \[sec:auxiliary function\], we then perform a Deift-Zhou steepest descent analysis [@DZ93] on this RH problem for large positive $s$ in Section \[sec:asymanalyX\]. This asymptotic outcome, together with the differential identities for $F(s;\rho)$, will finally lead to the proof of Theorem \[main-thm\], as presented in Section \[sec:proof\].
Differential identities for the Fredholm determinant {#sec:DiffIdentity}
====================================================
A Riemann-Hilbert characterization of the Pearcey kernel
--------------------------------------------------------
The starting point toward the proof of Theorem \[main-thm\] is an alternative representation of the Pearcey kernel $K^{\mathrm{Pe}}$ via a $3 \times 3$ RH problem, as shown in [@BK3] and stated next.
\[rhp: Pearcey\] We look for | the gap probability [ @BRD08 ]; see also [ @BE; @E10; @E06 ] for the approach of hustler hypothesis. It seem unlikely that these method are applicable in the present lawsuit. One rough idea to harness this trouble is based on the observation that, as the parameter $ \rho$ tends to $ -\infty$, the cusp singularity at the origin melt and the origin becomes a unconstipated distributor point inside the bulk. Thus, we expect that $ F(s;\rho)$ might be related to the determinant of (generalize) sine kernel under certain scaling limits when $ \rho \to -\infty$, from which the constant terminus can be derived. We will leave this issue to a future publication.
ultimately, we note that the asymptotics of Fredholm determinant associated with the Pearcey kernel is also investigated from the point of view of phase transition in [ @ACV; @BC1 ], i.e., to show how the Pearcey process becomes an Airy process by sending both $ s$ and the parameter $ \rho$ to positive infinity. We emphasize the asymptotic results therein are essentially different from ours.
The rest of this paper is devoted to the proof of Theorem \[main - thm\ ]. We chiefly keep up the general scheme prove in [ @Bor: Dei2002; @DIZ97 ]. In Section \[sec: DiffIdentity\ ], we refer the partial derivatives of $ F(s;\rho)$ to a $ 3 \times 3 $ releasing hormone problem with constant jumps, which is essential in the validation. After introducing some auxiliary functions defined on a Riemann surface with a specified sheet structure in Section \[sec: auxiliary function\ ], we then perform a Deift - Zhou steepest descent psychoanalysis [ @DZ93 ] on this RH problem for big convinced $ s$ in Section \[sec: asymanalyX\ ]. This asymptotic outcome, together with the differential identities for $ F(s;\rho)$, will finally lead to the proof of Theorem \[main - thm\ ], as presented in Section \[sec: proof\ ].
Differential identity for the Fredholm determinant { # sec: DiffIdentity }
= = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = =
A Riemann - Hilbert characterization of the Pearcey kernel
--------------------------------------------------------
The starting compass point toward the proof of Theorem \[main - thm\ ] is an alternate representation of the Pearcey kernel $ K^{\mathrm{Pe}}$ via a $ 3 \times 3 $ RH problem, as shown in [ @BK3 ] and stated next.
\[rhp: Pearcey\ ] We search for | thf gap probability [@BRD08]; set also [@BE; @E10; @E06] for the ap'roach kf operagor theory. It seems unlikely tyat tyese methods are applizable in nhe presebt cese. One rough idxz to taghle fmis pxoulem is based ok the obsereation that, as tfe parameter $\rho$ tends to $-\infty$, the ctsp sinbuparity at the jrigpn disziptars and the origin becomes a reghlar popnt inside the bukk. Thus, we expect that $F(s;\rjo)$ mlght be related to the determunane of (generalixed) sine ktruel under cgrtain scaling limits when $\rho \go -\inyty$, from whucy tjg constant txrm cag be derived. We will leave yhis issue to s fnturw publication.
Finally, xe note that the asyiptotics mf Fredholm determibabt asvocidted qitf tge Psarcey kecnel is alsk investigared from the viewpoone of phase trahsitiog yn [@ACV; @BC1], i.e., to show how the Pearcey prmceas becomes an Airy procwss by sending both $s$ and the [arameter $\rho$ to positive infinity. We emphasize tve asbmotouig fwsklts therein are essentially different from otds.
Uhe rest of this iaper is devoted tp hhr proof of Theurem \[mcjn-fhm\]. We mainly folllw the denerql stratedy extablished in [@Bor:Dei2002; @DIZ97]. In Section \[sec:BifdIdentity\], we relatz the partiau detivatifes of $F(s;\rho)$ to a $3 \timer 3$ RG problem wlth constzvt jumps, which ir exsantial in the proof. After yntroduciig soke auxiuiari functyons definfd on a Riemann surface aith c spewified sheft structure in Section \[sec:auxiliary function\], wa tven perfjrm a Deift-Zhou stqepest descent analysns [@DZ93] un this RH problek for large positive $s$ it Section \[sec:esymanalyV\]. Thus awymptotkz outcome, togeyher with the diffeeential identities fot $R(s;\rho)$, will finaoly lead to the prpof of Tjeprqk \[main-thm\], as prerengrd in Section \[seg:pruof\].
Dofferential identitias fkr the Fredholm deyevminant {#sgc:DiffIdegtity}
====================================================
A Riemanm-Hilbert characterlzatimn uf thr Pgarcey kernel
--------------------------------------------------------
The starting point toward tje iroof of Theowem \[nain-thm\] is au alternative representation of the Pearrey kernel $K^{\mathrm{Pe}}$ viq a $3 \times 3$ RH probktm, as shown ii [@BK3] agd stated next.
\[rhp: Pearcey\] We oook for | the gap probability [@BRD08]; see also [@BE; for approach of theory. It seems applicable the present case. rough idea to this problem is based on the that, as the parameter $\rho$ tends to $-\infty$, the cusp singularity at the disappears and the origin becomes a regular point inside the bulk. Thus, we that might related the determinant of (generalized) sine kernel under certain scaling limits when $\rho \to -\infty$, from which constant term can be derived. We will leave issue to a future Finally, we note that the of determinant associated the kernel also investigated from viewpoint of phase transition in [@ACV; @BC1], i.e., to show how the Pearcey process becomes an Airy by sending and the $\rho$ positive We emphasize the therein are essentially different from ours. this paper is devoted to the proof of \[main-thm\]. We follow the general strategy established in @DIZ97]. In Section \[sec:DiffIdentity\], we relate the partial of $F(s;\rho)$ to a $3 \times 3$ RH problem with constant jumps, which is essential proof. After introducing some functions defined on Riemann with specified structure in \[sec:auxiliary function\], we then perform a Deift-Zhou steepest descent analysis [@DZ93] this RH problem for large positive $s$ in Section \[sec:asymanalyX\]. outcome, with the differential for $F(s;\rho)$, will finally to proof of Theorem \[main-thm\], in \[sec:proof\]. the determinant ==================================================== A Riemann-Hilbert characterization the Pearcey kernel -------------------------------------------------------- The point toward the proof alternative representation of the Pearcey kernel $K^{\mathrm{Pe}}$ via $3 \times 3$ RH problem, as shown [@BK3] and stated next. \[rhp: Pearcey\] We look for | the gap probability [@BRD08]; see alSo [@BE; @E10; @E06] for tHe appRoaCh oF oPeraTor tHeory. It seems unLIkelY that these methods are apPlicaBlE In thE PrEsent Case. One ROuGH IdeA tO tAckLe THiS probLem Is based On the obserVatIoN that, as the paRAmEter $\rho$ tenDs tO $-\infty$, the cusP siNgularItY at THe oriGin DisapPears aND the orIgin becomEs A RegulaR Point inSIDe The bUlk. Thus, we expect thAT $F(S;\Rho)$ might be relaTed to tHe DEtERMinAnt Of (generaliZeD) sine KErnel unDEr CERTaiN Scaling limits When $\rho \to -\inFTy$, fRom whiCh The COnstanT term CaN Be dErived. We wilL leaVe this issUe to a fUTure pubLIcation.
finallY, we NotE thaT ThE aSymPtOTicS Of freDHolM determiNaNt AssocIateD WITH the peaRcey KerneL is also investIgaTed fROm tHe vieWpoinT of pHaSe traNsitioN in [@ACv; @Bc1], i.e., to show how the pearCey procesS beCoMes An airy pROcess bY seNdiNg both $s$ And the pARamEtER $\RHo$ To positive infinity. we EMPhAsize the AsymptOTiC rESults theReIn aRe esSENtialLy diFFeRent from Ours.
ThE ReSt Of this pApEr is deVoTed To tHe proOF of THeorem \[Main-thm\]. WE mainLY follow the geneRAl strategy estABlISHeD In [@BoR:DeI2002; @DIZ97]. In SectiOn \[seC:diffidenTItY\], we RElate The paRtIAl DErivatives of $F(s;\rho)$ to A $3 \tImes 3$ RH ProblEm with constanT jumps, whicH IS EssentiaL in tHE pROof. After introdUcing Some auxiliARy functiOns deFined on a riemann suRFAce with a SpeCifIed SheET StRucture in SectION \[sec:AuXiliary FunCtion\], we TheN peRfoRm a deIft-Zhou stEepest deScEnT aNaLysIs [@DZ93] oN This RH prObLem FoR laRge poSItive $s$ In SecTion \[SeC:aSYmaNalyX\]. ThIS aSYMptoTiC oUtcoMe, tOgEther With THe dIfferenTial identItiES for $f(s;\RhO)$, will fiNally lead to thE pRoof of TheoReM \[maIn-thm\], aS PResented In Section \[sec:proof\].
DifferENtial idEntIties For tHe FredholM deTerminAnt {#SEc:DiffidentiTy}
====================================================
A RiEmAnn-hILbert CHArActErIzation of tHE peaRcey kErNel
--------------------------------------------------------
THe startIng point toward the pROof Of Theorem \[main-Thm\] Is an ALTeRnaTIvE RepReSEntATIon of the Pearcey Kernel $K^{\matHrM{pe}}$ Via a $3 \times 3$ Rh ProBlEm, as shoWn in [@BK3] aNd staTEd next.
\[rHp: Pearcey\] we look for | the gap probability [@BRD 08]; see a lso [ @BE ; @ E1 0; @ E06] for the appro a ch o f operator theory. Itseems u n like l ythatthese m e th o d s a re a ppl ic a bl e inthe presen t case. On e r ou gh idea to t a ck le this pr obl em is basedonthe ob se rva t ion t hat , asthe pa r ameter $\rho$ t en d s to $ - \infty$ , th e cu sp singularity at th e origin disapp ears a nd th e ori gin becomes a r egula r pointi ns i d e th e bulk. Thus,we expect t h at$F(s;\ rh o)$ mightbe re la t edto the dete rmin ant of (g eneral i zed) si n e kerne l unde r c ert ains ca li ngli m its wh en$ \rh o \to - \i nf ty$,from w h i ch t hecons tantterm can be d eri ved. Wewillleave thi sissue to afutur epublication.
F inal ly, we no teth atth e asy m ptotic s o f F redholm determ i nan ta s s oc iated with the Pea rc e y k ernel is alsoi nv es t igated f ro m t he v i e wpoin t of ph ase tran sition in [ @ACV; @ BC 1], i. e. , t o s how h o w th e Pear cey proc ess b e comes an Airyp rocess by sen d in g bo t h $s $ a nd the para mete r $\r ho$t opos i tiveinfin it y .W e emphasize the asy mp toticresul ts therein ar e essentia l l y differe nt f r om ours.
The res t ofthis paper is devot ed to the pro of of The o r em \[mai n-t hm\ ].Wem a in ly follow the g ener al strate gyestabli she d i n [ @Bo r: Dei2002;@DIZ97]. I nSe ct ion \[se c :DiffIde nt ity \] , w e rel a te the part ialde ri v ati ves of$ F( s ; \rho )$ t o a$3\t imes3$ R H pr oblem w ith const ant jump s, w hich is essential in t he proof.Af ter intro d u cing som e auxiliary functions d e fined o n a Riem annsurface w ith a spe cif i ed she et str uctur einS e ction \ [s ec: au xiliary fu n c tio n\],we the n perfo rm a Deift-Zhou st e epe st descent an aly sis[ @ DZ 93] on thi sR H p r o blem for largepositive $ s$ in Section \ [ sec :a symanal yX\]. T his a s ymptoti c outcome , togethe rwith t hedifferenti al ident ities for $F(s; \ rh o)$,wil l fina ll y l ead t o thep roo f ofTheore m\[main -thm\ ], as pres ented in Section \[sec: proof\ ].
D iff erentialide n tit ies for t he F redholm de ter min ant { #se c :Diff Iden t it y}= ===== ==== = ========= = == === = = == =========== = = = === =====
A Rieman n-Hi lbert characteriz a tion of the Pe arce y ker nel ---- -- -------------- --- -- - - -------- -- ----------- --------
T he st arting point toward t he proofof T heo rem \[mai n-t hm \ ] is an a lt e rnativ e re pr esenta tion o f the P earcey kernel $K ^{\ma t h rm{Pe } }$via a $ 3 \time s 3$RH problem , as shownin [@B K3]and s tated n ex t.
\[ rhp :Pearcey\]W e look fo r | the_gap probability_[@BRD08]; see also [@BE;_@E10; @E06]_for_the approach_of_operator theory. It_seems unlikely that_these methods are applicable_in the present_case._One rough idea to tackle this problem is based on the observation that, as_the_parameter $\rho$_tends_to_$-\infty$, the cusp singularity at_the origin disappears and the_origin becomes_a regular point inside the bulk. Thus, we_expect_that $F(s;\rho)$ might_be related to the determinant of (generalized) sine kernel_under certain scaling limits when $\rho_ \to -\infty$,_from_which_the constant term can_be derived. We will leave this_issue to a future publication.
Finally, we_note that the asymptotics of Fredholm determinant_associated with the Pearcey kernel is_also investigated from the viewpoint_of phase_transition in [@ACV; @BC1], i.e.,_to show how_the Pearcey_process becomes an_Airy process by sending both $s$_and the parameter_$\rho$ to positive infinity. We emphasize_the_asymptotic results therein_are_essentially_different from_ours.
The rest of_this_paper is_devoted_to the proof of Theorem \[main-thm\]._We_mainly follow the general strategy established in_[@Bor:Dei2002; @DIZ97]. In Section_\[sec:DiffIdentity\],_we relate the partial_derivatives of $F(s;\rho)$ to a_$3 \times 3$ RH problem with_constant jumps,_which is_essential in the proof. After introducing some auxiliary functions defined on_a Riemann surface with a specified_sheet structure in Section_\[sec:auxiliary function\],_we_then perform a_Deift-Zhou_steepest descent_analysis [@DZ93] on this RH problem for_large positive_$s$ in Section \[sec:asymanalyX\]. This asymptotic_outcome, together with the_differential_identities for $F(s;\rho)$, will finally lead_to the proof of Theorem \[main-thm\],_as presented in Section \[sec:proof\].
Differential_identities_for_the Fredholm determinant {#sec:DiffIdentity}
====================================================
A Riemann-Hilbert_characterization of the Pearcey kernel
--------------------------------------------------------
The starting_point toward the_proof of Theorem \[main-thm\] is an alternative_representation_of the Pearcey kernel $K^{\mathrm{Pe}}$ via_a_$3 \times 3$ RH problem, as_shown_in_[@BK3] and stated next.
\[rhp: Pearcey\]_We look for |
2)^{n-1}}{({{\sqrt{-1}\partial\bar{\partial}}}r^2)^n}=\xi_0(v)\cdot \frac{1}{r}.$$
Since $T$ acts on $X$, $T$ also acts on the set of functions on $X$ by $\tau\circ f(x)=f(\tau^{-1}x)$ for any $\tau\in T$ and $x\in X$. For convenience, we introduce the following
Denote by $PSH(X,\xi_0)$ the set of bounded real functions ${{\varphi}}$ on $X$ that satisfies:
1. $\tau\circ {{\varphi}}={{\varphi}}$ for any $\tau\in T$;
2. $r^2_{{\varphi}}:=r^2 e^{{{\varphi}}}$ is a proper strictly plurisubharmonic function on $X$.
\[defn-RFKC\] We say that $r^2_{{\varphi}}:=r^2 e^{{{\varphi}}}$ where ${{\varphi}}\in PSH(X, \xi_0)$ is the radius function of a Ricci-flat Kähler cone metric on $(X, \xi_0)$ if ${{\varphi}}$ is smooth on $X^{\rm reg}$ and there exists a positive constant $C>0$ such that $$\label{eq-RFKC}
({{\sqrt{-1}\partial\bar{\partial}}}r^2_{{\varphi}})^n=C\cdot dV.$$
Compared with the weak Kähler-Einstein case, it is expected that the regularity condition in the above definition is automatically satisfied. With this regularity assumption, on the regular part $X^{\rm reg}$, both sides of are smooth volume forms and we have $r_{{{\varphi}}}\partial_{r_{{\varphi}}}=2 {\rm Re}(\xi_0)$ or, equivalently, $\xi_0=r_{{\varphi}}\partial_{r_{{\varphi}}}-i J(r_{{\varphi}}\partial_{r_{{\varphi}}})$. Moreover, taking $\mathcal{L}_{r_{{\varphi}}\partial_{r_{{\varphi}}}}$ on both sides gives us the identity $\mathcal{L}_{r_{{\varphi}}\partial_{r_{{\varphi}}}}dV=2n\; dV$. Equivalently we have: $${{\mathcal{L}}}_{\xi_0}s= m n\cdot s,$$ where $s\in |-mK_X|$ is the chosen $T | 2)^{n-1}}{({{\sqrt{-1}\partial\bar{\partial}}}r^2)^n}=\xi_0(v)\cdot \frac{1}{r}.$$
Since $ T$ acts on $ X$, $ T$ also acts on the set of function on $ X$ by $ \tau\circ f(x)=f(\tau^{-1}x)$ for any $ \tau\in T$ and $ x\in X$. For public toilet, we introduce the following
Denote by $ PSH(X,\xi_0)$ the set of bound real functions $ { { \varphi}}$ on $ X$ that satisfies:
1. $ \tau\circ { { \varphi}}={{\varphi}}$ for any $ \tau\in T$;
2. $ r^2_{{\varphi}}:=r^2 e^{{{\varphi}}}$ is a proper rigorously plurisubharmonic function on $ X$.
\[defn - RFKC\ ] We say that $ r^2_{{\varphi}}:=r^2 e^{{{\varphi}}}$ where $ { { \varphi}}\in PSH(X, \xi_0)$ is the radius affair of a Ricci - flat Kähler cone metric on $ (adam, \xi_0)$ if $ { { \varphi}}$ is smooth on $ X^{\rm reg}$ and there exists a incontrovertible constant $ C>0 $ such that $ $ \label{eq - RFKC }
({ { \sqrt{-1}\partial\bar{\partial}}}r^2_{{\varphi}})^n = C\cdot dV.$$
Compared with the weak Kähler - Einstein case, it is expected that the regularity condition in the above definition is mechanically satisfied. With this regularity assumption, on the regular function $ X^{\rm reg}$, both sides of are smooth bulk forms and we experience $ r_{{{\varphi}}}\partial_{r_{{\varphi}}}=2 { \rm Re}(\xi_0)$ or, equivalently, $ \xi_0 = r_{{\varphi}}\partial_{r_{{\varphi}}}-i J(r_{{\varphi}}\partial_{r_{{\varphi}}})$. Moreover, taking $ \mathcal{L}_{r_{{\varphi}}\partial_{r_{{\varphi}}}}$ on both sides gives us the identity $ \mathcal{L}_{r_{{\varphi}}\partial_{r_{{\varphi}}}}dV=2n\; dV$. Equivalently we have: $ $ { { \mathcal{L}}}_{\xi_0}s= m n\cdot s,$$ where $ s\in |-mK_X|$ is the choose $ T | 2)^{n-1}}{({{\sqgt{-1}\partial\bar{\partial}}}r^2)^n}=\xi_0(v)\gdot \frac{1}{r}.$$
Since $J$ qcts oi $X$, $T$ amso acts on the set of functions on $E$ by $\tau\curc f(x)=f(\tau^{-1}x)$ for any $\taj\in T$ and $x\in X$. Fir cibvenience, xs introduce tgc folniwing
Denote by $PSH(X,\xi_0)$ tha set of boundad rzal functions ${{\varphi}}$ on $X$ that satissies:
1. $\tsu\firc {{\varphi}}={{\varkhi}}$ fpw ang $\tau\in T$;
2. $r^2_{{\varphi}}:=r^2 e^{{{\varphi}}}$ is a pdoper surictly plurisubhatmonic function on $X$.
\[defn-RFNC\] Wf say that $r^2_{{\varphi}}:=g^2 e^{{{\varphi}}}$ wyere ${{\carphi}}\in PSH(B, \xi_0)$ is tht xadius funcjion of a Ricci-flat Kähler cone mdtric on $(X, \xi_0)$ id ${{\carovi}}$ is smooti on $X^{\gm reg}$ and thcge exisds a poxitive constanb $C>0$ snch rhat $$\label{eq-RFKC}
({{\sqrt{-1}\pertial\bar{\partial}}}r^2_{{\varkhi}})^n=C\cdot gV.$$
Eompared with the weaj Jähler-Ginstain zqse, it ix sxpectfd vhat the refularity cobdition in the abovt dqdinition is ahtomatycwlly satisfied. With this regularity asslmptjon, on the regular part $X^{\rm reg}$, both sides ov are smojth volume forms and we have $r_{{{\varphi}}}\partial_{r_{{\varphh}}}=2 {\rm Ce}(\bi_0)$ ir, dwulvalently, $\xi_0=r_{{\varphi}}\partial_{r_{{\varphi}}}-i J(r_{{\varphi}}\pawfisl_{g_{{\varphi}}})$. Moreover, baking $\mathcal{L}_{r_{{\vatpji}}\lwrtial_{r_{{\varphi}}}}$ on bocg aides gives us the identijy $\matycal{L}_{r_{{\varkhi}}\pattial_{r_{{\varphi}}}}dV=2n\; dV$. Equivalebtly we have: $${{\nathcal{L}}}_{\xi_0}s= m n\cdoc s,$$ where $s\iu |-mK_X|$ os thr chosen $T | 2)^{n-1}}{({{\sqrt{-1}\partial\bar{\partial}}}r^2)^n}=\xi_0(v)\cdot \frac{1}{r}.$$ Since $T$ acts on $X$, acts the set functions on $X$ $\tau\in and $x\in X$. convenience, we introduce following Denote by $PSH(X,\xi_0)$ the set bounded real functions ${{\varphi}}$ on $X$ that satisfies: 1. $\tau\circ {{\varphi}}={{\varphi}}$ for any T$; 2. $r^2_{{\varphi}}:=r^2 e^{{{\varphi}}}$ is a proper strictly plurisubharmonic function on $X$. \[defn-RFKC\] say $r^2_{{\varphi}}:=r^2 where PSH(X, \xi_0)$ is the radius function of a Ricci-flat Kähler cone metric on $(X, \xi_0)$ if is smooth on $X^{\rm reg}$ and there exists positive constant $C>0$ such $$\label{eq-RFKC} ({{\sqrt{-1}\partial\bar{\partial}}}r^2_{{\varphi}})^n=C\cdot dV.$$ Compared with weak case, it expected the condition in the definition is automatically satisfied. With this regularity assumption, on the regular part $X^{\rm reg}$, both sides of smooth volume we have {\rm or, $\xi_0=r_{{\varphi}}\partial_{r_{{\varphi}}}-i J(r_{{\varphi}}\partial_{r_{{\varphi}}})$. Moreover, on both sides gives us the Equivalently we have: $${{\mathcal{L}}}_{\xi_0}s= m n\cdot s,$$ where |-mK_X|$ is chosen $T | 2)^{n-1}}{({{\sqrt{-1}\partial\bar{\partial}}}r^2)^n}=\xI_0(v)\cdot \frac{1}{R}.$$
SincE $T$ aCts On $x$, $T$ alSo acTs on the set of fuNCtioNs on $X$ by $\tau\circ f(x)=f(\tau^{-1}x)$ fOr any $\TaU\In T$ aND $x\In X$. FoR convenIEnCE, We iNtRoDucE tHE fOllowIng
denote bY $PSH(X,\xi_0)$ the Set Of Bounded real fUNcTions ${{\varphI}}$ on $x$ that satisfiEs:
1. $\tAu\circ {{\VaRphI}}={{\VarphI}}$ foR any $\tAu\in T$;
2. $r^2_{{\VArphi}}:=r^2 E^{{{\varphi}}}$ is A pROper stRIctly plURIsUbhaRmonic function on $X$.
\[DEfN-rFKC\] We say that $r^2_{{\Varphi}}:=R^2 e^{{{\VArPHI}}}$ whEre ${{\Varphi}}\in PSh(X, \Xi_0)$ is tHE radius FUnCTIOn oF A Ricci-flat KähLer cone metrIC on $(x, \xi_0)$ if ${{\vArPhi}}$ IS smootH on $X^{\rM rEG}$ anD there existS a poSitive conStant $C>0$ SUch that $$\LAbel{eq-RfKC}
({{\sqrT{-1}\paRtiAl\baR{\PaRtIal}}}R^2_{{\vARphI}})^N=C\CdoT DV.$$
COmpared wItH tHe weaK KähLER-eInstEin Case, It is eXpected that thE reGulaRIty CondiTion iN the AbOve deFinitiOn is aUtOmatically satisFied. with this rEguLaRitY aSsumpTIon, on tHe rEguLar part $x^{\rm reg}$, bOTh sIdES OF aRe smooth volume formS aND We Have $r_{{{\varPhi}}}\parTIaL_{r_{{\VArphi}}}=2 {\rm RE}(\xI_0)$ or, EquiVALentlY, $\xi_0=r_{{\VArPhi}}\partiAl_{r_{{\varPHi}}}-I J(R_{{\varphi}}\PaRtial_{r_{{\VaRphI}}})$. MoReoveR, TakiNg $\mathCal{L}_{r_{{\varPhi}}\paRTial_{r_{{\varphi}}}}$ on bOTh sides gives uS ThE IDeNTity $\MatHcal{L}_{r_{{\varphI}}\parTIal_{r_{{\VarpHI}}}}dv=2n\; dv$. equivAlentLy WE hAVe: $${{\mathcal{L}}}_{\xi_0}s= m n\cdot S,$$ wHere $s\iN |-mK_X|$ iS the chosen $T | 2)^{n-1}}{({{\sqrt{-1}\par tial\bar{\ parti al} }}r ^2 )^n} =\xi _0(v)\cdot \fr a c{1} {r}.$$
Since $T$ acts on $ X$ , $T$ al so ac ts on t h es e t o ffu nct io n son $X $ b y $\tau \circ f(x) =f( \t au^{-1}x)$ f o rany $\tau\ inT$ and $x\in X$ . Forco nve n ience , w e int roduce the fo llowing
De n ote by $PSH(X, \ x i_ 0)$the set of bounde d r e al functions $ {{\var ph i }} $ on$X$ that sati sf ies:1. $\t a u\ c i r c { { \varphi}}={{\ varphi}}$ f o r a ny $\t au \in T$;
2 . $r ^2 _ {{\ varphi}}:=r ^2 e ^{{{\varp hi}}}$ is a pr o per str ictlyplu ris ubha r mo ni c f un c tio n o n $ X $.
\[defn- RF KC \] We say t h a t $r ^2_ {{\v arphi }}:=r^2 e^{{{ \va rphi } }}$ wher e ${{ \var ph i}}\i n PSH( X, \x i_ 0)$ is the radi us f unction o f a R icc i- flatK ählercon e m etric o n $(X,\ xi_ 0) $ i f${{\varphi}}$ is s mo o t hon $X^{\ rm reg } $an d there e xi sts a p o s itive con s ta nt $C>0$ sucht ha t$$\labe l{ eq-RFK C}
({ {\s qrt{- 1 }\pa rtial\ bar{\par tial} } }r^2_{{\varphi } })^n=C\cdot d V .$ $
C o mpar edwith the we ak K ä hler -Ein s te inc ase,it is e x pe c ted that the regula ri ty con ditio n in the abov e definiti o n is autom atic a ll y satisfied. Wi th th is regular i ty assum ption , on the regularp a rt $X^{\ rmreg }$, bo t h s ides of are s m o othvo lume fo rms and we ha ve$r_ {{{ \v arphi}}}\ partial_ {r _{ {\ va rph i}}}= 2 {\rm Re }( \xi _0 )$or, e q uivale ntly, $\x i_ 0= r _{{ \varphi } }\ p a rtia l_ {r _{{\ var ph i}}}- i J( r _{{ \varphi }}\partia l_{ r _{{\ va rp hi}}})$ . Moreover, t ak ing $\math ca l{L }_{r_{ { \ varphi}} \partial_{r_{{\varphi}} } }$ on b oth side s gi ves us th e i dentit y $ \ mathca l{L}_{ r_{{\ va rph i } }\par t i al _{r _{ {\varphi}} } } dV= 2n\;dV $. E quivale ntly we have: $${{ \ mat hcal{L}}}_{\x i_0 }s=m n\ cdo t s , $$wh e re$ s \in |-mK_X|$ is the chose n$ T | 2)^{n-1}}{({{\sqrt{-1}\partial\bar{\partial}}}r^2)^n}=\xi_0(v)\cdot \frac{1}{r}.$$
Since_$T$ acts_on $X$, $T$ also_acts on_the_set of_functions_on $X$ by_$\tau\circ f(x)=f(\tau^{-1}x)$ for_any $\tau\in T$ and_$x\in X$. For_convenience,_we introduce the following
Denote by $PSH(X,\xi_0)$ the set of bounded real functions ${{\varphi}}$ on_$X$_that satisfies:
1.__$\tau\circ_{{\varphi}}={{\varphi}}$ for any $\tau\in T$;
2._ $r^2_{{\varphi}}:=r^2 e^{{{\varphi}}}$ is a_proper strictly_plurisubharmonic function on $X$.
\[defn-RFKC\] We say that $r^2_{{\varphi}}:=r^2_e^{{{\varphi}}}$_where ${{\varphi}}\in PSH(X,_\xi_0)$ is the radius function of a Ricci-flat Kähler_cone metric on $(X, \xi_0)$ if_${{\varphi}}$ is smooth_on_$X^{\rm_reg}$ and there exists_a positive constant $C>0$ such that_$$\label{eq-RFKC}
({{\sqrt{-1}\partial\bar{\partial}}}r^2_{{\varphi}})^n=C\cdot dV.$$
Compared with the weak Kähler-Einstein_case, it is expected that the regularity_condition in the above definition is_automatically satisfied. With this regularity_assumption, on_the regular part $X^{\rm reg}$,_both sides of_are smooth_volume forms and_we have $r_{{{\varphi}}}\partial_{r_{{\varphi}}}=2 {\rm Re}(\xi_0)$ or,_equivalently, $\xi_0=r_{{\varphi}}\partial_{r_{{\varphi}}}-i J(r_{{\varphi}}\partial_{r_{{\varphi}}})$._Moreover, taking $\mathcal{L}_{r_{{\varphi}}\partial_{r_{{\varphi}}}}$ on both sides_gives_us the identity_$\mathcal{L}_{r_{{\varphi}}\partial_{r_{{\varphi}}}}dV=2n\;_dV$._Equivalently we_have: $${{\mathcal{L}}}_{\xi_0}s= m_n\cdot_s,$$ where_$s\in_|-mK_X|$ is the chosen $T |
, we introduce the class of Burgers vectors $S=\{b_1,...,b_s\}$. In what follows we assume that $S$ contains at least two (independent) vectors, so that $$\label{R2}
{\rm Span}_{{\mathbb{R}}} S={\mathbb{R}}^2$$ and this will imply that the function $\varphi$ in the energy (\[Gammalimit\]) is finite in whole of ${\mathbb{R}}^2$. The case of only one Burgers vector is easier and it implies that $\varphi$ is finite only on a one dimensional subspace of ${\mathbb{R}}^2$. We denote by $\mathbb{S}$ the span of $S$ with integer coefficients ($\mathbb{S}={\rm Span}_{{\mathbb{Z}}}S$), i.e., the set of Burgers vectors for “multiple dislocations".
As in Section \[Scalings\], $N_{\varepsilon}$ represents the number of dislocations present in the crystal, corresponding to the internal scale ${\varepsilon}$. We introduce also the sequence $\rho_{\varepsilon}$ representing the radius of the hard core surrounding the dislocations, and we require
- $\lim_{{\varepsilon}\to 0} \frac{\rho_{\varepsilon}}{{\varepsilon}_n^s} = \infty$ for every fixed $0<s<1$;
- $\lim_{{\varepsilon}\to 0} |N_{\varepsilon}| \rho_{\varepsilon}^2 = 0$.
Condition i) says that the hard core region contains almost all the self-energy, while condition ii) says that the area of the hard core region tends to zero, and hence that its complement contains almost all the interaction energy. Note that conditions i) and ii) are compatible whenever
$$\label{ipn}
N_{\varepsilon}{\varepsilon}^s \to 0 \qquad \text{ for every fixed } s>0.$$
We assume that the distance between any pair of dislocation points is at least $2\rho_{\varepsilon}$ and we define the class $X_{{\varepsilon}}$ of admissible dislocations by $$\begin{gathered}
\label{xn}
X_{{\varepsilon}}:= \Big\{\mu\in {{\mathcal M}}({\Omega};{\mathbb{R}}^2): \, \mu=\sum_{i=1}^M \xi_i \, \delta_{x_i}, \, M\in{\mathbb{N}},\, B_{{\rho_{\varepsilon}}}(x_i) \subset {\ | , we introduce the class of Burgers vectors $ S=\{b_1,... ,b_s\}$. In what follows we assume that $ S$ control at least two (autonomous) vectors, so that $ $ \label{R2 }
{ \rm Span}_{{\mathbb{R } } } S={\mathbb{R}}^2$$ and this will imply that the affair $ \varphi$ in the department of energy (\[Gammalimit\ ]) is finite in whole of $ { \mathbb{R}}^2$. The case of only one Burgers vector is easier and it imply that $ \varphi$ is finite only on a one dimensional subspace of $ { \mathbb{R}}^2$. We denote by $ \mathbb{S}$ the couple of $ S$ with integer coefficients ($ \mathbb{S}={\rm Span}_{{\mathbb{Z}}}S$), i.e., the set of Burgers vector for “ multiple dislocations ".
As in Section \[Scalings\ ], $ N_{\varepsilon}$ represents the phone number of dislocation present in the crystal, corresponding to the internal scale $ { \varepsilon}$. We introduce besides the sequence $ \rho_{\varepsilon}$ representing the radius of the arduous core surrounding the dislocation, and we require
- $ \lim_{{\varepsilon}\to 0 } \frac{\rho_{\varepsilon}}{{\varepsilon}_n^s } = \infty$ for every repair $ 0 < s<1 $;
- $ \lim_{{\varepsilon}\to 0 } |N_{\varepsilon}| \rho_{\varepsilon}^2 = 0$.
Condition i) says that the hard core region contains about all the self - energy, while condition ii) says that the area of the hard core area tend to zero, and hence that its complement contains almost all the interaction energy. Note that conditions i) and ii) are compatible whenever
$ $ \label{ipn }
N_{\varepsilon}{\varepsilon}^s \to 0 \qquad \text { for every fixed } s>0.$$
We assume that the distance between any pair of dislocation point is at least $ 2\rho_{\varepsilon}$ and we define the class $ X_{{\varepsilon}}$ of admissible dislocation by $ $ \begin{gathered }
\label{xn }
X_{{\varepsilon}}:= \Big\{\mu\in { { \mathcal M}}({\Omega};{\mathbb{R}}^2 ): \, \mu=\sum_{i=1}^M \xi_i \, \delta_{x_i }, \, M\in{\mathbb{N}},\, B_{{\rho_{\varepsilon}}}(x_i) \subset { \ | , we introduce the class of Nurgers vectors $S=\{b_1,...,b_s\}$. In what rollows de assume that $S$ contains at lwast uco (independent) vectofs, so than $$\label{R2}
{\rn Spen}_{{\mathbb{R}}} S={\mathbu{D}}^2$$ and tmns wimp im'lb that the funcjion $\varphi$ hn the energy (\[Caomclimit\]) is finite in whole of ${\mathbb{R}}^2$. The caxe of only one Borgerx vecfor is easier and it implies that $\barphi$ ps finite only on a one dimensional subspacf of ${\mathbb{R}}^2$. We denote by $\mathbb{S}$ the wpan of $S$ wigh integer coefficienta ($\mathbb{S}={\rm Span}_{{\mathbb{Z}}}S$), i.e., thd set of Burgerw cechmrs for “mulviple qislocations".
As in Sewtion \[Svalings\], $N_{\varepxilmn}$ eepresents the number of dislocations ptesent in dhz crystal, correspondibg to tve ittervql rcame ${\vzrepsipon}$. We introdhce also thw sequence $\rho_{\varepxijin}$ representihg the rwdius of the hard core surrounding the gismocations, and we requirw
- $\lim_{{\varepsilon}\to 0} \vrac{\rho_{\vawepsilon}}{{\varepsilon}_n^s} = \infty$ for every fixed $0<s<1$;
- $\lhm_{{\varxprilin}\bo 0} |V_{\cagepsilon}| \rho_{\varepsilon}^2 = 0$.
Condition i) says that fht hsrd core regiok contains almost slp yre self-energy, while cohdition ii) says thwt the wrea if the hawd cpre region tends to zero, ane hence that uts complement concains almost all the onteraction energy. Note thaf conditiond i) and ij) are compatible dhekevar
$$\label{ikv}
N_{\varepsilon}{\varepfilon}^s \to 0 \qauad \tebt{ fpr evewy fixed } d>0.$$
We assume that the distwnce yetwean any paig of dislocation points is at lxest $2\rho_{\varepsolmn}$ dnd we dzfine bhe class $X_{{\vare[silon}}$ of admixsible bislocxtions by $$\gegin{gavhered}
\label{xg}
X_{{\varepsilon}}:= \Tlg\{\mu\in {{\mathcel M}}({\Omega};{\iathvb{R}}^2): \, \mu=\sum_{i=1}^O \xi_i \, \delta_{x_i}, \, M\in{\mathby{U}},\, B_{{\rho_{\varwpsilon}}}(x_i) \subset {\ | , we introduce the class of Burgers In follows we that $S$ contains so $$\label{R2} {\rm Span}_{{\mathbb{R}}} and this will that the function $\varphi$ in the (\[Gammalimit\]) is finite in whole of ${\mathbb{R}}^2$. The case of only one Burgers is easier and it implies that $\varphi$ is finite only on a one subspace ${\mathbb{R}}^2$. denote $\mathbb{S}$ the span of $S$ with integer coefficients ($\mathbb{S}={\rm Span}_{{\mathbb{Z}}}S$), i.e., the set of Burgers vectors “multiple dislocations". As in Section \[Scalings\], $N_{\varepsilon}$ represents number of dislocations present the crystal, corresponding to the scale We introduce the $\rho_{\varepsilon}$ the radius of hard core surrounding the dislocations, and we require - $\lim_{{\varepsilon}\to 0} \frac{\rho_{\varepsilon}}{{\varepsilon}_n^s} = \infty$ for every fixed - $\lim_{{\varepsilon}\to \rho_{\varepsilon}^2 = Condition says the hard core almost all the self-energy, while condition the area of the hard core region tends zero, and that its complement contains almost all interaction energy. Note that conditions i) and ii) compatible whenever $$\label{ipn} N_{\varepsilon}{\varepsilon}^s \to 0 \qquad \text{ for every fixed } s>0.$$ We assume distance between any pair dislocation points is least and define class $X_{{\varepsilon}}$ admissible dislocations by $$\begin{gathered} \label{xn} X_{{\varepsilon}}:= \Big\{\mu\in {{\mathcal M}}({\Omega};{\mathbb{R}}^2): \, \mu=\sum_{i=1}^M \, \delta_{x_i}, \, M\in{\mathbb{N}},\, B_{{\rho_{\varepsilon}}}(x_i) \subset {\ | , we introduce the class of BurgErs vectors $s=\{b_1,...,b_s\}$. IN whAt fOlLows We asSume that $S$ contaINs at Least two (independent) vecTors, sO tHAt $$\laBEl{r2}
{\rm SpAn}_{{\mathbB{r}}} S={\MAThbB{R}}^2$$ AnD thIs WIlL implY thAt the fuNction $\varpHi$ iN tHe energy (\[GammALiMit\]) is finitE in Whole of ${\mathbB{R}}^2$. THe case Of OnlY One BuRgeRs vecTor is eASier anD it implieS tHAt $\varpHI$ is finiTE OnLy on A one dimensional suBSpACe of ${\mathbb{R}}^2$. We dEnote bY $\mAThBB{s}$ thE spAn of $S$ with iNtEger cOEfficieNTs ($\MATHbb{s}={\Rm Span}_{{\mathbb{Z}}}s$), i.e., the set of bUrgErs vecToRs fOR “multiPle diSlOCatIons".
As in SecTion \[scalings\], $N_{\VarepsILon}$ reprESents thE numbeR of DisLocaTIoNs PreSeNT in THe CrySTal, CorrespoNdInG to thE intERNAL scaLe ${\vArepSilon}$. we introduce alSo tHe seQUenCe $\rho_{\VarepSiloN}$ rEpresEnting The raDiUs of the hard core SurrOunding thE diSlOcaTiOns, anD We requIre
- $\Lim_{{\VarepsiLon}\to 0} \frAC{\rhO_{\vAREPsIlon}}{{\varepsilon}_n^s} = \inFtY$ FOr Every fixEd $0<s<1$;
- $\lim_{{\VArEpSIlon}\to 0} |N_{\vArEpsIlon}| \RHO_{\varePsilON}^2 = 0$.
COndition I) says tHAt ThE hard coRe Region CoNtaIns AlmosT All tHe self-Energy, whIle coNDition ii) says thAT the area of the HArD COrE RegiOn tEnds to zero, aNd heNCe thAt itS CoMplEMent cOntaiNs ALmOSt all the interaction EnErgy. NoTe thaT conditions i) aNd ii) are comPATIble whenEver
$$\LAbEL{ipn}
N_{\varepsiloN}{\varePsilon}^s \to 0 \qQUad \text{ fOr eveRy fixed } s>0.$$
we assume tHAT the distAncE beTweEn aNY PaIr of dislocatiON PoinTs Is at leaSt $2\rHo_{\varepSilOn}$ aNd wE deFiNe the clasS $X_{{\varepsIlOn}}$ Of AdMisSible DIslocatiOnS by $$\BeGin{GatheREd}
\labeL{xn}
X_{{\vArepSiLoN}}:= \big\{\Mu\in {{\matHCaL m}}({\omegA};{\mAtHbb{R}}^2): \, \Mu=\sUm_{I=1}^M \xi_i \, \DeltA_{X_i}, \, M\In{\mathbB{N}},\, B_{{\rho_{\varEpsILon}}}(x_I) \sUbSet {\ | , we introduce the class o f Burgersvecto rs$S= \{ b_1, ..., b_s\}$. In wha t fol lows we assume that $S $ con ta i ns a t l easttwo (in d ep e n den t) v ect or s ,so th at$$\labe l{R2}
{\rm Sp an }_{{\mathbb{ R }} } S={\math bb{ R}}^2$$ andthi s will i mpl y that th e fun ction$ \varph i$ in the e n ergy ( \ [Gammal i m it \])is finite in whol e o f ${\mathbb{R}} ^2$. T he ca s e of on ly one Bur ge rs ve c tor ise as i e r an d it implies t hat $\varph i $ i s fini te on l y on a onedi m ens ional subsp aceof ${\mat hbb{R} } ^2$. We denoteby $\m ath bb{ S}$t he s pan o f $S $ w ith int eger coe ff ic ients ($\ m a t h bb{S }={ \rm Span }_{{\mathbb{Z }}} S$), i.e ., th e set ofBu rgers vecto rs fo r“multiple dislo cati ons".
As in S ect io n \[S c alings \], $N _{\vare psilon} $ re pr e s e nt s the number of di sl o c at ions pre sent i n t he crystal, c orr espo n d ing t o th e i nternalscale$ {\ va repsilo n} $. Wein tro duc e als o the seque nce $\rh o_{\v a repsilon}$ rep r esenting ther ad i u so f th e h ard core su rrou n ding the di slo c ation s, an dw er equire
- $\lim_{ {\ vareps ilon} \to 0} \frac{ \rho_{\var e p s ilon}}{{ \var e ps i lon}_n^s} = \i nfty$ for every fixed $0 <s<1$ ;
- $ \lim_{{\v a r epsilon} \to 0} |N _{\ v a re psilon}| \rho _ { \var ep silon}^ 2 = 0$.
C ond iti oni)sa ys that t he hardco re r eg ion cont a ins almo st al lthe self - energy , whi le c on di t ion ii) sa y st h at t he a reaofth e har d co r e r egion t ends to z ero , and h en ce that its compleme nt containsal mos t allt h e intera ction energy. Note that conditi ons i) a nd i i) are co mpa tiblewhe n ever
$$\lab el{ip n}
N_ { \ varep s i lo n}{ \v arepsilon} ^ s \t o 0 \q quad \text{ for every fixed } s>0 .$$
We assum e t hatt h edis t an c e b et w een a ny pair of disl ocation po in t sis at leas t $2 \r ho_{\va repsilo n}$ a n d we de fine theclass $X_ {{ \var e p sil on}}$ of a dmissibl e disloca t ionsb y$$\be gin {gathe re d}\labe l{xn}X _{{ \vare psilon }} := \Bi g\{\m u\ in {{\ma thcal M}}({\Omega};{\ma thbb{R }}^2) : \ , \mu=\su m_{ i =1} ^M \xi_i\, \ delta_{x_i },\,M\in{ \ma t hbb{N }},\ , B _{{ \ rho_{ \var e psilon}}} ( x_ i)\ s ub set {\ | , we_introduce the_class of Burgers vectors_$S=\{b_1,...,b_s\}$. In_what_follows we_assume_that $S$ contains_at least two_(independent) vectors, so that_$$\label{R2}
{\rm Span}_{{\mathbb{R}}} S={\mathbb{R}}^2$$_and_this will imply that the function $\varphi$ in the energy (\[Gammalimit\]) is finite in_whole_of ${\mathbb{R}}^2$._The_case_of only one Burgers vector_is easier and it implies_that $\varphi$_is finite only on a one dimensional subspace_of_${\mathbb{R}}^2$. We denote_by $\mathbb{S}$ the span of $S$ with integer coefficients_($\mathbb{S}={\rm Span}_{{\mathbb{Z}}}S$), i.e., the set_of Burgers vectors_for_“multiple_dislocations".
As in Section \[Scalings\],_$N_{\varepsilon}$ represents the number of dislocations_present in the crystal, corresponding to_the internal scale ${\varepsilon}$. We introduce also_the sequence $\rho_{\varepsilon}$ representing the radius_of the hard core surrounding_the dislocations,_and we require
- _$\lim_{{\varepsilon}\to 0} \frac{\rho_{\varepsilon}}{{\varepsilon}_n^s}_= \infty$_for every fixed_$0<s<1$;
- $\lim_{{\varepsilon}\to 0} |N_{\varepsilon}|_\rho_{\varepsilon}^2 = 0$.
Condition_i) says that the hard core_region_contains almost all_the_self-energy,_while condition_ii) says that_the_area of_the_hard core region tends to zero,_and_hence that its complement contains almost all_the interaction energy. Note_that_conditions i) and ii)_are compatible whenever
$$\label{ipn}
N_{\varepsilon}{\varepsilon}^s \to 0_ \qquad \text{ for every fixed_} s>0.$$
We_assume that_the distance between any pair of dislocation points is at least_$2\rho_{\varepsilon}$ and we define the class_$X_{{\varepsilon}}$ of admissible dislocations_by $$\begin{gathered}
\label{xn}
X_{{\varepsilon}}:=_\Big\{\mu\in_{{\mathcal M}}({\Omega};{\mathbb{R}}^2): \,_\mu=\sum_{i=1}^M_\xi_i \,_\delta_{x_i}, \, M\in{\mathbb{N}},\, B_{{\rho_{\varepsilon}}}(x_i) \subset {\ |
}$, we have $$\hat{\ell} _A(f)=\cases$$(d\_s+iae\^[-s]{})f\_[(s,t)]{} $$&\text{if $\tilde {A}$ is defined by (6) }\\$$(d\_s+ibe\^s)f\_[(s,t)]{} $$&\text{if $\tilde {A}$ is defined by (7)}\\$$(d\_s +i\[ae\^[-s]{}+ be\^s\])f\_[(s,t)]{} $$&\text{if $\tilde {A}$ is defined by (8)}$$\
$$\Big((d+b\gamma e^s)\partial_s+$$\
$$i[ae^{-s}+b(\alpha\beta-\gamma\delta)e^s +c\gamma ]\Big)f\vert_{(s,t)}.&\text{if $\tilde {A}$ is defined by (9)$$}\\$$((d-be\^s)\_s+$$\\$$i\[-ae\^[-s]{}-b(-)e\^s +c\])f\_[(s,t)]{}.& $$ ]{}
With above notations we obtain the operators : $$\hat{\ell} _A=\cases$$\_A\^[(2)]{}=(d\_s+iae\^[-s]{})\_[(s,t)]{} $$&\\$$\_A\^[(3)]{}=(d\_s+ibe\^s)\_[(s,t)]{} $$&\\$$\_A\^[(4)]{}=(d\_s+i\[ae\^[-s]{}+ be\^s\])\_[(s,t)]{} $$& \\$$\_A\^[(5)]{}=((d+be\^s)\_s+i\[ae\^[-s]{}+b(-)e\^s+c\])\_[(s,t)]{}.&\
$$\hat\ell_A^{(5')}=\Big((d-b\gamma e^s)\partial_s+i[-ae^{-s}-b(\alpha\beta-\gamma\delta)e^s+c\gamma]\Big)\vert_{(s,t)}
\endcases$$ which provides the representations of the Lie algebra $\mathfrak g$=$Lie(\mathbb R\ltimes \mathbb H_3)$.
Furthermore, $\forall A,B \in\mathfrak g$, $$\hat\ell_A\circ\hat\ell_ B- \hat\ell_ | } $, we have $ $ \hat{\ell } _ A(f)=\cases$$(d\_s+iae\^[-s]{})f\_[(s, t) ] { } $ $ & \text{if $ \tilde { A}$ is defined by (6) } \\$$(d\_s+ibe\^s)f\_[(s, t) ] { } $ $ & \text{if $ \tilde { A}$ is defined by (7)}\\$$(d\_s + i\[ae\^[-s]{}+ be\^s\])f\_[(s, t) ] { } $ $ & \text{if $ \tilde { A}$ is defined by (8)}$$\
$ $ \Big((d+b\gamma e^s)\partial_s+$$\
$ $ i[ae^{-s}+b(\alpha\beta-\gamma\delta)e^s + c\gamma ] \Big)f\vert_{(s, t)}.&\text{if $ \tilde { A}$ is define by (9)$$}\\$$((d - be\^s)\_s+$$\\$$i\[-ae\^[-s]{}-b(-)e\^s + c\])f\_[(s, t) ] { }. & $ $ ] { }
With above notation we obtain the operators: $ $ \hat{\ell } _ A=\cases$$\_A\^[(2)]{}=(d\_s+iae\^[-s]{})\_[(s, t) ] { } $ $ & \\$$\_A\^[(3)]{}=(d\_s+ibe\^s)\_[(s, t) ] { } $ $ & \\$$\_A\^[(4)]{}=(d\_s+i\[ae\^[-s]{}+ be\^s\])\_[(s, t) ] { } $ $ & \\$$\_A\^[(5)]{}=((d+be\^s)\_s+i\[ae\^[-s]{}+b(-)e\^s+c\])\_[(s, t)]{}.&\
$ $ \hat\ell_A^{(5')}=\Big((d - b\gamma e^s)\partial_s+i[-ae^{-s}-b(\alpha\beta-\gamma\delta)e^s+c\gamma]\Big)\vert_{(s, t) }
\endcases$$ which provides the representation of the Lie algebra $ \mathfrak g$=$Lie(\mathbb R\ltimes \mathbb H_3)$.
Furthermore, $ \forall A, B \in\mathfrak g$, $ $ \hat\ell_A\circ\hat\ell _ B- \hat\ell _ | }$, we have $$\hat{\ell} _A(f)=\cases$$(d\_s+iat\^[-s]{})f\_[(s,t)]{} $$&\text{if $\tildg {Q}$ is vefined by (6) }\\$$(d\_s+ice\^s)f\_[(s,t)]{} $$&\text{if $\tilde {A}$ is defmned by (7)}\\$$(e\_s +i\[ae\^[-s]{}+ be\^s\])f\_[(s,t)]{} $$&\text{if $\gilde {A}$ id definee by (8)}$$\
$$\Vig((d+b\gamma e^s)\partial_s+$$\
$$i[ae^{-a}+n(\alphc\bxta-\gamma\delta)e^s +c\gamma ]\Bic)f\vert_{(s,t)}.&\text{if $\tklbe {A}$ is defined by (9)$$}\\$$((d-be\^s)\_s+$$\\$$i\[-ae\^[-s]{}-b(-)e\^s +c\])f\_[(s,e)]{}.& $$ ]{}
With ahove notations we pftaih the operators : $$\hat{\ell} _A=\cases$$\_A\^[(2)]{}=(d\_s+ize\^[-s]{})\_[(s,t)]{} $$&\\$$\_A\^[(3)]{}=(v\_s+ibe\^s)\_[(s,t)]{} $$&\\$$\_A\^[(4)]{}=(d\_s+i\[ae\^[-x]{}+ be\^s\])\_[(s,t)]{} $$& \\$$\_A\^[(5)]{}=((d+be\^s)\_s+i\[ae\^[-s]{}+b(-)e\^s+c\])\_[(s,t)]{}.&\
$$\hwt\elp_A^{(5')}=\Big((d-b\gamma e^s)\parhial_s+i[-ae^{-s}-b(\aopha\fwta-\gamma\deltx)e^s+c\gamma]\Bpy)\vert_{(s,t)}
\enddases$$ which provides the represdntatnons of the Lue wngebra $\mathhrak g$=$Jie(\mathbb R\lbpmes \madhbb H_3)$.
Firthermore, $\forsll A,B \in\mathfrak g$, $$\hat\ell_A\rirc\hat\ell_ B- \hat\ell_ | }$, we have $$\hat{\ell} _A(f)=\cases$$(d\_s+iae\^[-s]{})f\_[(s,t)]{} $$&\text{if $\tilde defined (6) }\\$$(d\_s+ibe\^s)f\_[(s,t)]{} $\tilde {A}$ is $$&\text{if {A}$ is defined (8)}$$\ $$\Big((d+b\gamma e^s)\partial_s+$$\ +c\gamma ]\Big)f\vert_{(s,t)}.&\text{if $\tilde {A}$ is defined (9)$$}\\$$((d-be\^s)\_s+$$\\$$i\[-ae\^[-s]{}-b(-)e\^s +c\])f\_[(s,t)]{}.& $$ ]{} With above notations we obtain the operators : $$\hat{\ell} $$&\\$$\_A\^[(3)]{}=(d\_s+ibe\^s)\_[(s,t)]{} $$&\\$$\_A\^[(4)]{}=(d\_s+i\[ae\^[-s]{}+ be\^s\])\_[(s,t)]{} $$& \\$$\_A\^[(5)]{}=((d+be\^s)\_s+i\[ae\^[-s]{}+b(-)e\^s+c\])\_[(s,t)]{}.&\ $$\hat\ell_A^{(5')}=\Big((d-b\gamma e^s)\partial_s+i[-ae^{-s}-b(\alpha\beta-\gamma\delta)e^s+c\gamma]\Big)\vert_{(s,t)} \endcases$$ which provides the representations of Lie $\mathfrak R\ltimes H_3)$. Furthermore, $\forall A,B \in\mathfrak g$, $$\hat\ell_A\circ\hat\ell_ B- \hat\ell_ | }$, we have $$\hat{\ell} _A(f)=\cases$$(d\_s+iae\^[-s]{})F\_[(s,t)]{} $$&\text{if $\tIlde {A}$ Is dEfiNeD by (6) }\\$$(d\_S+ibe\^S)f\_[(s,t)]{} $$&\text{if $\tildE {a}$ is dEfined by (7)}\\$$(d\_s +i\[ae\^[-s]{}+ be\^s\])f\_[(s,t)]{} $$&\teXt{if $\tIlDE {A}$ is DEfIned bY (8)}$$\
$$\Big((d+b\gAMmA E^S)\paRtIaL_s+$$\
$$i[Ae^{-S}+B(\aLpha\bEta-\Gamma\deLta)e^s +c\gammA ]\BiG)f\Vert_{(s,t)}.&\text{if $\TIlDe {A}$ is definEd bY (9)$$}\\$$((d-be\^s)\_s+$$\\$$i\[-ae\^[-s]{}-b(-)e\^S +c\])f\_[(S,t)]{}.& $$ ]{}
With AbOve NOtatiOns We obtAin the OPeratoRs : $$\hat{\ell} _A=\CaSEs$$\_A\^[(2)]{}=(d\_s+iAE\^[-s]{})\_[(s,t)]{} $$&\\$$\_A\^[(3)]{}=(d\_s+IBE\^s)\_[(S,t)]{} $$&\\$$\_A\^[(4)]{}=(d\_S+i\[ae\^[-s]{}+ be\^s\])\_[(s,t)]{} $$& \\$$\_A\^[(5)]{}=((d+be\^s)\_s+i\[AE\^[-s]{}+B(-)E\^s+c\])\_[(s,t)]{}.&\
$$\hat\ell_A^{(5')}=\BiG((d-b\gamMa E^S)\pARTiaL_s+i[-Ae^{-s}-b(\alpha\bEtA-\gammA\Delta)e^s+C\GaMMA]\big)\VErt_{(s,t)}
\endcases$$ Which providES thE repreSeNtaTIons of The LiE aLGebRa $\mathfrak g$=$lie(\mAthbb R\ltiMes \matHBb H_3)$.
FurtHErmore, $\fOrall A,b \in\MatHfraK G$, $$\hAt\Ell_a\cIRc\hAT\eLl_ B- \HAt\eLl_ | }$, we have $$\hat{\ell} _ A(f)=\case s$$(d \_s +ia e\ ^[-s ]{}) f\_[(s,t)]{} $ $ &\te xt{if $\tilde {A}$ is defi ne d by( 6) }\\$ $(d\_s+ i be \ ^ s)f \_ [( s,t )] { }$$&\t ext {if $\t ilde {A}$isde fined by (7 ) }\ \$$(d\_s + i\[ ae\^[-s]{}+be\ ^s\])f \_ [(s , t)]{} $$ &\tex t{if $ \ tilde{A}$ is d ef i ned by (8)}$$\ $ $\ Big( (d+b\gamma e^s)\p a rt i al_s+$$\
$$i[a e^{-s} +b ( \a l p ha\ bet a-\gamma\d el ta)e^ s +c\gam m a] \ B ig) f \vert_{(s,t)} .&\text{if$ \ti lde {A }$ is define d by ( 9 )$$ }\\$$((d-be \^s) \_s+$$\\$ $i\[-a e \^[-s]{ } -b(-)e\ ^s +c\ ])f \_[ (s,t ) ]{ }. & $ $] {}Wi tha bov e notati on swe ob tain t h e ope rat ors: $$\ hat{\ell} _A= \ca ses$ $ \_A \^[(2 )]{}= (d\_ s+ iae\^ [-s]{} )\_[( s, t)]{} $$&\\$$\_ A\^[ (3)]{}=(d \_s +i be\ ^s )\_[( s ,t)]{} $$ &\\ $$\_A\^ [(4)]{} = (d\ _s + i \ [a e\^[-s]{}+ be\^s\] )\ _ [ (s ,t)]{} $ $& \\$ $ \_ A\ ^ [(5)]{}= (( d+b e\^s ) \ _s+i\ [ae\ ^ [- s]{}+b(- )e\^s+ c \] )\ _[(s,t) ]{ }.&\
$ $\ hat \el l_A^{ ( 5')} =\Big( (d-b\gam ma e^ s )\partial_s+i[ - ae^{-s}-b(\al p ha \ b et a -\ga mma \delta)e^s+ c\ga m ma]\ Big) \ ve rt_ { (s,t) }
\e nd c as e s$$ which providesth e repr esent ations of the Lie algeb r a $\mathfr ak g $ =$ L ie(\mathbb R\l times \mathbb H _ 3)$.
Fu rther more, $\ forall A, B \in\math fra k g $,$$\ h a t\ ell_A\circ\ha t \ ell_ B - \hat\ ell _ | }$, we_have $$\hat{\ell}__A(f)=\cases$$(d\_s+iae\^[-s]{})f\_[(s,t)]{} $$&\text{if $\tilde {A}$_ is_defined_by (6)_}\\$$(d\_s+ibe\^s)f\_[(s,t)]{}_$$&\text{if $\tilde {A}$_is defined by_ (7)}\\$$(d\_s +i\[ae\^[-s]{}+ be\^s\])f\_[(s,t)]{}_$$&\text{if $\tilde {A}$_is_defined by (8)}$$\
$$\Big((d+b\gamma e^s)\partial_s+$$\
$$i[ae^{-s}+b(\alpha\beta-\gamma\delta)e^s +c\gamma ]\Big)f\vert_{(s,t)}.&\text{if $\tilde {A}$ is defined by (9)$$}\\$$((d-be\^s)\_s+$$\\$$i\[-ae\^[-s]{}-b(-)e\^s +c\])f\_[(s,t)]{}.&_$$_]{}
With above_notations_we_obtain the operators : $$\hat{\ell}__A=\cases$$\_A\^[(2)]{}=(d\_s+iae\^[-s]{})\_[(s,t)]{} $$&\\$$\_A\^[(3)]{}=(d\_s+ibe\^s)\_[(s,t)]{} $$&\\$$\_A\^[(4)]{}=(d\_s+i\[ae\^[-s]{}+ be\^s\])\_[(s,t)]{} $$&_\\$$\_A\^[(5)]{}=((d+be\^s)\_s+i\[ae\^[-s]{}+b(-)e\^s+c\])\_[(s,t)]{}.&\
$$\hat\ell_A^{(5')}=\Big((d-b\gamma e^s)\partial_s+i[-ae^{-s}-b(\alpha\beta-\gamma\delta)e^s+c\gamma]\Big)\vert_{(s,t)}_
\endcases$$ which provides the representations of the Lie_algebra_$\mathfrak g$=$Lie(\mathbb R\ltimes_\mathbb H_3)$.
Furthermore, $\forall A,B \in\mathfrak g$, $$\hat\ell_A\circ\hat\ell_ B- \hat\ell_ |
similar nature.
Secondly, we did not compare serendipitous interactions to any performance metric. In the research lab, it is difficult to find a well-defined performance metric apart from publications and research impact, which do not change frequently enough to warrant real-time tracking of informal interactions. However, this does not mean that such interactions do not have an important effect, just that it is difficult to measure formally using existing metrics. As a next step, we are planning to collect subjective daily measures (e.g., frustration, satisfaction) from our participants using experience sampling, and to investigate any relationship between these measures and sensed serendipitous interactions.
A third limitation is that we have not studied exactly how the sensing of employees’ inter-group interactions, and meetings in serendipitous spaces, could be used in practice by an organization, for example, to foster more such interactions. One could envisage that making workers aware of the sensed information could prompt them to consider their own interactions and might have a positive effect on productivity by increasing the exchange of information and ideas.
Conclusions
===========
In this work, we have used state-of-the-art active RFID tags to track serendipitous interactions in the workplace between individuals from different groups, in a less obtrusive and therefore potentially more accurate way than was possible using previous sensing technology.
Our results suggest that cultural differences between individuals in a highly international office environment can affect the likelihood that people engage in serendipitous interactions with others from different groups. We have also been able to characterize certain spaces in the workplace, such as kitchens and printers, as being particularly likely to host serendipitous interactions, and shown that different kinds of spaces may be more likely to host inter-group interactions, depending on the kind of group considered. Knowledge of these effects could be important for organizations, given that such interactions have been shown by previous work to be beneficial for group productivity, by enabling the exchange of information and ideas.
Acknowledgements
================
We thank the ISI Foundation and Bitmanufaktur for their support. We acknowledge the SocioPatterns collaboration (`http://www.sociopatterns.org`). We also thank Jon Crowcroft, and Kerstin Sailer for valuable discussions and feedback. Chloë Brown is a recipient of the Google Europe Fellowship in Mobile Computing, and this research is supported in part by this Google Fellowship.
| similar nature.
Secondly, we did not compare serendipitous interactions to any performance measured. In the inquiry lab, it is unmanageable to find a well - defined performance measured apart from publications and research impingement, which do not change frequently enough to warrant actual - time tracking of cozy interactions. However, this does not mean that such interactions do not have an authoritative effect, just that it is difficult to measure formally using exist metrics. As a next step, we are design to collect subjective daily measures (for example, frustration, satisfaction) from our player using experience sampling, and to investigate any relationship between these measure and sensed serendipitous interactions.
A third limitation is that we have not studied exactly how the sensing of employee ’ inter - group interaction, and meetings in serendipitous spaces, could be used in practice by an organization, for example, to foster more such interactions. One could envisage that making worker aware of the sensed information could motivate them to think their own interaction and might have a positive consequence on productivity by increasing the exchange of information and theme.
Conclusions
= = = = = = = = = = =
In this work, we have used state - of - the - artwork active RFID tags to track serendipitous interactions in the workplace between individuals from different groups, in a less obtrusive and therefore potentially more accurate manner than was possible using previous sensing technology.
Our result suggest that cultural differences between individuals in a highly external office environment can feign the likelihood that people engage in serendipitous interaction with others from different group. We have also been able to characterize sealed spaces in the workplace, such as kitchens and printers, as being particularly probable to host serendipitous interactions, and shown that unlike kinds of spaces may be more likely to host inter - group interactions, depending on the kind of group view. Knowledge of these effects could be important for organizations, given that such interactions have been shown by previous work to be beneficial for group productiveness, by enabling the exchange of data and ideas.
Acknowledgements
= = = = = = = = = = = = = = = =
We thank the ISI Foundation and Bitmanufaktur for their support. We notice the SocioPatterns collaboration (` http://www.sociopatterns.org `). We also thank Jon Crowcroft, and Kerstin Sailer for valuable discussions and feedback. Chloë Brown is a recipient of the Google Europe Fellowship in Mobile Computing, and this inquiry is supported in part by this Google Fellowship. | sililar nature.
Secondly, we aid not compare serendmpitous interacgions to any performance metcic. Un tht research lab, it ir difficupt to fibd a qell-definev performance jctric epart from publlcations ang research impdcg, chich do not change frequently enougr to watrwnt real-time ttackimd of pnnormal interactions. However, this does nmt mean that xuch interactions do not hwve wn important effech, just that it yw difficult go measure formally usjng existing metrics. As a next rtep, ce are planbibg hm collect snbjectpve daily measures (e.g., frustrstion, satisfacbion) hrom our participants usiig experience sampligg, and to iuvestigate any relatiinwhip tetwaen gyesd mtasnrea and denaed serendjpitous intwractions.
A third likieqtion is that we hade not studied exactly how the sensing of emlloyees’ inter-group inteeactions, and meetings in serenqipitous spaces, could be used in practice by an osganivagiou, for dzalple, to foster more such interactions. One coujs tnvpsage that making workers awsrf ps the sensed knformcfikn could prompt thfm to cjnsidwr their jwn onteractions and might have a positive vffext on productivity by increasnng thg exchsnge of information and idezs.
Conclusiojs
===========
In this dork, we have usea snate-mf-the-art active RFID tags eo track wereudipitour injeractijns in the workinace between indivlduald xrom diffegent groups, in a less obtrusive and therefore pmtettially iore sccurate way ehan was possiyle usiny prevkous sensihg techiology.
Our refults suggest jhat cultural differegces betqeen inakviduals in a nighly innexnational office environmenb can zffect the likeoihiod that people eneagq pn xewandipitous itterxctkpns wkth othtxs nroo digferent groups. We haee amso been able to cnavacterize certain spaces in thr workplace, such ad kitrhens end prontgrs, as being particularly likelg to host sevendipitous igterqctions, and xhown that different kinds of spaces mab be more likely to hosr inter-group interaetlons, dependiig on ehe kind mf group considered. Jnowledge of thest effects could be impodtant xor ogganizations, given that such interactions have been shown by previous worj to bx feneficial ror broup pxodbctivitr, by xnabling the exchsnge of information and ideas.
Acynowledgamznts
================
We thank the ISI Foundatipn and Bitmanufxktur for their support. Se acknoeledge the SocioPatterns collabpration (`http://www.sociopattcrns.org`). Ww alsk thank Joi Crowcroft, and Kersyin Sanler fir valuqble discussions xnd fcedback. Chpoë Bcowu is a recipient of the Google Eueope Fwlpowship in Mobilx Computing, wkd thia resewrch is supportes in oart cy this Googlr Fellowship.
| similar nature. Secondly, we did not compare to performance metric. the research lab, a performance metric apart publications and research which do not change frequently enough warrant real-time tracking of informal interactions. However, this does not mean that such do not have an important effect, just that it is difficult to measure using metrics. a step, we are planning to collect subjective daily measures (e.g., frustration, satisfaction) from our participants using sampling, and to investigate any relationship between these and sensed serendipitous interactions. third limitation is that we not exactly how sensing employees’ interactions, and meetings serendipitous spaces, could be used in practice by an organization, for example, to foster more such interactions. could envisage workers aware the information prompt them to own interactions and might have a productivity by increasing the exchange of information and Conclusions =========== this work, we have used state-of-the-art RFID tags to track serendipitous interactions in the between individuals from different groups, in a less obtrusive and therefore potentially more accurate way possible using previous sensing Our results suggest cultural between in highly international environment can affect the likelihood that people engage in serendipitous interactions others from different groups. We have also been able to spaces the workplace, such kitchens and printers, as particularly to host serendipitous interactions, that kinds be likely host inter-group interactions, depending the kind of group considered. of these effects could that such interactions have been shown by previous to be beneficial for group productivity, by the exchange of information and ideas. Acknowledgements ================ We thank the ISI and Bitmanufaktur support. We acknowledge the SocioPatterns collaboration (`http://www.sociopatterns.org`). We thank Jon Crowcroft, and Sailer for valuable discussions and feedback. Chloë Brown is recipient the Google Fellowship in Mobile and this research supported in part Google Fellowship. | similar nature.
Secondly, we diD not comparE sereNdiPitOuS intEracTions to any perfORmanCe metric. In the research lAb, it iS dIFficULt To finD a well-dEFiNED peRfOrManCe MEtRic apArt From pubLications aNd rEsEarch impact, wHIcH do not chanGe fRequently enoUgh To warrAnT reAL-time TraCking Of infoRMal intEractions. hoWEver, thIS does noT MEaN thaT such interactions DO nOT have an importaNt effeCt, JUsT THat It iS difficult To MeasuRE formalLY uSING exISting metrics. AS a next step, wE Are PlanniNg To cOLlect sUbjecTiVE daIly measures (E.g., frUstration, SatisfACtion) frOM our parTicipaNts UsiNg exPErIeNce SaMPliNG, aNd tO InvEstigate AnY rElatiOnshIP BETweeN thEse mEasurEs and sensed seRenDipiTOus InterActioNs.
A tHiRd limItatioN is thAt We have not studieD exaCtly how thE seNsIng Of EmploYEes’ intEr-gRouP interaCtions, aND meEtINGS iN serendipitous spacEs, COUlD be used iN practICe By AN organizAtIon, For eXAMple, tO fosTEr More such InteraCTiOnS. One couLd EnvisaGe ThaT maKing wORkerS aware Of the senSed inFOrmation could pROmpt them to conSIdER ThEIr owN inTeractions aNd miGHt haVe a pOSiTivE EffecT on prOdUCtIVity by increasing the ExChange Of infOrmation and idEas.
ConclusIONS
===========
In this wOrk, wE HaVE used state-of-thE-art aCtive RFID tAGs to tracK sereNdipitouS interactIONs in the wOrkPlaCe bEtwEEN iNdividuals froM DIffeReNt groupS, in A less obTruSivE anD thErEfore poteNtially mOrE aCcUrAte Way thAN was possIbLe uSiNg pRevioUS sensiNg tecHnolOgY.
OUR reSults suGGeST That CuLtUral DifFeRenceS betWEen IndividUals in a hiGhlY InteRnAtIonal ofFice environmeNt Can affect tHe LikElihooD THat peoplE engage in serendipitous iNTeractiOns With oTherS from diffEreNt grouPs. WE Have alSo been Able tO cHarACTerizE CErTaiN sPaces in the WORkpLace, sUcH as kItchens And printers, as being PArtIcularly likelY to Host SEReNdiPItOUs iNtERacTIOns, and shown that Different kInDS oF spaces may BE moRe Likely tO host inTer-grOUp interActions, dePending on ThE kinD OF grOup consideRed. KnowlEdge of theSE effeCTs Could Be iMportaNt For OrganIzatioNS, giVen thAt such InTeractIons hAvE been shoWn by previous work to be benEficiaL for gRouP productiVitY, By eNabling thE excHange of infOrmAtiOn and IdeAS.
AcknOwleDGeMenTS
================
We thAnk tHE ISI FoundATiOn aND biTmanufaktur FOR TheIr supPorT. we acknOwleDge the SocioPatterNS collaboration (`Http://WWW.soCioPAtteRnS.org`). We also thanK JoN CROWcroft, anD KErstin SaileR for valuAbLE discUssionS and feEdback. CHLOë bRown is A recIpiEnt of the GOogLe eUrope FeLlOwSHip in MObilE COmputiNg, and tHIs reSEArch is supported iN part BY This GOOglE FellOwShip.
| similar nature.
Secondly , we did n ot co mpa rese rend ipit ous interactio n s to any performance metri c. In t h e re s ea rch l ab, iti sd i ffi cu lt to f i nd a we ll- defined performan ceme tric apart f r om publicati ons and researc h i mpact, w hic h do n otchang e freq u entlyenough to w a rrantr eal-tim e tr acki ng of informal in t er a ctions. Howeve r, thi sd oe s not me an that su ch inte r actions do n o t h a ve an importa nt effect,j ust thatit is diffic ult t om eas ure formall y us ing exist ing me t rics. A s a next step, we ar e pl a nn in g t oc oll e ct su b jec tive dai ly m easur es ( e . g . , fr ust rati on, s atisfaction)fro m ou r pa rtici pants usi ng expe rience samp li ng, and to inve stig ate any r ela ti ons hi p bet w een th ese me asuresand sen s edse r e n di pitous interaction s. Athird li mitati o nis that weha venots t udied exa c tl y how th e sens i ng o f emplo ye es’ in te r-g rou p int e ract ions,and meet ingsi n serendipitou s spaces, coul d b e us e d in pr actice by a n or g aniz atio n ,for examp le, t of os t er more such intera ct ions.One c ould envisage that maki n g workersawar e o f the sensed in forma tion could prompt t hem t o consid er theiro w n intera cti ons an d m i g ht have a posit i v e ef fe ct on p rod uctivit y b y i ncr eas in g the exc hange of i nf or ma tio n and ideas.
Co ncl us ion s
=== = ====== =
In thi swo r k,we have us e d sta te -o f-th e-a rt acti ve R F IDtags to track se ren d ipit ou sinterac tions in thewo rkplace be tw een indiv i d uals fro m different groups, ina less o btr usive and therefor e p otenti all y moreaccura te wa ytha n was p o s si ble u sing previ o u s s ensin gtech nology.
Our results sugg e stthat cultural di ffer e n ce s b e tw e enin d ivi d u als in a highly internati on a loffice env i ron me nt canaffectthe l i kelihoo d that pe ople enga ge ins e ren dipitous i nteracti ons witho thers fr om di ffe rent g ro ups . Wehave a l sobeenable t ocharac teriz ecertainspaces in the workplace , such as k itc hens andpri n ter s, as bei ng p articularl y l ike ly to ho s t ser endi p it ous inter acti o ns, and s h ow n t h a tdifferent k i n d s o f spa ces may be mor e likely to hosti nter-group int erac t i ons , d e pend in g on the kindofgr o u p consid er ed. Knowled ge of th es e effe cts co uld be import a n tf or org aniz ati ons, give n t ha t such i nt er a ctions hav ebeen s hown b y pre v i ous work to be b enefi c i al fo r gr oup p ro ductivi t y, b y enabling the exchan ge ofinfo rmati on andid eas.
Ack no wledgement s
======== ===== ===
We t hank th e ISIFoun d a tionandBi tma nufakturf o rt he ir sup port . Weac know ledge the SocioPat ter n s colla bo rat i o n (`ht t p: / / www.sociop att erns. o r g`). We al s o th a nk Jon C rowcro ft, an d Kerst i n S ai ler for va l u able disc ussions a n d f ee dbac k. Chloë B rown i s a r ecipient of the GoogleEurope F e l lowshi p i n Mob i le Com pu ting, and th is re se a rch iss u ppo rted in par t b y t hisGoo gleF ellowship.
| similar_nature.
Secondly, we_did not compare serendipitous_interactions to_any_performance metric._In_the research lab,_it is difficult_to find a well-defined_performance metric apart_from_publications and research impact, which do not change frequently enough to warrant real-time tracking_of_informal interactions._However,_this_does not mean that such_interactions do not have an_important effect,_just that it is difficult to measure formally_using_existing metrics. As_a next step, we are planning to collect subjective_daily measures (e.g., frustration, satisfaction) from_our participants using_experience_sampling,_and to investigate any_relationship between these measures and sensed_serendipitous interactions.
A third limitation is that_we have not studied exactly how the_sensing of employees’ inter-group interactions, and_meetings in serendipitous spaces, could_be used_in practice by an organization,_for example, to_foster more_such interactions. One_could envisage that making workers aware_of the sensed_information could prompt them to consider_their_own interactions and_might_have_a positive_effect on productivity_by_increasing the_exchange_of information and ideas.
Conclusions
===========
In this work,_we_have used state-of-the-art active RFID tags to_track serendipitous interactions in_the_workplace between individuals from_different groups, in a less_obtrusive and therefore potentially more accurate_way than_was possible_using previous sensing technology.
Our results suggest that cultural differences between individuals_in a highly international office environment_can affect the likelihood_that people_engage_in serendipitous interactions_with_others from_different groups. We have also been able_to characterize_certain spaces in the workplace, such_as kitchens and printers,_as_being particularly likely to host serendipitous_interactions, and shown that different kinds_of spaces may be more_likely_to_host inter-group interactions, depending on_the kind of group considered. Knowledge_of these effects_could be important for organizations, given that_such_interactions have been shown by previous_work_to be beneficial for group productivity,_by_enabling_the exchange of information and_ideas.
Acknowledgements
================
We thank the ISI Foundation and_Bitmanufaktur for their support. We acknowledge the SocioPatterns collaboration_(`http://www.sociopatterns.org`). We also_thank Jon Crowcroft, and Kerstin_Sailer_for_valuable discussions and feedback. Chloë Brown is a recipient of_the Google_Europe Fellowship in_Mobile Computing, and this research is supported in part by_this Google Fellowship.
|
BB-8]{}, and [Toymail]{}; the rest have not responded even after 6 months. Overall, most responses (and lack thereof) indicate that these companies take privacy matters much more frivolously than expected.
Related work {#sec:related}
============
Smart electronics for children have made the news for security breaches numerous times in the past; e.g., see Rapid7 [@rapid72015iot] (strangers could hack baby monitors to view children sleeping and even talk to them directly), Norwegian Consumer Council [@bouvet2016toys] (highlight dangerous security vulnerabilities in 3 toys: i-Que, My Friend Cayla, and [Hello Barbie]{}), Motherboard [@motherboard2016gps] (strangers could track children’s location and message them through the HereO smart watch), and BBC [@bbcgermany2017cayla] (German ban on Cayla in February 2017, branding it as an “illegal espionage apparatus”). Like wearables for adults, many smart toys are Bluetooth-enabled and accompany children outside the home, and previous research on fitness trackers has uncovered numerous vulnerabilities such as location tracking [@Hilts2016fitness].
Several security breaches involving the [VTech InnoTab MAX]{} have been reported. In 2015, VTech’s Learning Lodge database, used by the VTech suite of children’s tablets, suffered a severe breach, exposing the PII of 6.3 million children’s profiles [@vtech2017breach]. The breach was due to a combination of insecure practices, e.g., PII was sent in plaintext (no HTTPS), the database was vulnerable to a simple SQL injection attack, and account passwords were stored using an MD5 hash (no salting). The breach ultimately led to a \$650,000 settlement with the U.S. Federal Trade Commission for unauthorized collection of children’s personal information and for failing to secure that information [@ftc2018vtechsettlement], and a related finding from Canada’s Privacy Commissioner that VTech was in breach of its privacy laws (but no corresponding settlement) [@pipeda2018vtech]. Also in 2016, the UK-based security firm Pen Test Partners found that the [VTech InnoTab MAX]{} is vulnerable to trivial data extraction [@ptpvtech2015extraction]. (We found that this flaw has not been fixed.) Later in 2016, it was found that | BB-8 ] { }, and [ Toymail ] { }; the rest have not responded even after 6 months. Overall, most responses (and lack thence) bespeak that these companies take privacy count much more frivolously than expect.
Related oeuvre { # sec: relate }
= = = = = = = = = = = =
Smart electronics for child have made the news for security system breaches numerous times in the past; e.g., see Rapid7 [ @rapid72015iot ] (stranger could hack baby monitors to view children sleep and even talk to them directly), Norwegian Consumer Council [ @bouvet2016toys ] (highlight dangerous security system vulnerabilities in 3 toys: i - Que, My ally Cayla, and [ Hello Barbie ] { }), Motherboard [ @motherboard2016gps ] (strangers could chase children ’s location and message them through the HereO smart watch), and BBC [ @bbcgermany2017cayla ] (German ban on Cayla in February 2017, brand it as an “ illegal espionage apparatus ”). Like wearables for adults, many smart toys are Bluetooth - enabled and accompany children outside the home, and previous research on fitness trackers has uncovered numerous vulnerabilities such as location tracking [ @Hilts2016fitness ].
Several security system breach involving the [ VTech InnoTab MAX ] { } have been report. In 2015, VTech ’s Learning Lodge database, use by the VTech suite of children ’s tablets, suffered a severe breach, exposing the PII of 6.3 million child ’s profiles [ @vtech2017breach ]. The breach was due to a combination of insecure practices, e.g., PII was sent in plaintext (no HTTPS), the database was vulnerable to a bare SQL injection attack, and account passwords were stored use an MD5 hashish (no salting). The rupture ultimately head to a \$650,000 settlement with the U.S. Federal Trade Commission for unauthorized collection of children ’s personal data and for failing to secure that information [ @ftc2018vtechsettlement ], and a related finding from Canada ’s Privacy Commissioner that VTech was in breach of its privacy laws (but no corresponding settlement) [ @pipeda2018vtech ]. besides in 2016, the UK - based security firm Pen Test Partners found that the [ VTech InnoTab MAX ] { } is vulnerable to trivial data origin [ @ptpvtech2015extraction ]. (We found that this flaw has not been fixed .) Later in 2016, it was find that | BB-8]{}, and [Toymail]{}; the rest haye not responded even ahter 6 mknths. Ovdrall, most responses (and lacn rhereif) indicate that these companied take peivary matters much more friyjlouapy tkai expected.
Relatgd work {#sec:ralated}
============
Smart elacgrlnics for children have made the ners for xefurity breachef nukqroua times in the past; e.g., see Rapid7 [@ralid72015iot] (vtrangers coukd hack baby monitors to vlew fhildren sleeping wnd even taok tj them directuy), Norwegian Consumer Douncil [@bouvet2016toys] (highlight daneeroux security vylnftabilities ii 3 toyf: i-Que, My Frlvnd Cayna, and [Nello Barbie]{}), Mpthxrboqrd [@motherboard2016gps] (strengers could track crildren’s noeation and message thwm throogh tve HdeeO smzrv wztch), ajd UBC [@bbcgermahy2017cayla] (Gernan ban on Cayla in Fqvruary 2017, brandjng it af an “illegal espionage apparatus”). Like wtarabmes for adults, many smaet toys are Bluetooth-gnabled anq accompany children outside the home, and previouv resxafch ok figbeds trackers has uncovered numerous vulnerabilyfits xuch as locatijn tracking [@Niptx2016sitness].
Severau secuxjtg breaches involvijg the [DTech InnoTab IAX]{} hsve been reported. In 2015, VTech’w Learning Ljege database, used yy the VTech suije of vhildren’s tablets, suffexed a aevere breafh, exposihe the PII of 6.3 miulipn children’s profiles [@vtech2017brqach]. The ureack was dud to a comfination ov inscwure practices, e.g., OII wcs sett in plaijtext (no HTTPS), the database was vulnerable to a sikple SQL injegtion attack, anq account passcords wexe stofed using zn MD5 hesh (no saltigg). The breach oltimately lev to a \$650,000 sqttlwmenr with gfe U.S. Federal Yrade Commission foe unauthorized colkecjikn of children’s petwonal informatoon anq voc faijhng to secura thxt kmformxtion [@ftc2018vteghsdttlrment], and a related xindjng from Canada’s Ptiyacy Commussioner that VTech wss in breach of itd pritacy lews (buy nj corresponding settlement) [@pipesa2018vtech]. Apso in 2016, the UK-bwsed security fixm Pen Test Partners found that the [VTeci InnoTab MAX]{} is vulneravle to trivial data txtraction [@ptpttech2015evtraction]. (We found that this dlaw has not been fixed.) Later in 2016, it wzs foutd thwt | BB-8]{}, and [Toymail]{}; the rest have not after months. Overall, responses (and lack take matters much more than expected. Related {#sec:related} ============ Smart electronics for children made the news for security breaches numerous times in the past; e.g., see [@rapid72015iot] (strangers could hack baby monitors to view children sleeping and even talk them Norwegian Council (highlight dangerous security vulnerabilities in 3 toys: i-Que, My Friend Cayla, and [Hello Barbie]{}), Motherboard [@motherboard2016gps] could track children’s location and message them through HereO smart watch), and [@bbcgermany2017cayla] (German ban on Cayla February branding it an espionage Like wearables for many smart toys are Bluetooth-enabled and accompany children outside the home, and previous research on fitness trackers uncovered numerous as location [@Hilts2016fitness]. security involving the [VTech have been reported. In 2015, VTech’s used by the VTech suite of children’s tablets, a severe exposing the PII of 6.3 million profiles [@vtech2017breach]. The breach was due to a of insecure practices, e.g., PII was sent in plaintext (no HTTPS), the database was vulnerable simple SQL injection attack, account passwords were using MD5 (no The breach led to a \$650,000 settlement with the U.S. Federal Trade Commission unauthorized collection of children’s personal information and for failing to information and a related from Canada’s Privacy Commissioner VTech in breach of its (but corresponding in the security firm Pen Test found that the [VTech InnoTab is vulnerable to trivial that this flaw has not been fixed.) Later 2016, it was found that | BB-8]{}, and [Toymail]{}; the rest have noT responded Even aFteR 6 moNtHs. OvEralL, most responses (ANd laCk thereof) indicate that tHese cOmPAnieS TaKe priVacy matTErS MUch MoRe FriVoLOuSly thAn eXpected.
related worK {#seC:rElated}
============
Smart eLEcTronics for ChiLdren have madE thE news fOr SecURity bReaChes nUmerouS Times iN the past; e.G., sEE Rapid7 [@RApid72015iot] (STRaNgerS could hack baby monIToRS to view childreN sleepInG AnD EVen TalK to them dirEcTly), NoRWegian CONsUMER CoUNcil [@bouvet2016toyS] (highlight dANgeRous seCuRitY VulnerAbiliTiES in 3 Toys: i-Que, My FRienD Cayla, and [hello BARbie]{}), MotHErboard [@MotherBoaRd2016gPs] (stRAnGeRs cOuLD trACk ChiLDreN’s locatiOn AnD messAge tHEM THrouGh tHe HeReO smArt watch), and BBc [@bbCgerMAny2017Cayla] (germaN ban On cayla In FebrUary 2017, bRaNding it as an “illeGal eSpionage aPpaRaTus”). liKe weaRAbles fOr aDulTs, many sMart toyS Are blUETOoTh-enabled and accompAnY CHiLdren outSide thE HoMe, ANd previoUs ResEarcH ON fitnEss tRAcKers has uNcoverED nUmErous vuLnErabilItIes SucH as loCAtioN trackIng [@Hilts2016FitneSS].
Several securiTY breaches invoLViNG ThE [vTecH InNoTab MAX]{} havE beeN RepoRted. iN 2015, VtecH’S LearNing LOdGE dATabase, used by the VTecH sUite of ChildRen’s tablets, suFfered a sevERE Breach, exPosiNG tHE PII of 6.3 million cHildrEn’s profileS [@Vtech2017breAch]. ThE breach wAs due to a cOMBination Of iNseCurE prACTiCes, e.g., PII was seNT In plAiNtext (no hTTpS), the daTabAse Was VulNeRable to a sImple SQL InJeCtIoN atTack, aND account PaSswOrDs wEre stORed usiNg an Md5 hasH (nO sALtiNg). The brEAcH ULtimAtElY led To a \$650,000 SeTtlemEnt wITh tHe U.S. FedEral Trade comMIssiOn FoR unauthOrized collectIoN of childreN’s PerSonal iNFOrmation And for failing to secure thAT informAtiOn [@ftc2018VtecHsettlemeNt], aNd a relAteD FindinG from CAnada’S PRivACY CommISSiOneR tHat VTech waS IN brEach oF iTs prIvacy laWs (but no correspondiNG seTtlement) [@pipedA2018vtEch]. ALSO iN 2016, thE uK-BAseD sECurITY firm Pen Test ParTners found ThAT tHe [VTech InnOtab mAx]{} is vulnErable tO trivIAl data eXtraction [@Ptpvtech2015eXtRactION]. (We Found that tHis flaw hAs not been FIxed.) LATeR in 2016, it Was Found tHaT | BB-8]{}, and [Toymail]{}; the resthavenot re sp onde d ev en after 6 mon t hs.Overall, most response s (an dl ackt he reof) indica t et h atth es e c om p an ies t ake privac y mattersmuc hmore frivolo u sl y than exp ect ed.
Related wo rk {#s ec :re l ated}
== ===== =====Smartelectroni cs for ch i ldren h a v emade the news for sec u ri t y breaches num erousti m es i n t hepast; e.g. ,see R a pid7 [@ r ap i d 7 201 5 iot] (strange rs could ha c k b aby mo ni tor s to vi ew ch il d ren sleeping a nd e ven talkto the m direct l y), Nor wegian Co nsu merC ou nc il[@ b ouv e t2 016 t oys ] (highl ig ht dang erou s s e curi tyvuln erabi lities in 3 t oys : i- Q ue, My F riend Cay la , and [Hell o Bar bi e]{}), Motherbo ard[@motherb oar d2 016 gp s] (s t ranger s c oul d track childr e n’s l o c a ti on and message the mt h ro ugh theHereOs ma rt watch),an d B BC [ @ b bcger many 2 01 7cayla](Germa n b an on Cay la in Fe br uar y 2 017,b rand ing it as an “ illeg a l espionage ap p aratus”). Lik e w e a ra b lesfor adults, ma ny s m arttoys ar e B l uetoo th-en ab l ed and accompany child re n outs ide t he home, andprevious r e s e arch onfitn e ss trackers has u ncove red numero u s vulner abili ties suc h as loca t i on track ing [@ Hil ts2 0 1 6f itness].
Sev e r al s ec urity b rea ches in vol vin g t he[V Tech Inno Tab MAX] {} h av ebee n rep o rted. In 2 015 ,VTe ch’sL earnin g Lod ge d at ab a se, used b y t h e VTe ch s uite of c hildr en’s tab lets, s uffered a se v erebr ea ch, exp osing the PII o f 6.3 mill io n c hildre n ’ s profil es [@vtech2017breach].T he brea chwas d ue t o a combi nat ion of in s ecurepracti ces,e. g., P II wa s se ntin plaintext ( noHTTPS ), the databa se was vulnerablet o a simple SQL i nje ctio n at tac k ,a ndac c oun t passwords werestored usi ng an MD5 hash( nosa lting). The br eachu ltimate ly led to a \$650, 00 0 se t t lem ent with t he U.S.Federal T r ade C o mm issio n f or una ut hor izedcollec t ion of c hildre n’ s pers onalin formatio n and for failing to se cure t hat i nfo rmation [ @ft c 201 8vtechset tlem ent], anda r ela ted f ind i ng fr om C a na da’ s Priv acyC ommission e rtha t VT ech was inb r e ach of i tsp rivacy law s (but no corresp o nding settleme nt)[ @ pip eda 2 018v te ch]. Also in 2 016 ,t h e UK-bas ed security f irm PenTe s t Par tnersfoundthat th e [V T ech In noTa b M AX]{} isvul ne r able to t ri v ial da ta e xt ractio n [@pt p vtec h 2 015extraction].(We f o u nd th a t t his f la w has n o t be en fixed.) Later in 2 016, i t wa s fou nd that | BB-8]{},_and [Toymail]{};_the rest have not_responded even_after_6 months._Overall,_most responses (and_lack thereof) indicate_that these companies take_privacy matters much_more_frivolously than expected.
Related work {#sec:related}
============
Smart electronics for children have made the news for security_breaches_numerous times_in_the_past; e.g., see Rapid7 [@rapid72015iot] (strangers_could hack baby monitors to_view children_sleeping and even talk to them directly), Norwegian_Consumer_Council [@bouvet2016toys] (highlight dangerous_security vulnerabilities in 3 toys: i-Que, My Friend Cayla,_and [Hello Barbie]{}), Motherboard [@motherboard2016gps] (strangers could_track children’s location_and_message_them through the HereO_smart watch), and BBC [@bbcgermany2017cayla] (German ban_on Cayla in February 2017, branding_it as an “illegal espionage apparatus”). Like_wearables for adults, many smart toys_are Bluetooth-enabled and accompany children_outside the_home, and previous research on_fitness trackers has_uncovered numerous_vulnerabilities such as_location tracking [@Hilts2016fitness].
Several security breaches involving the_[VTech InnoTab MAX]{} have_been reported. In 2015, VTech’s Learning_Lodge_database, used by_the_VTech_suite of_children’s tablets, suffered_a_severe breach,_exposing_the PII of 6.3 million children’s_profiles [@vtech2017breach]._The breach was due to a combination_of insecure practices, e.g.,_PII_was sent in plaintext_(no HTTPS), the database was_vulnerable to a simple SQL injection_attack, and_account passwords_were stored using an MD5 hash (no salting). The breach ultimately_led to a \$650,000 settlement with_the U.S. Federal Trade_Commission for_unauthorized_collection of children’s_personal_information and_for failing to secure that information [@ftc2018vtechsettlement], and_a related_finding from Canada’s Privacy Commissioner that_VTech was in breach_of_its privacy laws (but no corresponding_settlement) [@pipeda2018vtech]. Also in 2016, the UK-based_security firm Pen Test Partners_found_that_the [VTech InnoTab MAX]{} is vulnerable_to trivial data extraction [@ptpvtech2015extraction]. (We found_that this flaw_has not been fixed.) Later in 2016,_it_was found that |
zero temperature because thermal fluctuations can destroy long-range order in two-dimensions (the so-called, Hohenberg-Mermin-Wagner theorem [@MW66]). At a finite temperature $T$, topological defects such as dislocations are always present. Furthermore, because of the particular structure of the honeycomb lattice, the dynamics of lattice defects in graphene planes belong to the generic class of kinetically constrained models[@DS00; @RS03], where defects are never completely annealed since their number decreases only as a logarithmic function of the annealing time [@DS00]. Indeed, defects are ubiquitous in carbon allotropes with sp$^2$ coordination and have been observed in these systems [@Hetal04b]. As a consequence of the presence of topological defects, the electronic properties discussed previously, are significantly modified leading to qualitatively new physics. As we show below, extended defects can lead to the phenomenon of self-doping with the formation of electron or hole pockets close to the Dirac points. We show, however, that the presence of such defects can still lead to long electronic mean free paths. We present next an analysis of the physical properties of graphene as a function of the density of defects, at zero and finite temperature, frequency, and magnetic field. The defects analyzed here, like boundaries (edges), dislocations, vacancies, can be considered strong distortions of the perfect system. In this respect, our work complements the studies of defects and interactions in systems described by the two-dimensional Dirac equation [@r1].
The role of disorder on the electronic properties of coupled graphene planes shows also its importance on the unexpected appearance of ferromagnetism in proton irradiated graphite [@Ketal00; @Eetal02; @MHM02; @Ketal03b; @Eetal03b; @MP05]. In a recent publication, the role of the exchange mechanism on a disordered graphene plane was addressed [@PGN05b]. It was found that disorder can stabilizes a ferromagnetic phase in the presence of long-range Coulomb interactions. On the other hand, the effect of disorder on the density of states of a single graphene plane amounts to the creation of a finite density of states at zero energy. Therefore, a certain amount of screening should be present and the question of whether the interplay of disorder and short-range Coulomb interaction may stabilize a ferromagnetic ground state has to be addressed as well.
Moreover, with the current experimental techniques, it is possible to study not only a single layer of graphene but also graphene | zero temperature because thermal fluctuations can destroy long - range decree in two - dimension (the so - called, Hohenberg - Mermin - Wagner theorem [ @MW66 ]). At a finite temperature $ T$, topological defects such as dislocation are always present. Furthermore, because of the particular social organization of the honeycomb lattice, the dynamics of lattice defects in graphene airplane belong to the generic class of kinetically stiffen models[@DS00; @RS03 ], where defects are never completely anneal since their number decreases only as a logarithmic function of the annealing time [ @DS00 ]. Indeed, defects are ubiquitous in carbon paper allotropes with sp$^2 $ coordination and have been observed in these systems [ @Hetal04b ]. As a consequence of the bearing of topological defects, the electronic properties discussed previously, are significantly modified lead to qualitatively new physics. As we show below, extended defects can run to the phenomenon of self - doping with the formation of electron or hole pockets close to the Dirac points. We usher, however, that the presence of such defects can still lead to long electronic mean free paths. We present next an analysis of the forcible property of graphene as a function of the concentration of defect, at zero and finite temperature, frequency, and charismatic field. The defects analyzed here, like boundaries (edges), dislocations, vacancies, can be considered firm distortions of the perfect system. In this respect, our work complement the studies of defects and interactions in systems described by the two - dimensional Dirac equation [ @r1 ].
The role of disorderliness on the electronic property of coupled graphene planes shows also its importance on the unexpected appearance of ferromagnetism in proton irradiated graphite [ @Ketal00; @Eetal02; @MHM02; @Ketal03b; @Eetal03b; @MP05 ]. In a recent publication, the function of the exchange mechanism on a disordered graphene plane was addressed [ @PGN05b ]. It was discover that disorderliness can stabilizes a ferromagnetic phase in the presence of long - range Coulomb interactions. On the early hand, the effect of disorder on the density of department of state of a single graphene plane amounts to the creation of a finite density of department of state at zero energy. Therefore, a certain amount of screening should be present and the question of whether the interplay of disorder and short - range Coulomb interaction may brace a ferromagnetic ground state get to be addressed as well.
Moreover, with the current experimental techniques, it is possible to study not only a individual layer of graphene but also graphene | zego temperature because tmermal fluctuations can destrky long-rxnge order in two-dimensions (vhe wo-caloed, Hohenberg-Mermin-Wagver theorvm [@MW66]). At q fiiite temperature $T$, topological fefeetw such as dislpcations ase always presang. Yurthermore, because of the particulaw strucyuge of the honeicomb jattjbe, the dynamics of lattice defedts in jraphene planes belong to the generic clads ov kinetically conshrained modgms[@DF00; @RS03], where dewects are never complejely annealed since their number decrzases only qs a pmgarithmic hunctijn of the ankvaling dime [@DS00]. Indeed, defectx ace uviquitous in carbon anlotropes with sp$^2$ soordinathou and have been obsercee in jhese sysgwms [@Heual04u]. Aa a cojsesuence of fhe presencw of topological degesnx, the electrknic pwo[erties discussed previously, are signifpcanfly modified leading to qualitatively new phisics. As wq show below, extended defects can lead to the phetomenkv oy self-aiplng with the formation of electron or hole posieus blose to the Dirag points. We show, hpwfvrt, that the prerence of auch defects can shill lewd to long elestromic mean free paths. We preswnt next an cnaoysis of the physieal propertizs of braphrne as a function of thz densjty of defefts, at zedu and finite temoerstgre, frequency, and magnetic field. Thx defzcts anauyzec here, like bounfaries (edges), dislocationd, vaccnciev, can be clnsidered strong distortions of the perfect sistam. Pn this rzspect, our work com[lements the sjudies of defezts and inneractionv in systemf described bf the two-dimeisional Dyrac equqtion [@r1].
Gfe role of disprder on nht electronuc properties of cpupusd graphene plauts whows also its ompurtwnbe pn dhe unexpectad aopextance of ferromannegism in proton irradiateg grzphite [@Ketal00; @Eetal02; @MMM02; @Ketal03b; @Eetal03b; @IP05]. In a receny publication, the gole mf vhe exvhagge mechanism on a disordered fraphene olake was addresfed [@IGN05b]. It was fobnd that disorder can stabilizes a ferronagnetic phase in rhe presence of lony-rsnge Coulomu intewactions. Mn the other hand, thw effect of disorcer on the density of statev of w single graphene plane amounts to the creation of a finite density of stqtes av sero energy. Theteforv, a eertain amonnv of screening shpuld be present and the questioi of whethar the interplay of disorder amd short-range Cuulomb interaction may sfabilize a ferromagnetic ground state hss to be addressed as wekl.
Moreovwr, wifh the curcent experimental tevhniquzs, it us possuble to study not onky a single laytr oy graphene but also graphene | zero temperature because thermal fluctuations can destroy in (the so-called, theorem [@MW66]). At defects as dislocations are present. Furthermore, because the particular structure of the honeycomb the dynamics of lattice defects in graphene planes belong to the generic class kinetically constrained models[@DS00; @RS03], where defects are never completely annealed since their number only a function the annealing time [@DS00]. Indeed, defects are ubiquitous in carbon allotropes with sp$^2$ coordination and have observed in these systems [@Hetal04b]. As a consequence the presence of topological the electronic properties discussed previously, significantly leading to new As show below, extended can lead to the phenomenon of self-doping with the formation of electron or hole pockets close to Dirac points. however, that presence such can still lead electronic mean free paths. We present of the physical properties of graphene as a of the of defects, at zero and finite frequency, and magnetic field. The defects analyzed here, boundaries (edges), dislocations, vacancies, can be considered strong distortions of the perfect system. In this work complements the studies defects and interactions systems by two-dimensional equation [@r1]. role of disorder on the electronic properties of coupled graphene planes also its importance on the unexpected appearance of ferromagnetism in graphite @Eetal02; @MHM02; @Ketal03b; @MP05]. In a recent the of the exchange mechanism disordered plane It found disorder can stabilizes a phase in the presence of Coulomb interactions. On the disorder on the density of states of a graphene plane amounts to the creation of finite density of states at zero energy. Therefore, a certain amount of should be the question of whether the interplay of disorder short-range Coulomb interaction may a ferromagnetic ground state has to be addressed as Moreover, the current techniques, it is to study not a single layer but also | zero temperature because theRmal fluctuAtionS caN deStRoy lOng-rAnge order in two-DImenSions (the so-called, HohenbErg-MeRmIN-WagNEr TheorEm [@MW66]). At a FInITE teMpErAtuRe $t$, ToPologIcaL defectS such as disLocAtIons are alwayS PrEsent. FurthErmOre, because of The ParticUlAr sTRuctuRe oF the hOneycoMB lattiCe, the dynaMiCS of latTIce defeCTS iN graPhene planes belong TO tHE generic class oF kinetIcALlY COnsTraIned models[@dS00; @rS03], wheRE defectS ArE NEVer COmpletely anneAled since thEIr nUmber dEcReaSEs only As a loGaRIthMic function Of thE annealinG time [@Ds00]. indeed, dEFects arE ubiquItoUs iN carBOn AlLotRoPEs wITh Sp$^2$ cOOrdInation aNd HaVe beeN obsERVED in tHesE sysTems [@HEtal04b]. As a conseQueNce oF The PreseNce of TopoLoGical DefectS, the eLeCtronic propertiEs diScussed prEviOuSly, ArE signIFicantLy mOdiFied leaDing to qUAliTaTIVElY new physics. As we shoW bELOw, Extended DefectS CaN lEAd to the pHeNomEnon OF Self-dOpinG WiTh the forMation OF eLeCtron or HoLe pockEtS clOse To the dIrac Points. we show, hoWever, THat the presence OF such defects cAN sTILl LEad tO loNg electroniC meaN Free PathS. we PreSEnt neXt an aNaLYsIS of the physical propeRtIes of gRapheNe as a function Of the densiTY OF defects, At zeRO aND finite temperaTure, fRequency, anD Magnetic Field. the defecTs analyzeD HEre, like bOunDarIes (EdgES), DiSlocations, vacANCies, CaN be consIdeRed stroNg dIstOrtIonS oF the perfeCt system. in ThIs ReSpeCt, our WOrk complEmEntS tHe sTudieS Of defeCts anD intErAcTIonS in systEMs DEScriBeD bY the Two-DiMensiOnal dIraC equatiOn [@r1].
The rolE of DIsorDeR oN the eleCtronic properTiEs of coupleD gRapHene plANEs shows aLso its importance on the unEXpected AppEaranCe of FerromagnEtiSm in prOtoN IrradiAted grAphitE [@KEtaL00; @eEtal02; @Mhm02; @keTal03B; @EEtal03b; @MP05]. In a RECenT publIcAtioN, the rolE of the exchange mechANisM on a disordereD grApheNE PlAne WAs ADdrEsSEd [@Pgn05B]. It was found that Disorder caN sTAbIlizes a ferROmaGnEtic phaSe in the PreseNCe of lonG-range CouLomb interAcTionS. oN thE other hand, The effecT of disordER on thE DeNsity Of sTates oF a SinGle grAphene PLanE amouNts to tHe CreatiOn of a FiNite densIty of states at zero energy. therefOre, a cErtAin amount Of sCReeNing shoulD be pResent and tHe qUesTion oF whETher tHe inTErPlaY Of disOrdeR And short-rANgE CoULOmB interactioN MAY stAbiliZe a FErromaGnetIc ground state has tO Be addressed as wEll.
MOREovEr, wITh thE cUrrent experimeNtaL tECHniques, iT iS possible to Study not OnLY a sinGle layEr of grAphene bUT AlSO graphEne | zero temperature becausethermal fl uctua tio nsca n de stro y long-range o r derin two-dimensions (the so-c al l ed,H oh enber g-Mermi n -W a g ner t he ore m[ @M W66]) . A t a fin ite temper atu re $T$, topolo g ic al defects su ch as disloc ati ons ar ealw a ys pr ese nt. F urther m ore, b ecause of t h e part i cular s t r uc ture of the honeycomb la t tice, the dyna mics o fl at t i cedef ects in gr ap henep lanes b e lo n g tot he generic cl ass of kine t ica lly co ns tra i ned mo dels[ @D S 00; @RS03], wh eredefects a re nev e r compl e tely an nealed si nce the i rnu mbe rd ecr e as eso nly as a lo ga ri thmic fun c t i o n of th e an neali ng time [@DS0 0]. Ind e ed, defe cts a re u bi quito us incarbo nallotropes with sp$ ^2$ coord ina ti onan d hav e beenobs erv ed in t hese sy s tem s[ @ H et al04b]. As a conse qu e n ce of thepresen c eof topologi ca l d efec t s , the ele c tr onic pro pertie s d is cussedpr evious ly , a resigni f ican tly mo dified l eadin g to qualitativ e ly new physic s .A s w e sho w b elow, exten dedd efec ts c a nlea d to t he ph en o me n on of self-doping w it h theforma tion of elect ron or hol e p ockets c lose to the Dirac poin ts. W e show, ho w ever, th at th e presen ce of suc h defectscan st ill le a d t o long electr o n ic m ea n freepat hs. Wepre sen t n ext a n analysi s of the p hy si ca l p roper t ies of g ra phe ne as a fu n ctionof th e de ns it y of defect s ,a t zer oan d fi nit etempe ratu r e,frequen cy, and m agn e ticfi el d. Thedefects analy ze d here, li ke bo undari e s (edges) , dislocations, vacanci e s, canbeconsi dere d strongdis tortio nso f theperfec t sys te m.I n this r es pec t, our workc o mpl ement sthestudies of defects and in t era ctions in sys tem s de s c ri bed by the t w o-d i m ensional Diracequation [ @r 1 ].
The role ofdi sorderon theelect r onic pr opertiesof couple dgrap h e neplanes sho ws alsoits impor t anceo nthe u nex pected a ppe aranc e of f e rro magne tism i nproton irra di ated gra phite [@Ketal00; @Eetal 02; @M HM02; @K etal03b;@Ee t al0 3b; @MP05 ]. I n a recent pu bli catio n,t he ro le o f t hee xchan ge m e chanism o n a di s o rd ered graphe n e pla ne wa s a d dresse d [@ PGN05b]. It was f o und that disor derc a n s tab i lize sa ferromagneti c p ha s e in thepr esence of l ong-rang eC oulom b inte ractio ns. Ont h eo ther h and, th e effectofdi s order o nth e densi ty o fstates of as ingl e graphene plane a mount s to th e cr eatio nof a fi n itedensity of states atzero e nerg y. Th erefore ,a cert ain a mount of s c reening s hould be pre se nt a ndthe qu esti o n of w heth er th e interpl a y o f d is o rde r an d sho rt -ran ge Coulom b interac tio n may st ab ili z e a fer r om a g netic grou ndstate h as to be a d dres s ed as we ll.
M oreove r, with the c urrentexp e r imental t echniques , it i s po ssible t ostud ynot o nly a single la yer of g rapheneb u t also gr a phen e | zero_temperature because_thermal fluctuations can destroy_long-range order_in_two-dimensions (the_so-called,_Hohenberg-Mermin-Wagner theorem [@MW66])._At a finite_temperature $T$, topological defects_such as dislocations_are_always present. Furthermore, because of the particular structure of the honeycomb lattice, the dynamics_of_lattice defects_in_graphene_planes belong to the generic_class of kinetically constrained models[@DS00;_@RS03], where_defects are never completely annealed since their number_decreases_only as a_logarithmic function of the annealing time [@DS00]. Indeed, defects_are ubiquitous in carbon allotropes with_sp$^2$ coordination and_have_been_observed in these systems_[@Hetal04b]. As a consequence of the_presence of topological defects, the electronic_properties discussed previously, are significantly modified leading_to qualitatively new physics. As we_show below, extended defects can_lead to_the phenomenon of self-doping with_the formation of_electron or_hole pockets close_to the Dirac points. We show,_however, that the_presence of such defects can still_lead_to long electronic_mean_free_paths. We_present next an_analysis_of the_physical_properties of graphene as a function_of_the density of defects, at zero and_finite temperature, frequency, and_magnetic_field. The defects analyzed_here, like boundaries (edges), dislocations,_vacancies, can be considered strong distortions_of the_perfect system._In this respect, our work complements the studies of defects and_interactions in systems described by the_two-dimensional Dirac equation [@r1].
The_role of_disorder_on the electronic_properties_of coupled_graphene planes shows also its importance on_the unexpected_appearance of ferromagnetism in proton irradiated_graphite [@Ketal00; @Eetal02; @MHM02;_@Ketal03b;_@Eetal03b; @MP05]. In a recent publication,_the role of the exchange mechanism_on a disordered graphene plane_was_addressed_[@PGN05b]. It was found that_disorder can stabilizes a ferromagnetic phase_in the presence_of long-range Coulomb interactions. On the other_hand,_the effect of disorder on the_density_of states of a single graphene_plane_amounts_to the creation of a_finite density of states at zero_energy. Therefore, a certain amount of screening should be_present and the_question of whether the interplay_of_disorder_and short-range Coulomb interaction may stabilize a ferromagnetic ground state_has to_be addressed as_well.
Moreover, with the current experimental techniques, it is possible to_study not only a single layer of_graphene but also graphene |
$-\infty<\lambda <\alpha$.
\[lem\_hm\] Assume $k$ i.i.d. samples $x_j \sim S(\alpha<1, \beta=1, F_{(\alpha)})$, define the harmonic mean estimator $\hat{F}_{(\alpha),hm}$, [$$\begin{aligned}
\notag\vspace{-0.05in}
\hat{F}_{(\alpha),hm} = \frac{k\frac{\cos\left(\frac{\alpha\pi}{2}\right)}{\Gamma(1+\alpha)}}{\sum_{j=1}^k|x_j|^{-\alpha}},\end{aligned}$$]{} and the bias-corrected harmonic mean estimator $\hat{F}_{(\alpha),hm,c}$, [$$\begin{aligned}
\notag
\hat{F}_{(\alpha),hm,c} = \frac{k\frac{\cos\left(\frac{\alpha\pi}{2}\right)}{\Gamma(1+\alpha)}}{\sum_{j=1}^k|x_j|^{-\alpha}}
\left(1- \frac{1}{k}\left(\frac{2\Gamma^2(1+\alpha)}{\Gamma(1+2\alpha)}-1\right) \right).\end{aligned}$$]{} The bias and variance of $\hat{F}_{(\alpha),hm,c}$ are [$$\begin{aligned}
\notag
&\text{E}\left(\hat{F}_{(\alpha),hm,c}\right) = F_{(\alpha)}+O\left(\frac{1}{k^2}\right),\\\notag
&\text{Var}\left(\hat{F}_{(\alpha),hm,c}\right) = \frac{F^{2}_{(\alpha)}}{k}\left(\frac{2\Gamma^2(1+\alpha)}{\Gamma(1+2\alpha)}-1\right) + O\left(\frac{1}{k^2}\right).\end{aligned}$$]{}
The right tail bound of $\hat{F}_{(\alpha),hm}$ is, for $\epsilon>0$, [$$\begin{aligned}
\notag
&\mathbf{Pr}\left(
\hat{F}_{(\alpha),hm} - F_{(\alpha)} \geq \epsilon F_{(\alpha)}\right)
\leq\exp\left(-k\left(\frac{\epsilon^2}{G_{ | $ -\infty<\lambda < \alpha$.
\[lem\_hm\ ] Assume $ k$ i.i.d. samples $ x_j \sim S(\alpha<1, \beta=1, F_{(\alpha)})$, define the harmonic mean calculator $ \hat{F}_{(\alpha),hm}$, [ $ $ \begin{aligned }
\notag\vspace{-0.05 in }
\hat{F}_{(\alpha),hm } = \frac{k\frac{\cos\left(\frac{\alpha\pi}{2}\right)}{\Gamma(1+\alpha)}}{\sum_{j=1}^k|x_j|^{-\alpha}},\end{aligned}$$ ] { } and the diagonal - correct harmonic mean calculator $ \hat{F}_{(\alpha),hm, c}$, [ $ $ \begin{aligned }
\notag
\hat{F}_{(\alpha),hm, c } = \frac{k\frac{\cos\left(\frac{\alpha\pi}{2}\right)}{\Gamma(1+\alpha)}}{\sum_{j=1}^k|x_j|^{-\alpha } }
\left(1- \frac{1}{k}\left(\frac{2\Gamma^2(1+\alpha)}{\Gamma(1 + 2\alpha)}-1\right) \right).\end{aligned}$$ ] { } The diagonal and variance of $ \hat{F}_{(\alpha),hm, c}$ are [ $ $ \begin{aligned }
\notag
& \text{E}\left(\hat{F}_{(\alpha),hm, c}\right) = F_{(\alpha)}+O\left(\frac{1}{k^2}\right),\\\notag
& \text{Var}\left(\hat{F}_{(\alpha),hm, c}\right) = \frac{F^{2}_{(\alpha)}}{k}\left(\frac{2\Gamma^2(1+\alpha)}{\Gamma(1 + 2\alpha)}-1\right) + O\left(\frac{1}{k^2}\right).\end{aligned}$$ ] { }
The right fag end bound of $ \hat{F}_{(\alpha),hm}$ is, for $ \epsilon>0 $, [ $ $ \begin{aligned }
\notag
& \mathbf{Pr}\left (
\hat{F}_{(\alpha),hm } - F_{(\alpha) } \geq \epsilon F_{(\alpha)}\right)
\leq\exp\left(-k\left(\frac{\epsilon^2}{G _ { | $-\invty<\lambda <\alpha$.
\[lem\_hm\] Assmme $k$ i.i.d. samples $x_j \sik S(\alpga<1, \beta=1, W_{(\alpha)})$, define the harmonic mxan wstimqtor $\hat{F}_{(\alpha),hm}$, [$$\begin{xligned}
\nonag\vspace{-0.05un}
\hau{F}_{(\alpha),hm} = \frac{k\fczc{\cos\lenc(\frac{\zppha\'i}{2}\cight)}{\Gamma(1+\alpha)}}{\xum_{j=1}^k|x_j|^{-\alpva}},\end{aligned}$$]{} atd tke bias-corrected harmonic mean estimwtor $\hay{F}_{(\wlpha),hm,c}$, [$$\begin{wligmqd}
\nofag
\hat{F}_{(\alpha),hm,c} = \frac{k\frac{\cos\left(\fdac{\alphe\pi}{2}\right)}{\Gamma(1+\allha)}}{\sum_{j=1}^k|x_j|^{-\alpha}}
\left(1- \frac{1}{k}\lfft(\fgac{2\Gamma^2(1+\alpha)}{\Gamma(1+2\wlpha)}-1\right) \tjghe).\wnd{aligned}$$]{} Tfe bias and variance or $\hat{F}_{(\alpha),hm,c}$ are [$$\begin{aligned}
\votag
&\cext{E}\left(\haj{Y}_{(\qlpjd),hm,c}\right) = H_{(\alpha)}+J\left(\frac{1}{k^2}\rinnt),\\\notac
&\text{Vat}\left(\hat{F}_{(\alpha),mm,c}\rijht) = \frac{F^{2}_{(\alpha)}}{k}\left(\frac{2\Gemma^2(1+\alpha)}{\Gamma(1+2\alpha)}-1\ryght) + O\lext(\yrac{1}{k^2}\right).\end{aligned}$$]{}
Tye righj tain bojbd uf $\gav{F}_{(\ampha),hm}$ is, for $\epsilkn>0$, [$$\begin{alitned}
\notag
&\mathbf{Pr}\legt(
\rqt{F}_{(\alpha),hm} - R_{(\alpha)} \gqq \epsilon F_{(\alpha)}\right)
\leq\exp\left(-k\left(\frdc{\elsilon^2}{G_{ | $-\infty<\lambda <\alpha$. \[lem\_hm\] Assume $k$ i.i.d. samples S(\alpha<1, F_{(\alpha)})$, define harmonic mean estimator \frac{k\frac{\cos\left(\frac{\alpha\pi}{2}\right)}{\Gamma(1+\alpha)}}{\sum_{j=1}^k|x_j|^{-\alpha}},\end{aligned}$$]{} the bias-corrected harmonic estimator $\hat{F}_{(\alpha),hm,c}$, [$$\begin{aligned} \hat{F}_{(\alpha),hm,c} = \frac{k\frac{\cos\left(\frac{\alpha\pi}{2}\right)}{\Gamma(1+\alpha)}}{\sum_{j=1}^k|x_j|^{-\alpha}} \left(1- \frac{1}{k}\left(\frac{2\Gamma^2(1+\alpha)}{\Gamma(1+2\alpha)}-1\right) \right).\end{aligned}$$]{} bias and variance of $\hat{F}_{(\alpha),hm,c}$ are [$$\begin{aligned} \notag &\text{E}\left(\hat{F}_{(\alpha),hm,c}\right) = F_{(\alpha)}+O\left(\frac{1}{k^2}\right),\\\notag &\text{Var}\left(\hat{F}_{(\alpha),hm,c}\right) = + O\left(\frac{1}{k^2}\right).\end{aligned}$$]{} The right tail bound of $\hat{F}_{(\alpha),hm}$ is, for $\epsilon>0$, [$$\begin{aligned} \notag \hat{F}_{(\alpha),hm} F_{(\alpha)} \epsilon \leq\exp\left(-k\left(\frac{\epsilon^2}{G_{ | $-\infty<\lambda <\alpha$.
\[lem\_hm\] AssuMe $k$ i.i.d. sampLes $x_j \Sim s(\alPhA<1, \betA=1, F_{(\alPha)})$, define the haRMoniC mean estimator $\hat{F}_{(\alphA),hm}$, [$$\beGiN{AligNEd}
\Notag\Vspace{-0.05iN}
\HaT{f}_{(\AlpHa),Hm} = \FraC{k\FRaC{\cos\lEft(\Frac{\alpHa\pi}{2}\right)}{\GAmmA(1+\aLpha)}}{\sum_{j=1}^k|x_j|^{-\aLPhA}},\end{aligneD}$$]{} anD the bias-corrEctEd harmOnIc mEAn estImaTor $\haT{F}_{(\alphA),Hm,c}$, [$$\begIn{aligned}
\NoTAg
\hat{F}_{(\ALpha),hm,c} = \FRAc{K\fraC{\cos\left(\frac{\alpha\PI}{2}\rIGht)}{\Gamma(1+\alpha)}}{\sUm_{j=1}^k|x_j|^{-\AlPHa}}
\LEFt(1- \fRac{1}{K}\left(\frac{2\GAmMa^2(1+\alpHA)}{\Gamma(1+2\aLPhA)}-1\RIGht) \RIght).\end{aligneD}$$]{} The bias and VAriAnce of $\HaT{F}_{(\aLPha),hm,c}$ Are [$$\beGiN{AliGned}
\notag
&\teXt{E}\lEft(\hat{F}_{(\alPha),hm,c}\RIght) = F_{(\alPHa)}+O\left(\Frac{1}{k^2}\rIghT),\\\noTag
&\tEXt{vaR}\leFt(\HAt{F}_{(\ALpHa),hM,C}\riGht) = \frac{F^{2}_{(\AlPhA)}}{k}\lefT(\fraC{2\gAMMa^2(1+\alPha)}{\gammA(1+2\alphA)}-1\right) + O\left(\frAc{1}{k^2}\RighT).\End{AlignEd}$$]{}
The RighT tAil boUnd of $\hAt{F}_{(\alPhA),hm}$ is, for $\epsilon>0$, [$$\BegiN{aligned}
\nOtaG
&\mAthBf{pr}\lefT(
\Hat{F}_{(\alPha),Hm} - F_{(\Alpha)} \geQ \epsiloN f_{(\alPhA)}\RIGhT)
\leq\exp\left(-k\left(\frAc{\EPSiLon^2}{G_{ | $-\infty<\lambda <\alpha$ .
\[lem\_ hm\]Ass ume $ k$ i .i.d . samples $x_j \sim S(\alpha<1, \beta=1,F_{(\ al p ha)} ) $, defi ne theh ar m o nic m ea n e st i ma tor $ \ha t{F}_{( \alpha),hm }$, [ $$\begin{ali g ne d}
\notag\ vsp ace{-0.05in}
\h at{F}_ {( \al p ha),h m}= \fr ac{k\f r ac{\co s\left(\f ra c {\alph a \pi}{2} \ r ig ht)} {\Gamma(1+\alpha) } }{ \ sum_{j=1}^k|x_ j|^{-\ al p ha } } ,\e nd{ aligned}$$ ]{ } and the bia s -c o r r ect e d harmonic me an estimato r $\ hat{F} _{ (\a l pha),h m,c}$ ,[ $$\ begin{align ed}\notag
\h at{F}_ { (\alpha ) ,hm,c}= \fra c{k \fr ac{\ c os \l eft (\ f rac { \a lph a \pi }{2}\rig ht )} {\Gam ma(1 + \ a l pha) }}{ \sum _{j=1 }^k|x_j|^{-\a lph a}}\ lef t(1-\frac {1}{ k} \left (\frac {2\Ga mm a^2(1+\alpha)}{ \Gam ma(1+2\al pha )} -1\ ri ght)\ right) .\e nd{ aligned }$$]{}T hebi a s an d variance of $\ha t{ F } _{ (\alpha) ,hm,c} $ a re [$$\begi n{ ali gned } \nota g
&\ t ex t{E}\lef t(\hat { F} _{ (\alpha ), hm,c}\ ri ght ) = F_{( \ alph a)}+O\ left(\fr ac{1} { k^2}\right),\\ \ notag
&\text{ V ar } \ le f t(\h at{ F}_{(\alpha ),hm , c}\r ight ) = \f r ac{F^ {2}_{ (\ a lp h a)}}{k}\left(\frac{ 2\ Gamma^ 2(1+\ alpha)}{\Gamm a(1+2\alph a ) } -1\right ) +O \l e ft(\frac{1}{k^ 2}\ri ght).\end{ a ligned}$ $]{}
The rig ht tail b o u nd of $\ hat {F} _{( \al p h a) ,hm}$ is, for $ \eps il on>0$,[$$ \begin{ ali gne d}\no ta g
&\mathb f{Pr}\le ft (\h at {F} _{(\a l pha),hm} - F _{ (\a lpha) } \geq\epsi lonF_ {( \ alp ha)}\ri g ht ) \leq \e xp \lef t(- k\ left( \fra c {\e psilon^ 2}{G_{ | $-\infty<\lambda_<\alpha$.
\[lem\_hm\] Assume_$k$ i.i.d. samples $x_j_\sim S(\alpha<1,_\beta=1,_F_{(\alpha)})$, define_the_harmonic mean estimator_$\hat{F}_{(\alpha),hm}$, [$$\begin{aligned}
\notag\vspace{-0.05in}
\hat{F}_{(\alpha),hm} =_\frac{k\frac{\cos\left(\frac{\alpha\pi}{2}\right)}{\Gamma(1+\alpha)}}{\sum_{j=1}^k|x_j|^{-\alpha}},\end{aligned}$$]{} and the bias-corrected_harmonic mean estimator_$\hat{F}_{(\alpha),hm,c}$,_[$$\begin{aligned}
\notag
\hat{F}_{(\alpha),hm,c} = \frac{k\frac{\cos\left(\frac{\alpha\pi}{2}\right)}{\Gamma(1+\alpha)}}{\sum_{j=1}^k|x_j|^{-\alpha}}
\left(1- \frac{1}{k}\left(\frac{2\Gamma^2(1+\alpha)}{\Gamma(1+2\alpha)}-1\right) \right).\end{aligned}$$]{} The bias and variance of $\hat{F}_{(\alpha),hm,c}$ are [$$\begin{aligned}
\notag
&\text{E}\left(\hat{F}_{(\alpha),hm,c}\right) =_F_{(\alpha)}+O\left(\frac{1}{k^2}\right),\\\notag
&\text{Var}\left(\hat{F}_{(\alpha),hm,c}\right)_= \frac{F^{2}_{(\alpha)}}{k}\left(\frac{2\Gamma^2(1+\alpha)}{\Gamma(1+2\alpha)}-1\right)_+_O\left(\frac{1}{k^2}\right).\end{aligned}$$]{}
The_right tail bound of $\hat{F}_{(\alpha),hm}$_is, for $\epsilon>0$, [$$\begin{aligned}
\notag
&\mathbf{Pr}\left(
\hat{F}_{(\alpha),hm} -_ F_{(\alpha)}_\geq \epsilon F_{(\alpha)}\right)
\leq\exp\left(-k\left(\frac{\epsilon^2}{G_{ |
range interactions may considerably increase the rectification power and may avoid its decay with the system size, problems of the usual proposals of rectifiers. In particular, we show that such phenomenon occurs in graded systems, realizable materials in which thermal rectification ubiquitously holds. In short, our results indicate that graded materials are genuine candidates for the actual fabrication of thermal diodes.
We are in debt to an anonymous referee, who called our attention to the connection between some of our results and the Kirchhoff’s theorem for circuits [@Referee]. We thank R. Sardenberg, M. Matos, M. C. Aguiar and B. V. Costa for the help with numerical programs. This work was partially supported by CNPq (Brazil).
[32]{}
N. Li, J. Ren, L. Wang, G. Zhang, P. Hänggi, and B. Li, Rev. Mod. Phys. [**84**]{}, 1045 (2012); G. Casati, Nature Nanotech. [**2**]{}, 23 (2007); B. Hu, in [*Advanced Workshop on Energy Transport in Low-Dimensional System: Achievements and Mysteries*]{}, ICTP (2012), \[<http://cdsagenda5.ictp.trieste.it/full_display.php?ida=a11200>\].
M. Terraneo, M. Peyrard, and G. Casati, Phys. Rev. Lett. [**88**]{}, 094302 (2002).
B. Li, L. Wang, and G. Casati, Phys. Rev. Lett. [**93**]{}, 184301 (2004); L. Wang and B. Li, [*ibid.*]{} [**99**]{}, 177208 (2007); L. Wang and B. Li, [*ibid.*]{} [**101**]{}, 267203 (2008); B. Li [*et al.*]{}, Appl. Phys. Lett. [**88**]{}, 143501 (2006).
B. Hu, L. Yang, and Y. Zhang, Phys. Rev. Lett. [**97**]{}, 124302 (2006).
E. Pereira, Phys. Rev. E [**82**]{}, 040101 (R) (2010); E. Pereira, Physica A [**390**]{}, 4131 (2011).
C.W. Chang, D. Okawa, A. Majumdar, and A. Zettl, Science [**314**]{}, 1121 | range interactions may considerably increase the rectification exponent and may invalidate its decay with the system size, problems of the usual marriage proposal of rectifiers. In particular, we express that such phenomenon occur in graded systems, realizable materials in which thermal correction ubiquitously holds. In short, our results indicate that grade materials are genuine candidates for the actual fabrication of thermal diodes.
We are in debt to an anonymous reviewer, who called our attention to the connection between some of our solution and the Kirchhoff ’s theorem for circuits [ @Referee ]. We thank R. Sardenberg, M. Matos, M. C. Aguiar and B. V. Costa for the help with numeric programs. This employment was partially supported by CNPq (Brazil).
[ 32 ] { }
N. Li, J. Ren, L. Wang, G. Zhang, P. Hänggi, and B. Li, Rev. Mod. Phys. [ * * 84 * * ] { }, 1045 (2012); G. Casati, Nature Nanotech. [ * * 2 * * ] { }, 23 (2007); B. Hu, in [ * Advanced Workshop on Energy Transport in Low - Dimensional System: Achievements and Mysteries * ] { }, ICTP (2012), \[<http://cdsagenda5.ictp.trieste.it / full_display.php?ida = a11200>\ ].
M. Terraneo, M. Peyrard, and G. Casati, Phys. Rev. Lett. [ * * 88 * * ] { }, 094302 (2002).
B. Li, L. Wang, and G. Casati, Phys. Rev. Lett. [ * * 93 * * ] { }, 184301 (2004); L. Wang and B. Li, [ * ibid. * ] { } [ * * 99 * * ] { }, 177208 (2007); L. Wang and B. Li, [ * ibid. * ] { } [ * * 101 * * ] { }, 267203 (2008); B. Li [ * et al. * ] { }, Appl. Phys. Lett. [ * * 88 * * ] { }, 143501 (2006).
B. Hu, L. Yang, and Y. Zhang, Phys. Rev. Lett. [ * * 97 * * ] { }, 124302 (2006).
E. Pereira, Phys. Rev. E [ * * 82 * * ] { }, 040101 (R) (2010); E. Pereira, Physica A [ * * 390 * * ] { }, 4131 (2011).
C.W. Chang, D. Okawa, A. Majumdar, and A. Zettl, Science [ * * 314 * * ] { }, 1121 | rajge interactions may conriderably increcwe the rectirication power and may avoid its decey wuth tye system size, problemr of the lsual proposaow of rectihjers. In partidmlar, ce show that sucm phenomenot occurs in grdddd systems, realizable materials in whych thetmwl rectificatijn unyquifously holds. In short, our results jndicatt that graded matetials are genuine candidatfs flr the actual fabrlcation of jgeriql diodes.
We xre in debu co an anonyjous referee, who called our attdntiou to the cobnwctlmn between wome jf our resulbx and dhe Kirvhhoff’s theorek fmr xircuits [@Referee]. We tiank R. Sardenberg, M. Iatos, M. C. Ayuiar and B. V. Costa fir the velp witf nuoerjcel lrograls. Vhis work wzs partiallt supported by CNPq (Bwqzil).
[32]{}
N. Li, J. Reh, L. Wagg, G. Zhang, P. Hänggi, and B. Li, Rev. Mod. Phys. [**84**]{}, 1045 (2012); F. Casati, Nature Nanotecy. [**2**]{}, 23 (2007); B. Hu, in [*Advanced Workshop on Energy Transport in Low-Dimensional System: Achhevemxngs ckq Nydteries*]{}, ICTP (2012), \[<http://cdsagenda5.ictp.trieste.it/full_dyapkaj.php?ida=a11200>\].
M. Terranej, M. Peyrard, ajd D. Casati, Phys. Rev. Lzft. [**88**]{}, 094302 (2002).
B. Li, L. Wang, and G. Casaji, Phyw. Rev. Letu. [**93**]{}, 184301 (2004); K. Wang and B. Li, [*ibid.*]{} [**99**]{}, 177208 (2007); L. Wqng and B. Li, [*ubid.*]{} [**101**]{}, 267203 (2008); B. Li [*et al.*]{}, Appl. Phys. Uett. [**88**]{}, 143501 (2006).
B. Nu, L. Yang, and Y. Zhang, Pkys. Reb. Lett. [**97**]{}, 124302 (2006).
E. Oereira, Pghs. Rev. E [**82**]{}, 040101 (R) (2010); E. Oervira, Physica A [**390**]{}, 4131 (2011).
C.W. Chang, D. Okwwa, A. Majnmdar, and A. Xettk, Sciegce [**314**]{}, 1121 | range interactions may considerably increase the rectification may its decay the system size, of In particular, we that such phenomenon in graded systems, realizable materials in thermal rectification ubiquitously holds. In short, our results indicate that graded materials are candidates for the actual fabrication of thermal diodes. We are in debt to anonymous who our to the connection between some of our results and the Kirchhoff’s theorem for circuits [@Referee]. We R. Sardenberg, M. Matos, M. C. Aguiar and V. Costa for the with numerical programs. This work partially by CNPq [32]{} Li, Ren, L. Wang, Zhang, P. Hänggi, and B. Li, Rev. Mod. Phys. [**84**]{}, 1045 (2012); G. Casati, Nature Nanotech. [**2**]{}, (2007); B. [*Advanced Workshop Energy in System: Achievements and (2012), \[<http://cdsagenda5.ictp.trieste.it/full_display.php?ida=a11200>\]. M. Terraneo, M. Peyrard, Phys. Rev. Lett. [**88**]{}, 094302 (2002). B. Li, Wang, and Casati, Phys. Rev. Lett. [**93**]{}, 184301 L. Wang and B. Li, [*ibid.*]{} [**99**]{}, 177208 L. Wang and B. Li, [*ibid.*]{} [**101**]{}, 267203 (2008); B. Li [*et al.*]{}, Appl. Phys. 143501 (2006). B. Hu, Yang, and Y. Phys. Lett. 124302 E. Pereira, Rev. E [**82**]{}, 040101 (R) (2010); E. Pereira, Physica A [**390**]{}, (2011). C.W. Chang, D. Okawa, A. Majumdar, and A. Zettl, 1121 | range interactions may consiDerably incRease The RecTiFicaTion Power and may avoID its Decay with the system size, ProblEmS Of thE UsUal prOposals OF rECTifIeRs. in pArTIcUlar, wE shOw that sUch phenomeNon OcCurs in graded SYsTems, realizAblE materials in WhiCh therMaL reCTificAtiOn ubiQuitouSLy holdS. In short, oUr REsults INdicate THAt GradEd materials are genUInE Candidates for tHe actuAl FAbRICatIon Of thermal dIoDes.
We ARe in debT To AN ANonYMous referee, whO called our aTTenTion to ThE coNNectioN betwEeN SomE of our resulTs anD the KirchHoff’s tHEorem foR CircuitS [@ReferEe]. WE thAnk R. sArDeNbeRg, m. matOS, M. c. AgUIar And B. V. CosTa FoR the hElp wITH NUmerIcaL proGrams. this work was paRtiAlly SUppOrted By CNPQ (BraZiL).
[32]{}
N. Li, J. ren, L. WaNg, G. ZhAnG, P. Hänggi, and B. Li, REv. MoD. Phys. [**84**]{}, 1045 (2012); G. CasAti, naTurE NAnoteCH. [**2**]{}, 23 (2007); B. Hu, in [*advAncEd WorksHop on EnERgy trANSPoRt in Low-Dimensional sySTEm: achievemEnts anD myStERies*]{}, ICTP (2012), \[<HtTp://cDsagENDa5.ictP.triEStE.it/full_dIsplay.PHp?IdA=a11200>\].
M. TerrAnEo, M. PeyRaRd, aNd G. casatI, phys. rev. LetT. [**88**]{}, 094302 (2002).
B. Li, L. WanG, and G. cAsati, Phys. Rev. LeTT. [**93**]{}, 184301 (2004); L. Wang and B. Li, [*iBId.*]{} [**99**]{}, 177208 (2007); l. wAnG And B. li, [*iBid.*]{} [**101**]{}, 267203 (2008); B. Li [*et al.*]{}, APpl. PHYs. LeTt. [**88**]{}, 143501 (2006).
B. HU, l. YAng, ANd Y. ZhAng, PhYs. rEv. lEtt. [**97**]{}, 124302 (2006).
E. Pereira, Phys. Rev. E [**82**]{}, 040101 (r) (2010); E. pereirA, PhysIca A [**390**]{}, 4131 (2011).
C.W. Chang, D. OKawa, A. MajumDAR, And A. ZettL, SciENcE [**314**]{}, 1121 | range interactions may co nsiderably incr eas e t he rec tifi cation power a n d ma y avoid its decay with thesy s tems iz e, pr oblemso ft h e u su al pr op o sa ls of re ctifier s. In part icu la r, we show t h at such phen ome non occurs i n g radedsy ste m s, re ali zable mater i als in which th er m al rec t ificati o n u biqu itously holds. In sh o rt, our result s indi ca t et h atgra ded materi al s are genuine ca n d i dat e s for the act ual fabrica t ion of th er mal diodes .
We a r e i n debt to a n an onymous r eferee , who ca l led our atten tio n t o th e c on nec ti o n b e tw een som e of our r es ultsandt h e Kirc hho ff’s theo rem for circu its [@R e fer ee].We th ankR. Sard enberg , M.Ma tos, M. C. Agui ar a nd B. V.Cos ta fo rthe h e lp wit h n ume rical p rograms . Th is w o rk was partially sup po r t ed by CNPq (Braz i l) .[32]{}
N. Li , J. R en, L . Wa n g, G. Zhan g, P.H än gg i, andB. Li, R ev . M od. Phys . [** 84**]{ }, 1045(2012 ) ; G. Casati, N a ture Nanotech . [ * * 2* * ]{}, 23 (2007); B. Hu, in [ *Adv a nc edW orksh op on E n er g y Transport in Low- Di mensio nal S ystem: Achiev ements and M y steries* ]{}, IC T P (2012), \[<h ttp:/ /cdsagenda 5 .ictp.tr ieste .it/full _display. p h p?ida=a1 120 0>\ ].
M. T er raneo, M. Pey r a rd,an d G. Ca sat i, Phys . R ev. Le tt. [ **88**]{} , 094302 ( 20 02 ).
B . Li, L. Wang, a ndG. Ca sati, Phys.Rev.Lett .[* * 93* *]{}, 1 8 43 0 1 (20 04 ); L.Wan gand B . Li , [* ibid.*] {} [**99* *]{ } , 17 72 08 (2007) ; L. Wang and B . Li, [*ib id .*] {} [** 1 0 1**]{},267203 (2008); B. Li [* e t al.*] {}, Appl . Ph ys. Lett. [* *88**] {}, 143501 (2006 ).
B .Hu, L . Yan g , a ndY. Zhang, Ph y s . R ev. L et t. [ **97**] {}, 124302 (2006).
E. Pereira, Phy s.Rev. E [ **8 2 ** ] {}, 0 4 010 1 (R) (2010); E.Pereira, P hy s ic a A [**390 * *]{ }, 4131 ( 2011).
C.W. Chang,D. Okawa, A. Majum da r, a n d A. Zettl, Sc ience [* *314**]{} , 1121 | range_interactions may_considerably increase the rectification_power and_may_avoid its_decay_with the system_size, problems of_the usual proposals of_rectifiers. In particular,_we_show that such phenomenon occurs in graded systems, realizable materials in which thermal rectification_ubiquitously_holds. In_short,_our_results indicate that graded materials_are genuine candidates for the_actual fabrication_of thermal diodes.
We are in debt to an_anonymous_referee, who called_our attention to the connection between some of our_results and the Kirchhoff’s theorem for_circuits [@Referee]. We_thank_R._Sardenberg, M. Matos, M._C. Aguiar and B. V. Costa_for the help with numerical programs._This work was partially supported by CNPq_(Brazil).
[32]{}
N. Li, J. Ren, L. Wang,_G. Zhang, P. Hänggi, and_B. Li,_Rev. Mod. Phys. [**84**]{}, 1045_(2012); G. Casati,_Nature Nanotech._[**2**]{}, 23 (2007);_B. Hu, in [*Advanced Workshop on_Energy Transport in_Low-Dimensional System: Achievements and Mysteries*]{}, ICTP_(2012),_\[<http://cdsagenda5.ictp.trieste.it/full_display.php?ida=a11200>\].
M. Terraneo, M._Peyrard,_and_G. Casati,_Phys. Rev. Lett._[**88**]{},_094302 (2002).
B._Li,_L. Wang, and G. Casati, Phys._Rev._Lett. [**93**]{}, 184301 (2004); L. Wang and_B. Li, [*ibid.*]{} [**99**]{},_177208_(2007); L. Wang and_B. Li, [*ibid.*]{} [**101**]{}, 267203_(2008); B. Li [*et al.*]{}, Appl._Phys. Lett._[**88**]{}, 143501_(2006).
B. Hu, L. Yang, and Y. Zhang, Phys. Rev. Lett. [**97**]{},_124302 (2006).
E. Pereira, Phys. Rev. E_[**82**]{}, 040101 (R) (2010);_E. Pereira,_Physica_A [**390**]{}, 4131_(2011).
C.W._Chang, D._Okawa, A. Majumdar, and A. Zettl, Science_[**314**]{}, 1121 |
presence of plasma a major change in the value $R_{Bc}$, Eq.(\[RBc\]), and thus a major increase in the absolute yield of the $B_c$ mesons.
-0.2cm 0.5cm -0.5cm
-0.5cm
We cannot describe here all the intricacies of the population evolution equations that establish the final production rates of the $B_c$. We mentioned already that there are several channels of production. Beyond that the issue is to assess the prospects for the survival of the $B_{c} $ state in the hot parton environment. Using Kharzeev’s approach [@kh98], we have calculated the quarkonium dissociation rates for $J/\psi $, $B_{c} $ and $\Upsilon(1S) $, as shown in Fig.\[F1\]. The curves there indicate that $B_{c} $ should travel 5 to 10 times further than $J/\psi $ at $T = 300$MeV. It is worth noting that the probability of dissociation of the $B_{c} $ at $T=400$MeV is about the same as that of $J/\psi $ at $T=300$MeV.
Potential Model Approaches to the ${\bf B_{c}} $ System {#bcbound}
=======================================================
Another way to understand the unusual stability of the $B_c$ mesons in the plasma phase is to consider its bound state structure. The rich spectra of bound states of charmonium and the upsilon system below the flavor thresholds [@pdg98] provide enough information to determine the parameters of a nonrelativistic potential model for heavy quarks. The assumption of flavor independence extends this model to the $B_{c} $ system. We describe here one such recent calculation [@fu98] incorporating a model of a running coupling constant effects in the central potential and the full radiative one-loop expressions supplemented by the Gromes consistency condition to incorporate non-perturbative effects in the spin-dependent potentials: these calculations provided an excellent fit of the upsilon levels (avg. dev. = 4.3MeV), a good fit of the charmonium levels (avg. dev. = 19.9MeV) and a good account of the leptonic widths below threshold. The model predicts a ground state energy of $M(B_{c}) = 6286 ^{+15}_{-6}$MeV and that this state is 820MeV below the threshold for | presence of plasma a major change in the value $ R_{Bc}$, Eq.(\[RBc\ ]), and thus a major increase in the absolute output of the $ B_c$ meson.
-0.2 cm 0.5 cm -0.5 centimeter
-0.5 curium
We cannot describe here all the intricacies of the population development equations that establish the final product rates of the $ B_c$. We mentioned already that there be several channels of output. Beyond that the offspring is to assess the candidate for the survival of the $ B_{c } $ state in the hot parton environment. use Kharzeev ’s approach [ @kh98 ], we have calculated the quarkonium dissociation rates for $ J/\psi $, $ B_{c } $ and $ \Upsilon(1S) $, as show in Fig.\[F1\ ]. The curves there indicate that $ B_{c } $ should travel 5 to 10 times further than $ J/\psi $ at $ T = 300$MeV. It is deserving noting that the probability of dissociation of the $ B_{c } $ at $ T=400$MeV is approximately the same as that of $ J/\psi $ at $ T=300$MeV.
Potential Model Approaches to the $ { \bf B_{c } } $ System { # bcbound }
= = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = =
Another way to understand the strange stability of the $ B_c$ mesons in the plasma phase is to consider its bound state structure. The rich spectra of bound states of charmonium and the upsilon system below the flavor thresholds [ @pdg98 ] provide enough information to determine the parameter of a nonrelativistic potential model for big quark. The assumption of flavor independence extends this exemplar to the $ B_{c } $ system. We describe here one such recent calculation [ @fu98 ] incorporating a exemplar of a running coupling constant effects in the central potential and the broad radiative one - loop expressions supplemented by the Gromes consistency condition to integrate non - perturbative effects in the tailspin - dependent potentials: these calculations provided an excellent fit of the upsilon levels (avg. dev. = 4.3MeV), a effective fit of the charmonium levels (avg. dev. = 19.9MeV) and a good explanation of the leptonic widths below threshold. The model predicts a footing state department of energy of $ M(B_{c }) = 6286 ^{+15}_{-6}$MeV and that this state is 820MeV below the threshold for | prfsence of plasma a major change in the rqlue $R_{Uc}$, Eq.(\[RBd\]), and thjs a major increase in the ausolyte yueld of the $B_c$ mesons.
-0.2co 0.5cm -0.5cm
-0.5cm
Ae cannor dewxribe here all the intridwciev of the populajion evolutimn equations tvag zstablish the final production rates of the $B_f$. We mentioned alrtadr thzn uhere are several channels of prosuction. Beyond that yhe issue is to assess the prodpects for the surgival of thg $B_{c} $ state in thd hot parton environmeht. Using Kharzeev’s approach [@kh98], de hare calculatgb thf quarkonium dissjciation ratcx for $B/\psi $, $B_{v} $ and $\Upsilon(1X) $, es syown in Fig.\[F1\]. The curvxs there indicate thwt $B_{c} $ shmumd travel 5 to 10 tinew furjher dhan $H/\psk $ zt $T = 300$MeV. Lt ms worth nofing that tye probability of dosficiation of tge $B_{c} $ ae $T=400$MeV is about the same as that of $J/\psp $ af $T=300$MeV.
Potential Model Approaches to the ${\bf B_{f}} $ System {#bcbound}
=======================================================
Another way to understand the unusual statilitb uf uhc $C_x$ lesons in the plasma phase is to consider its goinc state structmre. The rich spectta og bound states of chcdmknium and the upsipon sysjem beoow the fjavot thresholds [@pdg98] provide enoygh informatpon ro determine the pcrameters of a npnrelstivistic potential modzl for heavy quarns. The asajmption of flavof ikde[endence extends this modej to the $U_{c} $ svstem. We desvribe rere one skch rcwent calculation [@fu98] incotporathng a modep of a running coupling constanv effects in jhe cettral pocentiak and the fulj radiative ong-loop expxessiovs supplemvnted by vhe Gromes cjnsistency cotfition to inrorporate non-pertyrbativd effects in thr spin-depvnbent potebtials: these calcukatkkns provided an exxellent fit of yhe upfipoi levqns (avg. dev. = 4.3KeV), x gupd fig of tht cmarooniim levels (avg. dev. = 19.9MaV) ahd a good account pf the lepjonic widehs below thrrshold. The model pgedicvs a gcound xtaje energy of $M(B_{c}) = 6286 ^{+15}_{-6}$MeV and that this stahe ls 820MeV below ehe bhrefhold for | presence of plasma a major change in $R_{Bc}$, and thus major increase in $B_c$ -0.2cm 0.5cm -0.5cm We cannot describe all the intricacies of the population equations that establish the final production rates of the $B_c$. We mentioned already there are several channels of production. Beyond that the issue is to assess prospects the of $B_{c} $ state in the hot parton environment. Using Kharzeev’s approach [@kh98], we have calculated the dissociation rates for $J/\psi $, $B_{c} $ and $, as shown in The curves there indicate that $ travel 5 10 further $J/\psi $ at = 300$MeV. It is worth noting that the probability of dissociation of the $B_{c} $ at $T=400$MeV about the that of $ $T=300$MeV. Model Approaches to B_{c}} $ System {#bcbound} ======================================================= Another the unusual stability of the $B_c$ mesons in plasma phase to consider its bound state structure. rich spectra of bound states of charmonium and upsilon system below the flavor thresholds [@pdg98] provide enough information to determine the parameters of potential model for heavy The assumption of independence this to $B_{c} $ We describe here one such recent calculation [@fu98] incorporating a model a running coupling constant effects in the central potential and radiative expressions supplemented by Gromes consistency condition to non-perturbative in the spin-dependent potentials: provided excellent upsilon (avg. = 4.3MeV), a good of the charmonium levels (avg. = 19.9MeV) and a widths below threshold. The model predicts a ground energy of $M(B_{c}) = 6286 ^{+15}_{-6}$MeV and this state is 820MeV below the threshold for | presence of plasma a major chaNge in the vaLue $R_{BC}$, Eq.(\[rBc\]), AnD thuS a maJor increase in tHE absOlute yield of the $B_c$ mesonS.
-0.2cm 0.5cm -0.5Cm
-0.5CM
We cANnOt desCribe heRE aLL The InTrIcaCiES oF the pOpuLation eVolution eqUatIoNs that establISh The final prOduCtion rates of The $b_c$. We meNtIonED alreAdy That tHere arE SeveraL channels Of PRoductIOn. BeyonD THaT the Issue is to assess thE PrOSpects for the suRvival Of THe $b_{C} $ StaTe iN the hot parToN enviROnment. USInG kHArzEEv’s approach [@kh98], We have calcuLAteD the quArKonIUm dissOciatIoN RatEs for $J/\psi $, $B_{c} $ And $\UPsilon(1S) $, as Shown iN fig.\[F1\]. The CUrves thEre indIcaTe tHat $B_{C} $ ShOuLd tRaVEl 5 tO 10 TiMes FUrtHer than $J/\PsI $ aT $T = 300$MeV. it is WORTH notIng That The prObability of diSsoCiatIOn oF the $B_{C} $ at $T=400$MEV is AbOut thE same aS that Of $j/\psi $ at $T=300$MeV.
PotenTial model ApprOacHeS to ThE ${\bf B_{c}} $ sYstem {#bCboUnd}
=======================================================
another Way to unDErsTaND THe Unusual stability of ThE $b_C$ mEsons in tHe plasMA pHaSE is to conSiDer Its bOUNd staTe stRUcTure. The rIch speCTrA oF bound sTaTes of cHaRmoNiuM and tHE upsIlon syStem beloW the fLAvor thresholds [@PDg98] provide enouGH iNFOrMAtioN to Determine thE parAMeteRs of A NoNreLAtiviStic pOtENtIAl model for heavy quarKs. the assUmptiOn of flavor indEpendence eXTENds this mOdel TO tHE $B_{c} $ system. We desCribe Here one sucH Recent caLculaTion [@fu98] inCorporatiNG A model of A ruNniNg cOupLINg Constant effecTS In thE cEntral pOteNtial anD thE fuLl rAdiAtIve one-looP expressIoNs SuPpLemEnted BY the GromEs ConSiSteNcy coNDition To incOrpoRaTe NOn-pErturbaTIvE EFfecTs In The sPin-DePendeNt poTEntIals: theSe calculaTioNS proViDeD an exceLlent fit of the UpSilon levelS (aVg. dEv. = 4.3MeV), a GOOd fit of tHe charmonium levels (avg. deV. = 19.9meV) and a GooD accoUnt oF the leptoNic Widths BelOW thresHold. ThE modeL pRedICTs a grOUNd StaTe Energy of $M(B_{C}) = 6286 ^{+15}_{-6}$mEV aNd thaT tHis sTate is 820MEV below the thresholD For | presence of plasma a majo r change i n the va lue $ R_{B c}$, Eq.(\[RBc\]), andthus a major increasein th ea bsol u te yiel d of th e $ B _ c$me so ns.
- 0. 2cm 0 .5c m -0.5c m
-0.5cm
We c annot descri b ehere all t heintricaciesofthe po pu lat i on ev olu tionequati o ns tha t establi sh the fi n al prod u c ti on r ates of the $B_c$ . W e mentioned alr eady t ha t t h e reare several c ha nnels of prod u ct i o n . B e yond that the issue is t o as sess t he pr o spects forth e su rvival of t he $ B_{c} $ s tate i n the ho t parton envir onm ent . Us i ng K har ze e v’s ap pro a ch[@kh98], w ehavecalc u l a t ed t hequar koniu m dissociatio n r ates for $J/\ psi $ , $B _{ c} $and $\ Upsil on (1S) $, as show n in Fig.\[F1 \]. T hecu rvest here i ndi cat e that$B_{c}$ sh ou l d tr avel 5 to 10 times f u r th er than$J/\ps i $ a t $T = 30 0$ MeV . It i s wor th n o ti ng thatthe pr o ba bi lity of d issoci at ion of the$ B_{c } $ at $T=400$ MeV i s about the sam e as that of $ J /\ p s i$ at$T= 300$MeV.
P oten t ialMode l A ppr o aches to t he ${ \ bf B_{c}} $ System{# bcboun d}
== ============= ========== = = = ======== ==== = == = ===========
A nothe r way to u n derstand theunusualstability o f the $B _c$ me son s i n th e plasma phas e is t oconside r i ts boun d s tat e s tru ct ure. Therich spe ct ra o fbou nd st a tes of c ha rmo ni umand t h e upsi lon s yste mbe l owthe fla v or t hres ho ld s [@ pdg 98 ] pro vide eno ugh inf ormationtod eter mi ne the pa rameters of a n onrelativi st icpotent i a l modelfor heavy quarks. The a s sumptio n o f fla vorindepende nce exten dst his mo del to the$B _{c } $ sys t e m. We d escribe he r e on e suc hrece nt calc ulation [@fu98] in c orp orating a mod elof a r un nin g c o upl in g co n s tant effects in the centr al po tential an d th efull ra diative one- l oop exp ressionssupplemen te d by t heGromes con sistency conditio n to i n co rpora tenon-pe rt urb ative effec t s i n the spin- de penden t pot en tials: t hese calculations provi ded an exce lle nt fit of th e up silon lev els(avg. dev. =4.3 MeV), ag ood f it o f t hec harmo nium levels (a v g. de v . = 19.9MeV) a n d a g ood a cco u nt oftheleptonic widths b e low threshold. The m ode l p r edic ts a ground stat e e ne r g y of $M( B_ {c}) = 6286 ^{+15}_ {- 6 }$MeV and t hat th is stat e is 820MeV bel owthe thres hol df or | presence_of plasma_a major change in_the value_$R_{Bc}$,_Eq.(\[RBc\]), and_thus_a major increase_in the absolute_yield of the $B_c$_mesons.
-0.2cm 0.5cm -0.5cm
-0.5cm
We_cannot_describe here all the intricacies of the population evolution equations that establish the final_production_rates of_the_$B_c$._We mentioned already that there_are several channels of production._Beyond that_the issue is to assess the prospects for_the_survival of the_$B_{c} $ state in the hot parton environment. Using_Kharzeev’s approach [@kh98], we have calculated_the quarkonium dissociation_rates_for_$J/\psi $, $B_{c} $_and $\Upsilon(1S) $, as shown in_Fig.\[F1\]. The curves there indicate that_$B_{c} $ should travel 5 to 10_times further than $J/\psi $ at_$T = 300$MeV. It is_worth noting_that the probability of dissociation_of the $B_{c}_$ at_$T=400$MeV is about_the same as that of $J/\psi_$ at $T=300$MeV.
Potential_Model Approaches to the ${\bf B_{c}}_$_System {#bcbound}
=======================================================
Another way_to_understand_the unusual_stability of the_$B_c$_mesons in_the_plasma phase is to consider its_bound_state structure. The rich spectra of bound_states of charmonium and_the_upsilon system below the_flavor thresholds [@pdg98] provide enough information_to determine the parameters of a_nonrelativistic potential_model for_heavy quarks. The assumption of flavor independence extends this model to_the $B_{c} $ system. We describe_here one such recent_calculation [@fu98] incorporating_a_model of a_running_coupling constant_effects in the central potential and the_full radiative_one-loop expressions supplemented by the Gromes_consistency condition to incorporate_non-perturbative_effects in the spin-dependent potentials: these_calculations provided an excellent fit of_the upsilon levels (avg. dev._=_4.3MeV),_a good fit of the_charmonium levels (avg. dev. = 19.9MeV)_and a good_account of the leptonic widths below threshold._The_model predicts a ground state energy_of_$M(B_{c}) = 6286 ^{+15}_{-6}$MeV and that_this_state_is 820MeV below the threshold_for |
him for sending us a copy of his paper [@korbas]. We would like to thank the anonymous referee for his detailed suggestions. In particular, we thank him for showing us the proof of Proposition\[secondlastprop\]. This is shorter and stronger than proof given by the authors.
[99]{}
Adams, J. F.: *Vector fields on Spheres*, Ann. Math. **75** (1962), 603-632.
Atiyah, M.—Hirzebruch, F.: *Bott periodicity and the parallelizability of the spheres*, In: Proc. Cambridge Philos. Soc., 57 (1961), pp. 223-226.
Balko, L’.—Korbaš, J.: *A note on the characteristic rank of a smooth manifold*, Group actions and homogeneous spaces, Fak. Mat. Fyziky Inform. Univ. Komenského, Bratislava, 2010, pp. 1-8.
Balko, L’.—Korbaš, J.: *A note of the characteristic rank of null-cobordant manifolds*, To appear in Acta. Math. Hungar, 2011/2012.
Borel, A.: *Sur la cohomologie des espaces fibrés principaux et des espaces homogènes de groupes de Lie compacts*, Ann. of Math. **57** (1953), 115-207.
Dold, A.: *Erzeugende der Thomschen Algebra $\mathfrak{N}$*, Math. Z. **65** (1956), 25-35.
Hatcher, A.: *Algebraic Topology*, Cambridge Univ. Press, 2002.
Husemoller, D.: *Fibre Bundles*, Springer-Verlag, New York, 1966.
Korbaš, J.: *Bounds for the cup-length of Poincaré spaces and their applications*, Topology Appl. **153** (2006), 2976-2986.
Korbaš, J.: *The cup-length of the oriented Grassmannians vs a new bound for zero cobordant manifolds*, Bull. Belg. Math. Soc.-Simon Stevin **17** (2010), 69-81.
Milnor, J.: *Some consequences of a theorem of Bott*, Ann. of Math. **68** (1958), 444 | him for sending us a copy of his paper [ @korbas ]. We would like to thank the anonymous reviewer for his detailed trace. In particular, we thank him for showing us the validation of Proposition\[secondlastprop\ ]. This is shorter and hard than proof given by the writer.
[ 99 ] { }
Adams, J. F.: * Vector fields on Spheres *, Ann. Math. * * 75 * * (1962), 603 - 632.
Atiyah, M.—Hirzebruch, F.: * Bott periodicity and the parallelizability of the sphere *, In: Proc. Cambridge Philos. Soc. , 57 (1961), pp. 223 - 226.
Balko, L’.—Korbaš, J.: * A note on the characteristic rank of a placid manifold *, Group actions and homogeneous spaces, Fak. Mat. Fyziky Inform. Univ. Komenského, Bratislava, 2010, pp. 1 - 8.
Balko, L’.—Korbaš, J.: * A eminence of the characteristic membership of null - cobordant manifolds *, To appear in Acta. Math. Hungar, 2011/2012.
Borel, A.: * Sur la cohomologie des espaces fibrés principaux et des espaces homogènes de groupes de Lie compacts *, Ann. of Math. * * 57 * * (1953), 115 - 207.
Dold, A.: * Erzeugende der Thomschen Algebra $ \mathfrak{N}$ *, Math. Z. * * 65 * * (1956), 25 - 35.
Hatcher, A.: * Algebraic Topology *, Cambridge Univ. Press, 2002.
Husemoller, D.: * Fibre Bundles *, Springer - Verlag, New York, 1966.
Korbaš, J.: * Bounds for the cup - length of Poincaré quad and their applications *, Topology Appl. * * 153 * * (2006), 2976 - 2986.
Korbaš, J.: * The cup - length of the oriented Grassmannians vs a modern bound for zero cobordant manifolds *, Bull. Belg. Math. Soc.-Simon Stevin * * 17 * * (2010), 69 - 81.
Milnor, J.: * Some consequences of a theorem of Bott *, Ann. of Math. * * 68 * * (1958), 444 | hil for sending us a copy uf his paper [@kotbqs]. We xould ljke to tfank the anonymous referee flr his eetailed suggestions. Iv particupar, we tyank yim for showing us the pdlof mh Proposition\[segondlastpro[\]. This is shordef cnd stronger than proof given by the authorx.
[99]{}
Afams, J. F.: *Vectot fiekqs oh Spheres*, Ann. Math. **75** (1962), 603-632.
Atiyah, M.—Hirzegruch, F.: *Bott periodivity and the parallelizabipity of the spheres*, In: Proc. Cambrudge Philos. Soc., 57 (1961), pp. 223-226.
Balko, L’.—Korbaš, J.: *A noje on the characteristic rank of a smpoth manifile*, Ggmup actions and romogeneous spaces, Fdk. Mat. Gyziky Inform. Mniv. Nomwnského, Bratislava, 2010, pp. 1-8.
Ualko, L’.—Korbaš, J.: *A notg of the cvaxacteristic rank of nylo-cobotdant mankdolas*, Uo eppsar in Acva. Math. Hunfar, 2011/2012.
Borel, A.: *Sur la cohomologie dqw espaces fibdés prigcypaux et des espaces homogènes de groupev ds Lie compacts*, Ann. of Mqth. **57** (1953), 115-207.
Dold, A.: *Erzeugenfe der Thjmschen Algebra $\mathfrak{N}$*, Math. Z. **65** (1956), 25-35.
Hatcher, A.: *Algetraic Gopiljet*, Fambridge Univ. Press, 2002.
Husemoller, D.: *Fibre Bundlqa*, Xpginger-Verlag, New Nork, 1966.
Korbaš, J.: *Boundx vot the cup-lengtf of Poindaré spaces and thelr applycatiins*, Topoljgy Sppl. **153** (2006), 2976-2986.
Korbaš, J.: *The cup-lengty of the orivntee Grassmannians vs a new bounb for eero cpbordant manifolds*, Bull. Belf. Math. Soc.-Slmon Stevjv **17** (2010), 69-81.
Milnor, J.: *Somd cpnvequences of a theorem of Fott*, Ann. if Mcth. **68** (1958), 444 | him for sending us a copy of [@korbas]. would like thank the anonymous In we thank him showing us the of Proposition\[secondlastprop\]. This is shorter and than proof given by the authors. [99]{} Adams, J. F.: *Vector fields on Ann. Math. **75** (1962), 603-632. Atiyah, M.—Hirzebruch, F.: *Bott periodicity and the parallelizability the In: Cambridge Soc., 57 (1961), pp. 223-226. Balko, L’.—Korbaš, J.: *A note on the characteristic rank of a manifold*, Group actions and homogeneous spaces, Fak. Mat. Inform. Univ. Komenského, Bratislava, pp. 1-8. Balko, L’.—Korbaš, J.: note the characteristic of manifolds*, appear in Acta. Hungar, 2011/2012. Borel, A.: *Sur la cohomologie des espaces fibrés principaux et des espaces homogènes de groupes Lie compacts*, Math. **57** 115-207. A.: der Thomschen Algebra Z. **65** (1956), 25-35. Hatcher, A.: Univ. Press, 2002. Husemoller, D.: *Fibre Bundles*, Springer-Verlag, York, 1966. J.: *Bounds for the cup-length of spaces and their applications*, Topology Appl. **153** (2006), Korbaš, J.: *The cup-length of the oriented Grassmannians vs a new bound for zero cobordant Belg. Math. Soc.-Simon Stevin (2010), 69-81. Milnor, *Some of theorem Bott*, Ann. Math. **68** (1958), 444 | him for sending us a copy of his Paper [@korbaS]. We woUld LikE tO thaNk thE anonymous refeREe foR his detailed suggestionS. In paRtICulaR, We Thank Him for sHOwING us ThE pRooF oF prOposiTioN\[secondLastprop\]. ThIs iS sHorter and strONgEr than prooF giVen by the authOrs.
[99]{}
adams, J. f.: *VEctOR fielDs oN SpheRes*, Ann. mAth. **75** (1962), 603-632.
AtiYah, M.—HirzeBrUCh, F.: *BotT PeriodiCITy And tHe parallelizabiliTY oF The spheres*, In: PrOc. CambRiDGe pHIloS. SoC., 57 (1961), pp. 223-226.
Balko, L’.—KOrBaš, J.: *A NOte on thE ChARACteRIstic rank of a sMooth manifoLD*, GrOup actIoNs aND homogEneouS sPAceS, Fak. Mat. FyziKy InForm. Univ. KOmenskÉHo, BratiSLava, 2010, pp. 1-8.
BAlko, L’.—KOrbAš, J.: *a notE Of ThE chArACteRIsTic RAnk Of null-coBoRdAnt maNifoLDS*, tO appEar In AcTa. MatH. Hungar, 2011/2012.
Borel, A.: *sur La coHOmoLogie Des esPaceS fIbrés PrinciPaux eT dEs espaces homogèNes dE groupes dE LiE cOmpAcTs*, Ann. OF Math. **57** (1953), 115-207.
DOld, a.: *ErZeugendE der ThoMSchEn aLGEbRa $\mathfrak{N}$*, Math. Z. **65** (1956), 25-35.
HaTcHER, A.: *algebraiC TopolOGy*, caMBridge UnIv. preSs, 2002.
HuSEMolleR, D.: *FiBRe bundles*, SPringeR-veRlAg, New YoRk, 1966.
korbaš, j.: *BOunDs fOr the CUp-leNgth of poincaré SpaceS And their applicATions*, Topology aPpL. **153** (2006), 2976-2986.
kOrBAš, J.: *THe cUp-length of tHe orIEnteD GraSSmAnnIAns vs A new bOuND fOR zero cobordant manifOlDs*, Bull. belg. MAth. Soc.-Simon StEvin **17** (2010), 69-81.
Milnor, j.: *sOMe conseqUencES oF A theorem of Bott*, ann. of math. **68** (1958), 444 | him for sending us a copy of his pa per [ @ko rba s] . We wou ld like to tha n k th e anonymous referee fo r his d e tail e dsugge stions. In p art ic ul ar, w e t hankhim for sh owing us t hepr oof of Propo s it ion\[secon dla stprop\]. Th isis sho rt era nd st ron ger t han pr o of giv en by the a u thors.
[99]{} Ad ams, J. F.: *Vector f i el d s on Spheres*, Ann.Ma t h. * *75 **(1962), 60 3- 632.Atiyah, M. — H i rze b ruch, F.: *Bo tt periodic i tyand th epar a lleliz abili ty ofthe spheres *, I n: Proc.Cambri d ge Phil o s. Soc. , 57 ( 196 1), pp. 22 3- 226 .Bal k o, L’ . —Ko rbaš, J. :*A note ont h e char act eris tic r ank of a smoo thmani f old *, Gr oup a ctio ns andhomoge neous s paces, Fak. Mat . Fy ziky Info rm. U niv .Komen s kého,Bra tis lava, 2 010, pp . 1- 8. B al ko, L’.—Korbaš, J. :* A n ote of t he cha r ac te r istic ra nk of nul l - cobor dant ma nifolds* , To a p pe ar in Act a. Math. H ung ar, 2011 / 2012 .
Bor el, A.:*Surl a cohomologied es espaces fi b ré s pr i ncip aux et des esp aces homo gène s d e g r oupes de L ie co m pacts*, Ann. of Mat h. **57* * (19 53), 115-207.
Dold, A. : * Erzeugen de d e rT homschen Algeb ra $\ mathfrak{N } $*, Math . Z.**65** ( 1956), 25 - 3 5.
Hatc her , A .:*Al g e br aic Topology* , Camb ri dge Uni v.Press,200 2.
Hu sem ol ler, D.:*Fibre B un dl es *, Sp ringe r -Verlag, N ewYo rk, 1966 .
Korb aš, J .: * Bo un d s f or thec up - l engt hof Poi nca ré spac es a n d t heir ap plication s*, Topo lo gy Appl.**153** (2006 ), 2976-2986 .
Ko rbaš,J . : *The c up-length of the orient e d Grass man nians vsa new bou ndfor ze roc oborda nt man ifold s* , B u l l. Be l g .Mat h. Soc.-Simo n Ste vin * *1 7**(2010), 69-81.
Milnor, J . : * Some conseque nce s of a t heo r em ofBo t t*, A nn. of Math. ** 68** (1958 ), 44 4 | him_for sending_us a copy of_his paper_[@korbas]._We would_like_to thank the_anonymous referee for_his detailed suggestions. In_particular, we thank_him_for showing us the proof of Proposition\[secondlastprop\]. This is shorter and stronger than proof_given_by the_authors.
[99]{}
Adams,_J._F.: *Vector fields on Spheres*,_Ann. Math. **75** (1962), 603-632.
Atiyah,_M.—Hirzebruch, F.:_*Bott periodicity and the parallelizability of the spheres*,_In:_Proc. Cambridge Philos._Soc., 57 (1961), pp. 223-226.
Balko, L’.—Korbaš, J.: *A note on_the characteristic rank of a smooth_manifold*, Group actions_and_homogeneous_spaces, Fak. Mat. Fyziky_Inform. Univ. Komenského, Bratislava, 2010, pp. 1-8.
Balko,_L’.—Korbaš, J.: *A note of the_characteristic rank of null-cobordant manifolds*, To appear_in Acta. Math. Hungar, 2011/2012.
Borel, A.:_*Sur la cohomologie des espaces_fibrés principaux_et des espaces homogènes de_groupes de Lie_compacts*, Ann._of Math. **57**_(1953), 115-207.
Dold, A.: *Erzeugende der Thomschen_Algebra $\mathfrak{N}$*, Math._Z. **65** (1956), 25-35.
Hatcher, A.: *Algebraic_Topology*,_Cambridge Univ. Press,_2002.
Husemoller,_D.:_*Fibre Bundles*,_Springer-Verlag, New York,_1966.
Korbaš,_J.: *Bounds_for_the cup-length of Poincaré spaces and_their_applications*, Topology Appl. **153** (2006), 2976-2986.
Korbaš, J.:_*The cup-length of the_oriented_Grassmannians vs a new_bound for zero cobordant manifolds*,_Bull. Belg. Math. Soc.-Simon Stevin **17**_(2010), 69-81.
Milnor,_J.: *Some_consequences of a theorem of Bott*, Ann. of Math. **68** (1958),_444 |