section \Optimizations for Code Float\ theory Optimize_Float imports "HOL-Library.Float" Optimize_Integer begin lemma compute_bitlen[code]: "bitlen a = (if a > 0 then log2 a + 1 else 0)" by (simp add: bitlen_alt_def log2_def) lemma compute_float_plus[code]: "Float m1 e1 + Float m2 e2 = (if m1 = 0 then Float m2 e2 else if m2 = 0 then Float m1 e1 else if e1 \ e2 then Float (m1 + m2 * power_int 2 (e2 - e1)) e1 else Float (m2 + m1 * power_int 2 (e1 - e2)) e2)" by (simp add: Float.compute_float_plus power_int_def) lemma compute_real_of_float[code]: "real_of_float (Float m e) = (if e \ 0 then m * 2 ^ nat e else m / power_int 2 (-e))" unfolding power_int_def[symmetric, of 2 e] using compute_real_of_float power_int_def by auto lemma compute_float_down[code]: "float_down p (Float m e) = (if p + e < 0 then Float (m div power_int 2 (-(p + e))) (-p) else Float m e)" by (simp add: Float.compute_float_down power_int_def) lemma compute_lapprox_posrat[code]: fixes prec::nat and x y::nat shows "lapprox_posrat prec x y = (let l = rat_precision prec x y; d = if 0 \ l then int x * power_int 2 l div y else int x div power_int 2 (- l) div y in normfloat (Float d (- l)))" by (auto simp add: Float.compute_lapprox_posrat power_int_def Let_def zdiv_int of_nat_power of_nat_mult) lemma compute_rapprox_posrat[code]: fixes prec x y defines "l \ rat_precision prec x y" shows "rapprox_posrat prec x y = (let l = l ; (r, s) = if 0 \ l then (int x * power_int 2 l, int y) else (int x, int y * power_int 2 (-l)) ; d = r div s ; m = r mod s in normfloat (Float (d + (if m = 0 \ y = 0 then 0 else 1)) (- l)))" by (auto simp add: l_def Float.compute_rapprox_posrat power_int_def Let_def zdiv_int of_nat_power of_nat_mult) lemma compute_float_truncate_down[code]: "float_round_down prec (Float m e) = (let d = bitlen (abs m) - int prec - 1 in if 0 < d then let P = power_int 2 d ; n = m div P in Float n (e + d) else Float m e)" by (simp add: Float.compute_float_round_down power_int_def cong: if_cong) lemma compute_int_floor_fl[code]: "int_floor_fl (Float m e) = (if 0 \ e then m * power_int 2 e else m div (power_int 2 (-e)))" by (simp add: Float.compute_int_floor_fl power_int_def) lemma compute_floor_fl[code]: "floor_fl (Float m e) = (if 0 \ e then Float m e else Float (m div (power_int 2 ((-e)))) 0)" by (simp add: Float.compute_floor_fl power_int_def) end