Datasets:
Tasks:
Text Generation
Modalities:
Text
Sub-tasks:
language-modeling
Languages:
English
Size:
100K - 1M
License:
(* | |
File: Eval_Numeral.thy | |
Author: Manuel Eberl <manuel@pruvisto.org> | |
Evaluation of terms involving rational numerals with the simplifier. | |
*) | |
section \<open>Evaluating expressions with rational numerals\<close> | |
theory Eval_Numeral | |
imports | |
Complex_Main | |
begin | |
lemma real_numeral_to_Ratreal: | |
"(0::real) = Ratreal (Frct (0, 1))" | |
"(1::real) = Ratreal (Frct (1, 1))" | |
"(numeral x :: real) = Ratreal (Frct (numeral x, 1))" | |
"(1::int) = numeral Num.One" | |
by (simp_all add: rat_number_collapse) | |
lemma real_equals_code: "Ratreal x = Ratreal y \<longleftrightarrow> x = y" | |
by simp | |
lemma Rat_normalize_idempotent: "Rat.normalize (Rat.normalize x) = Rat.normalize x" | |
apply (cases "Rat.normalize x") | |
using Rat.normalize_stable[OF normalize_denom_pos normalize_coprime] apply auto | |
done | |
lemma uminus_pow_Numeral1: "(-(x::_::monoid_mult)) ^ Numeral1 = -x" by simp | |
lemmas power_numeral_simps = power_0 uminus_pow_Numeral1 power_minus_Bit0 power_minus_Bit1 | |
lemma Fract_normalize: "Fract (fst (Rat.normalize (x,y))) (snd (Rat.normalize (x,y))) = Fract x y" | |
by (rule quotient_of_inject) (simp add: quotient_of_Fract Rat_normalize_idempotent) | |
lemma Frct_add: "Frct (a, numeral b) + Frct (c, numeral d) = | |
Frct (Rat.normalize (a * numeral d + c * numeral b, numeral (b*d)))" | |
by (auto simp: rat_number_collapse Fract_normalize) | |
lemma Frct_uminus: "-(Frct (a,b)) = Frct (-a,b)" by simp | |
lemma Frct_diff: "Frct (a, numeral b) - Frct (c, numeral d) = | |
Frct (Rat.normalize (a * numeral d - c * numeral b, numeral (b*d)))" | |
by (auto simp: rat_number_collapse Fract_normalize) | |
lemma Frct_mult: "Frct (a, numeral b) * Frct (c, numeral d) = Frct (a*c, numeral (b*d))" | |
by simp | |
lemma Frct_inverse: "inverse (Frct (a, b)) = Frct (b, a)" by simp | |
lemma Frct_divide: "Frct (a, numeral b) / Frct (c, numeral d) = Frct (a*numeral d, numeral b * c)" | |
by simp | |
lemma Frct_pow: "Frct (a, numeral b) ^ c = Frct (a ^ c, numeral b ^ c)" | |
by (induction c) (simp_all add: rat_number_collapse) | |
lemma Frct_less: "Frct (a, numeral b) < Frct (c, numeral d) \<longleftrightarrow> a * numeral d < c * numeral b" | |
by simp | |
lemma Frct_le: "Frct (a, numeral b) \<le> Frct (c, numeral d) \<longleftrightarrow> a * numeral d \<le> c * numeral b" | |
by simp | |
lemma Frct_equals: "Frct (a, numeral b) = Frct (c, numeral d) \<longleftrightarrow> a * numeral d = c * numeral b" | |
apply (intro iffI antisym) | |
apply (subst Frct_le[symmetric], simp)+ | |
apply (subst Frct_le, simp)+ | |
done | |
lemma real_power_code: "(Ratreal x) ^ y = Ratreal (x ^ y)" by (simp add: of_rat_power) | |
lemmas real_arith_code = | |
real_plus_code real_minus_code real_times_code real_uminus_code real_inverse_code | |
real_divide_code real_power_code real_less_code real_less_eq_code real_equals_code | |
lemmas rat_arith_code = | |
Frct_add Frct_uminus Frct_diff Frct_mult Frct_inverse Frct_divide Frct_pow | |
Frct_less Frct_le Frct_equals | |
lemma gcd_numeral_red: "gcd (numeral x::int) (numeral y) = gcd (numeral y) (numeral x mod numeral y)" | |
by (fact gcd_red_int) | |
lemma divmod_one: | |
"divmod (Num.One) (Num.One) = (Numeral1, 0)" | |
"divmod (Num.One) (Num.Bit0 x) = (0, Numeral1)" | |
"divmod (Num.One) (Num.Bit1 x) = (0, Numeral1)" | |
"divmod x (Num.One) = (numeral x, 0)" | |
unfolding divmod_def by simp_all | |
lemmas divmod_numeral_simps = | |
div_0 div_by_0 mod_0 mod_by_0 | |
fst_divmod [symmetric] | |
snd_divmod [symmetric] | |
divmod_cancel | |
divmod_steps [simplified rel_simps if_True] divmod_trivial | |
rel_simps | |
lemma Suc_0_to_numeral: "Suc 0 = Numeral1" by simp | |
lemmas Suc_to_numeral = Suc_0_to_numeral Num.Suc_1 Num.Suc_numeral | |
lemma rat_powr: | |
"0 powr y = 0" | |
"x > 0 \<Longrightarrow> x powr Ratreal (Frct (0, Numeral1)) = Ratreal (Frct (Numeral1, Numeral1))" | |
"x > 0 \<Longrightarrow> x powr Ratreal (Frct (numeral a, Numeral1)) = x ^ numeral a" | |
"x > 0 \<Longrightarrow> x powr Ratreal (Frct (-numeral a, Numeral1)) = inverse (x ^ numeral a)" | |
by (simp_all add: rat_number_collapse powr_minus) | |
lemmas eval_numeral_simps = | |
real_numeral_to_Ratreal real_arith_code rat_arith_code Num.arith_simps | |
Rat.normalize_def fst_conv snd_conv gcd_0_int gcd_0_left_int gcd.bottom_right_bottom gcd.bottom_left_bottom | |
gcd_neg1_int gcd_neg2_int gcd_numeral_red zmod_numeral_Bit0 zmod_numeral_Bit1 power_numeral_simps | |
divmod_numeral_simps numeral_One [symmetric] Groups.Let_0 Num.Let_numeral Suc_to_numeral power_numeral | |
greaterThanLessThan_iff atLeastAtMost_iff atLeastLessThan_iff greaterThanAtMost_iff rat_powr | |
Num.pow.simps Num.sqr.simps Product_Type.split of_int_numeral of_int_neg_numeral of_nat_numeral | |
ML \<open> | |
signature EVAL_NUMERAL = | |
sig | |
val eval_numeral_tac : Proof.context -> int -> tactic | |
end | |
structure Eval_Numeral : EVAL_NUMERAL = | |
struct | |
fun eval_numeral_tac ctxt = | |
let | |
val ctxt' = put_simpset HOL_ss ctxt addsimps @{thms eval_numeral_simps} | |
in | |
SELECT_GOAL (SOLVE (Simplifier.simp_tac ctxt' 1)) | |
end | |
end | |
\<close> | |
lemma "21254387548659589512*314213523632464357453884361*2342523623324234*564327438587241734743* | |
12561712738645824362329316482973164398214286 powr 2 / | |
(1130246312978423123+231212374631082764842731842*122474378389424362347451251263) > | |
(12313244512931247243543279768645745929475829310651205623844::real)" | |
by (tactic \<open>Eval_Numeral.eval_numeral_tac @{context} 1\<close>) | |
end | |