diff options
author | Josh Chen | 2018-05-29 12:28:13 +0200 |
---|---|---|
committer | Josh Chen | 2018-05-29 12:28:13 +0200 |
commit | 607c3971e08d1ded22bd9f1cabdd309653af1248 (patch) | |
tree | 256da6d96c3310c72a4fa2e87043382c77440cf0 | |
parent | 120879c099a2fb71e67a41a1c852c5db65e9eb4f (diff) |
More rigorous rules for Product type. Propositions and proofs all working, but have to think about maybe relaxing the computation rule, or else automating the currying of dependent functions.
Diffstat (limited to '')
-rw-r--r-- | HoTT.thy | 29 | ||||
-rw-r--r-- | HoTT_Theorems.thy | 67 |
2 files changed, 72 insertions, 24 deletions
@@ -54,24 +54,27 @@ subsubsection \<open>Dependent function/product\<close> consts Prod :: "[Term, (Term \<Rightarrow> Term)] \<Rightarrow> Term" + lambda :: "[Term, (Term \<Rightarrow> Term)] \<Rightarrow> Term" syntax - "_Prod" :: "[idt, Term, Term] \<Rightarrow> Term" ("(3\<Prod>_:_./ _)" 10) + "_Prod" :: "[idt, Term, Term] \<Rightarrow> Term" ("(3\<Prod>_:_./ _)" 10) + "__lambda" :: "[idt, Term, Term] \<Rightarrow> Term" ("(3\<^bold>\<lambda>_:_./ _)" 10) translations "\<Prod>x:A. B" \<rightleftharpoons> "CONST Prod A (\<lambda>x. B)" -(* The above syntax translation binds the x in the expression B *) + "\<^bold>\<lambda>x:A. b" \<rightleftharpoons> "CONST lambda A (\<lambda>x. b)" +(* The above syntax translations bind the x in the expressions B, b. *) abbreviation Function :: "[Term, Term] \<Rightarrow> Term" (infixr "\<rightarrow>" 30) where "A\<rightarrow>B \<equiv> \<Prod>_:A. B" axiomatization - lambda :: "(Term \<Rightarrow> Term) \<Rightarrow> Term" (binder "\<^bold>\<lambda>" 10) and - appl :: "[Term, Term] \<Rightarrow> Term" ("(3_`/_)" [10, 10] 60) + appl :: "[Term, Term] \<Rightarrow> Term" (infixl "`" 60) where Prod_form: "\<And>(A::Term) (B::Term \<Rightarrow> Term). \<lbrakk>A : U; B : A \<rightarrow> U\<rbrakk> \<Longrightarrow> \<Prod>x:A. B(x) : U" and - Prod_intro [intro]: "\<And>(A::Term) (B::Term \<Rightarrow> Term) (b::Term \<Rightarrow> Term). \<lbrakk>A : U; B : A \<rightarrow> U; \<And>x::Term. x : A \<Longrightarrow> b(x) : B(x)\<rbrakk> \<Longrightarrow> \<^bold>\<lambda>x. b(x) : \<Prod>x:A. B(x)" and + Prod_intro [intro]: "\<And>(A::Term) (B::Term \<Rightarrow> Term) (b::Term \<Rightarrow> Term). \<lbrakk>A : U; B : A \<rightarrow> U; \<And>x::Term. x : A \<Longrightarrow> b(x) : B(x)\<rbrakk> \<Longrightarrow> \<^bold>\<lambda>x:A. b(x) : \<Prod>x:A. B(x)" and Prod_elim [elim]: "\<And>(A::Term) (B::Term \<Rightarrow> Term) (f::Term) (a::Term). \<lbrakk>f : \<Prod>x:A. B(x); a : A\<rbrakk> \<Longrightarrow> f`a : B(a)" and - Prod_comp [simp]: "\<And>(b::Term \<Rightarrow> Term) (a::Term). (\<^bold>\<lambda>x. b(x))`a \<equiv> b(a)" and - Prod_uniq [simp]: "\<And>(A::Term) (B::Term \<Rightarrow> Term) (f::Term). f : \<Prod>x:A. B(x) \<Longrightarrow> \<^bold>\<lambda>x. (f`x) \<equiv> f" + Prod_comp [simp]: "\<And>(A::Term) (B::Term \<Rightarrow> Term) (b::Term \<Rightarrow> Term) (a::Term). \<lbrakk>A : U; B : A \<rightarrow> U; \<And>x::Term. x : A \<Longrightarrow> b(x) : B(x); a : A\<rbrakk> \<Longrightarrow> (\<^bold>\<lambda>x:A. b(x))`a \<equiv> b(a)" and + Prod_uniq [simp]: "\<And>(A::Term) (B::Term \<Rightarrow> Term) (f::Term). f : \<Prod>x:A. B(x) \<Longrightarrow> \<^bold>\<lambda>x:A. (f`x) \<equiv> f" +(* Thinking about the premises for the computation rule... they make simplification rather cumbersome, should I remove them? Would this potentially result in logical problems with being able to state untrue statements? (But probably not prove them?) *) text "Note that the syntax \<open>\<^bold>\<lambda>\<close> (bold lambda) used for dependent functions clashes with the proof term syntax (cf. \<section>2.5.2 of the Isabelle/Isar Implementation)." @@ -90,17 +93,19 @@ abbreviation Pair :: "[Term, Term] \<Rightarrow> Term" (infixr "\<times>" 50) where "A\<times>B \<equiv> \<Sum>_:A. B" axiomatization - pair :: "[Term, Term] \<Rightarrow> Term" ("(1'(_,/ _'))") and + pair :: "[Term, Term] \<Rightarrow> Term" ("(1'(_,/ _'))") and indSum :: "[Term \<Rightarrow> Term, Term \<Rightarrow> Term, Term] \<Rightarrow> Term" where Sum_form: "\<And>(A::Term) (B::Term \<Rightarrow> Term). \<lbrakk>A : U; B: A \<rightarrow> U\<rbrakk> \<Longrightarrow> \<Sum>x:A. B(x) : U" and - Sum_intro: "\<And>(A::Term) (B::Term \<Rightarrow> Term) (a::Term) (b::Term). \<lbrakk>A : U; B: A \<rightarrow> U; a : A; b : B(a)\<rbrakk> \<Longrightarrow> (a, b) : \<Sum>x:A. B(x)" and - Sum_elim: "\<And>(A::Term) (B::Term \<Rightarrow> Term) (C::Term \<Rightarrow> Term) (f::Term \<Rightarrow> Term) (p::Term). \<lbrakk>A : U; B: A \<rightarrow> U; C: \<Sum>x:A. B(x) \<rightarrow> U; \<And>x y::Term. \<lbrakk>x : A; y : B(x)\<rbrakk> \<Longrightarrow> f((x,y)) : C((x,y)); p : \<Sum>x:A. B(x)\<rbrakk> \<Longrightarrow> (indSum C f p) : C(p)" - Sum_comp: "" + Sum_intro [intro]: "\<And>(A::Term) (B::Term \<Rightarrow> Term) (a::Term) (b::Term). \<lbrakk>A : U; B: A \<rightarrow> U; a : A; b : B(a)\<rbrakk> \<Longrightarrow> (a, b) : \<Sum>x:A. B(x)" and + Sum_elim [elim]: "\<And>(A::Term) (B::Term \<Rightarrow> Term) (C::Term \<Rightarrow> Term) (f::Term \<Rightarrow> Term) (p::Term). \<lbrakk>A : U; B: A \<rightarrow> U; C: \<Sum>x:A. B(x) \<rightarrow> U; \<And>x y::Term. \<lbrakk>x : A; y : B(x)\<rbrakk> \<Longrightarrow> f((x,y)) : C((x,y)); p : \<Sum>x:A. B(x)\<rbrakk> \<Longrightarrow> (indSum C f p) : C(p)" and + Sum_comp [simp]: "\<And>(A::Term) (B::Term \<Rightarrow> Term) (C::Term \<Rightarrow> Term) (f::Term \<Rightarrow> Term) (a::Term) (b::Term). \<lbrakk>A : U; B: A \<rightarrow> U; C: \<Sum>x:A. B(x) \<rightarrow> U; \<And>x y::Term. \<lbrakk>x : A; y : B(x)\<rbrakk> \<Longrightarrow> f((x,y)) : C((x,y)); a : A; b : B(a)\<rbrakk> \<Longrightarrow> (indSum C f (a,b)) \<equiv> f((a,b))" -text "A choice had to be made for the elimination rule: we formalize the function \<open>f\<close> taking \<open>a : A\<close> and \<open>b : B(x)\<close> and returning \<open>C((a,b))\<close> as a meta-lambda \<open>f::Term \<Rightarrow> Term\<close> instead of an object dependent function \<open>f : \<Prod>x:A. B(x)\<close>. +text "A choice had to be made for the elimination rule: we formalize the function \<open>f\<close> taking \<open>a : A\<close> and \<open>b : B(x)\<close> and returning \<open>C((a,b))\<close> as a meta level \<open>f::Term \<Rightarrow> Term\<close> instead of an object logic dependent function \<open>f : \<Prod>x:A. B(x)\<close>. However we should be able to later show the equivalence of the formalizations." + + \<comment> \<open>Projection onto first component\<close> (* definition proj1 :: "Term \<Rightarrow> Term \<Rightarrow> Term" ("(proj1\<langle>_,_\<rangle>)") where diff --git a/HoTT_Theorems.thy b/HoTT_Theorems.thy index 33b0957..d83a08c 100644 --- a/HoTT_Theorems.thy +++ b/HoTT_Theorems.thy @@ -6,23 +6,25 @@ text "A bunch of theorems and other statements for sanity-checking, as well as t Things that *should* be automated: \<bullet> Checking that \<open>A\<close> is a well-formed type, when writing things like \<open>x : A\<close> and \<open>A : U\<close>. -" + \<bullet> Checking that the argument to a (dependent/non-dependent) function matches the type? Also the arguments to a pair?" \<comment> \<open>Turn on trace for unification and the simplifier, for debugging.\<close> -declare[[unify_trace_simp, unify_trace_types, simp_trace]] +declare[[unify_trace_simp, unify_trace_types, simp_trace, simp_trace_depth_limit=2]] section \<open>Functions\<close> +subsection \<open>Typing functions\<close> + text "Declaring \<open>Prod_intro\<close> with the \<open>intro\<close> attribute (in HoTT.thy) enables \<open>standard\<close> to prove the following." -lemma id_function: "A : U \<Longrightarrow> \<^bold>\<lambda>x. x : A\<rightarrow>A" .. +lemma id_function: "A : U \<Longrightarrow> \<^bold>\<lambda>x:A. x : A\<rightarrow>A" .. text "Here is the same result, stated and proved differently. The standard method invoked after the keyword \<open>proof\<close> is applied to the goal \<open>\<^bold>\<lambda>x. x: A\<rightarrow>A\<close>, and so we need to show the prover how to continue, as opposed to the previous lemma." lemma assumes "A : U" - shows "\<^bold>\<lambda>x. x : A\<rightarrow>A" + shows "\<^bold>\<lambda>x:A. x : A\<rightarrow>A" proof show "A : U" using assms . show "\<lambda>x. A : A \<rightarrow> U" using assms .. @@ -31,29 +33,29 @@ qed text "Note that there is no provision for declaring the type of bound variables outside of the scope of a lambda expression. More generally, we cannot write an assumption stating 'Let \<open>x\<close> be a variable of type \<open>A\<close>'." -proposition "\<lbrakk>A : U; A \<equiv> B\<rbrakk> \<Longrightarrow> \<^bold>\<lambda>x. x : B\<rightarrow>A" +proposition "\<lbrakk>A : U; A \<equiv> B\<rbrakk> \<Longrightarrow> \<^bold>\<lambda>x:A. x : B\<rightarrow>A" proof - assume 1: "A : U" and 2: "A \<equiv> B" - from id_function[OF 1] have 3: "\<^bold>\<lambda>x. x : A\<rightarrow>A" . + from id_function[OF 1] have 3: "\<^bold>\<lambda>x:A. x : A\<rightarrow>A" . from 2 have "A\<rightarrow>A \<equiv> B\<rightarrow>A" by simp - with 3 show "\<^bold>\<lambda>x. x : B\<rightarrow>A" .. + with 3 show "\<^bold>\<lambda>x:A. x : B\<rightarrow>A" .. qed text "It is instructive to try to prove \<open>\<lbrakk>A : U; B : U\<rbrakk> \<Longrightarrow> \<^bold>\<lambda>x. \<^bold>\<lambda>y. x : A\<rightarrow>B\<rightarrow>A\<close>. First we prove an intermediate step." -lemma constant_function: "\<lbrakk>A : U; B : U; x : A\<rbrakk> \<Longrightarrow> \<^bold>\<lambda>y. x : B\<rightarrow>A" .. +lemma constant_function: "\<lbrakk>A : U; B : U; x : A\<rbrakk> \<Longrightarrow> \<^bold>\<lambda>y:B. x : B\<rightarrow>A" .. text "And now the actual result:" proposition assumes 1: "A : U" and 2: "B : U" - shows "\<^bold>\<lambda>x. \<^bold>\<lambda>y. x : A\<rightarrow>B\<rightarrow>A" + shows "\<^bold>\<lambda>x:A. \<^bold>\<lambda>y:B. x : A\<rightarrow>B\<rightarrow>A" proof show "A : U" using assms(1) . - show "\<And>x. x : A \<Longrightarrow> \<^bold>\<lambda>y. x : B \<rightarrow> A" using assms by (rule constant_function) + show "\<And>x. x : A \<Longrightarrow> \<^bold>\<lambda>y:B. x : B \<rightarrow> A" using assms by (rule constant_function) from assms have "B \<rightarrow> A : U" by (rule Prod_formation) then show "\<lambda>x. B \<rightarrow> A: A \<rightarrow> U" using assms(1) by (rule constant_type_family) @@ -61,13 +63,54 @@ qed text "Maybe a nicer way to write it:" -proposition "\<lbrakk>A : U; B: U\<rbrakk> \<Longrightarrow> \<^bold>\<lambda>x. \<^bold>\<lambda>y. x : A\<rightarrow>B\<rightarrow>A" +proposition alternating_function: "\<lbrakk>A : U; B: U\<rbrakk> \<Longrightarrow> \<^bold>\<lambda>x:A. \<^bold>\<lambda>y:B. x : A\<rightarrow>B\<rightarrow>A" proof fix x - show "\<lbrakk>A : U; B : U; x : A\<rbrakk> \<Longrightarrow> \<^bold>\<lambda>y. x : B \<rightarrow> A" by (rule constant_function) + show "\<lbrakk>A : U; B : U; x : A\<rbrakk> \<Longrightarrow> \<^bold>\<lambda>y:B. x : B \<rightarrow> A" by (rule constant_function) show "\<lbrakk>A : U; B : U\<rbrakk> \<Longrightarrow> B\<rightarrow>A : U" by (rule Prod_formation) qed +subsection \<open>Function application\<close> + +lemma "\<lbrakk>A : U; a : A\<rbrakk> \<Longrightarrow> (\<^bold>\<lambda>x:A. x)`a \<equiv> a" by simp + +lemma + assumes + "A:U" and + "B:U" and + "a:A" and + "b:B" + shows "(\<^bold>\<lambda>x:A. \<^bold>\<lambda>y:B. x)`a`b \<equiv> a" +proof - + have "(\<^bold>\<lambda>x:A. \<^bold>\<lambda>y:B. x)`a \<equiv> \<^bold>\<lambda>y:B. a" + proof (rule Prod_comp[of A "\<lambda>_. B\<rightarrow>A"]) + have "B \<rightarrow> A : U" using constant_type_family[OF assms(1) assms(2)] assms(2) by (rule Prod_formation) + then show "\<lambda>x. B \<rightarrow> A: A \<rightarrow> U" using assms(1) by (rule constant_type_family[of "B\<rightarrow>A"]) + + show "\<And>x. x : A \<Longrightarrow> \<^bold>\<lambda>y:B. x : B \<rightarrow> A" using assms(2) assms(1) .. + show "A:U" using assms(1) . + show "a:A" using assms(3) . + qed (* Why do I need to do the above for the last two goals? Can't Isabelle do it automatically? *) + + then have "(\<^bold>\<lambda>x:A. \<^bold>\<lambda>y:B. x)`a`b \<equiv> (\<^bold>\<lambda>y:B. a)`b" by simp + + also have "(\<^bold>\<lambda>y:B. a)`b \<equiv> a" + proof (rule Prod_comp[of B "\<lambda>_. A"]) + show "\<lambda>y. A: B \<rightarrow> U" using assms(1) assms(2) by (rule constant_type_family) + show "\<And>y. y : B \<Longrightarrow> a : A" using assms(3) . + show "B:U" using assms(2) . + show "b:B" using assms(4) . + qed + + finally show "(\<^bold>\<lambda>x:A. \<^bold>\<lambda>y:B. x)`a`b \<equiv> a" . +qed + +text "Polymorphic identity function." + +consts Ui::Term +definition Id where "Id \<equiv> \<^bold>\<lambda>A:Ui. \<^bold>\<lambda>x:A. x" +(* Have to think about universes... *) + section \<open>Nats\<close> text "Here's a dumb proof that 2 is a natural number." |