/- Copyright (c) 2014 Jeremy Avigad. All rights reserved. Released under Apache 2.0 license as described in the file LICENSE. Author: Jeremy Avigad, Leonardo de Moura -/ import logic algebra.binary open eq.ops binary definition set [reducible] (X : Type) := X → Prop namespace set variable {X : Type} /- membership and subset -/ definition mem [reducible] (x : X) (a : set X) := a x infix `∈` := mem notation a ∉ b := ¬ mem a b theorem setext {a b : set X} (H : ∀x, x ∈ a ↔ x ∈ b) : a = b := funext (take x, propext (H x)) definition subset (a b : set X) := ∀⦃x⦄, x ∈ a → x ∈ b infix `⊆` := subset theorem subset.refl (a : set X) : a ⊆ a := take x, assume H, H theorem subset.trans (a b c : set X) (subab : a ⊆ b) (subbc : b ⊆ c) : a ⊆ c := take x, assume ax, subbc (subab ax) theorem subset.antisymm {a b : set X} (h₁ : a ⊆ b) (h₂ : b ⊆ a) : a = b := setext (λ x, iff.intro (λ ina, h₁ ina) (λ inb, h₂ inb)) -- an alterantive name theorem eq_of_subset_of_subset {a b : set X} (h₁ : a ⊆ b) (h₂ : b ⊆ a) : a = b := subset.antisymm h₁ h₂ definition strict_subset (a b : set X) := a ⊆ b ∧ a ≠ b infix `⊂`:50 := strict_subset theorem strict_subset.irrefl (a : set X) : ¬ a ⊂ a := assume h, absurd rfl (and.elim_right h) /- bounded quantification -/ abbreviation bounded_forall (a : set X) (P : X → Prop) := ∀⦃x⦄, x ∈ a → P x notation `forallb` binders `∈` a `,` r:(scoped:1 P, P) := bounded_forall a r notation `∀₀` binders `∈` a `,` r:(scoped:1 P, P) := bounded_forall a r abbreviation bounded_exists (a : set X) (P : X → Prop) := ∃⦃x⦄, x ∈ a ∧ P x notation `existsb` binders `∈` a `,` r:(scoped:1 P, P) := bounded_exists a r notation `∃₀` binders `∈` a `,` r:(scoped:1 P, P) := bounded_exists a r /- empty set -/ definition empty [reducible] : set X := λx, false notation `∅` := empty theorem not_mem_empty (x : X) : ¬ (x ∈ ∅) := assume H : x ∈ ∅, H theorem mem_empty_eq (x : X) : x ∈ ∅ = false := rfl theorem eq_empty_of_forall_not_mem {s : set X} (H : ∀ x, x ∉ s) : s = ∅ := setext (take x, iff.intro (assume xs, absurd xs (H x)) (assume xe, absurd xe !not_mem_empty)) theorem empty_subset (s : set X) : ∅ ⊆ s := take x, assume H, false.elim H theorem eq_empty_of_subset_empty {s : set X} (H : s ⊆ ∅) : s = ∅ := subset.antisymm H (empty_subset s) theorem subset_empty_iff (s : set X) : s ⊆ ∅ ↔ s = ∅ := iff.intro eq_empty_of_subset_empty (take xeq, by rewrite xeq; apply subset.refl ∅) /- universal set -/ definition univ : set X := λx, true theorem mem_univ (x : X) : x ∈ univ := trivial theorem mem_univ_eq (x : X) : x ∈ univ = true := rfl theorem empty_ne_univ [h : inhabited X] : (empty : set X) ≠ univ := assume H : empty = univ, absurd (mem_univ (inhabited.value h)) (eq.rec_on H (not_mem_empty _)) /- union -/ definition union [reducible] (a b : set X) : set X := λx, x ∈ a ∨ x ∈ b notation a ∪ b := union a b theorem mem_union (x : X) (a b : set X) : x ∈ a ∪ b ↔ x ∈ a ∨ x ∈ b := !iff.refl theorem mem_union_eq (x : X) (a b : set X) : x ∈ a ∪ b = (x ∈ a ∨ x ∈ b) := rfl theorem union_self (a : set X) : a ∪ a = a := setext (take x, !or_self) theorem union_empty (a : set X) : a ∪ ∅ = a := setext (take x, !or_false) theorem empty_union (a : set X) : ∅ ∪ a = a := setext (take x, !false_or) theorem union.comm (a b : set X) : a ∪ b = b ∪ a := setext (take x, or.comm) theorem union.assoc (a b c : set X) : (a ∪ b) ∪ c = a ∪ (b ∪ c) := setext (take x, or.assoc) theorem union.left_comm (s₁ s₂ s₃ : set X) : s₁ ∪ (s₂ ∪ s₃) = s₂ ∪ (s₁ ∪ s₃) := !left_comm union.comm union.assoc s₁ s₂ s₃ theorem union.right_comm (s₁ s₂ s₃ : set X) : (s₁ ∪ s₂) ∪ s₃ = (s₁ ∪ s₃) ∪ s₂ := !right_comm union.comm union.assoc s₁ s₂ s₃ /- intersection -/ definition inter [reducible] (a b : set X) : set X := λx, x ∈ a ∧ x ∈ b notation a ∩ b := inter a b theorem mem_inter (x : X) (a b : set X) : x ∈ a ∩ b ↔ x ∈ a ∧ x ∈ b := !iff.refl theorem mem_inter_eq (x : X) (a b : set X) : x ∈ a ∩ b = (x ∈ a ∧ x ∈ b) := rfl theorem inter_self (a : set X) : a ∩ a = a := setext (take x, !and_self) theorem inter_empty (a : set X) : a ∩ ∅ = ∅ := setext (take x, !and_false) theorem empty_inter (a : set X) : ∅ ∩ a = ∅ := setext (take x, !false_and) theorem inter.comm (a b : set X) : a ∩ b = b ∩ a := setext (take x, !and.comm) theorem inter.assoc (a b c : set X) : (a ∩ b) ∩ c = a ∩ (b ∩ c) := setext (take x, !and.assoc) theorem inter.left_comm (s₁ s₂ s₃ : set X) : s₁ ∩ (s₂ ∩ s₃) = s₂ ∩ (s₁ ∩ s₃) := !left_comm inter.comm inter.assoc s₁ s₂ s₃ theorem inter.right_comm (s₁ s₂ s₃ : set X) : (s₁ ∩ s₂) ∩ s₃ = (s₁ ∩ s₃) ∩ s₂ := !right_comm inter.comm inter.assoc s₁ s₂ s₃ theorem inter_univ (a : set X) : a ∩ univ = a := setext (take x, !and_true) theorem univ_inter (a : set X) : univ ∩ a = a := setext (take x, !true_and) /- distributivity laws -/ theorem inter.distrib_left (s t u : set X) : s ∩ (t ∪ u) = (s ∩ t) ∪ (s ∩ u) := setext (take x, !and.left_distrib) theorem inter.distrib_right (s t u : set X) : (s ∪ t) ∩ u = (s ∩ u) ∪ (t ∩ u) := setext (take x, !and.right_distrib) theorem union.distrib_left (s t u : set X) : s ∪ (t ∩ u) = (s ∪ t) ∩ (s ∪ u) := setext (take x, !or.left_distrib) theorem union.distrib_right (s t u : set X) : (s ∩ t) ∪ u = (s ∪ u) ∩ (t ∪ u) := setext (take x, !or.right_distrib) /- set-builder notation -/ -- {x : X | P} definition set_of (P : X → Prop) : set X := P notation `{` binder `|` r:(scoped:1 P, set_of P) `}` := r -- {x ∈ s | P} definition filter (P : X → Prop) (s : set X) : set X := λx, x ∈ s ∧ P x notation `{` binder ∈ s `|` r:(scoped:1 p, filter p s) `}` := r -- '{x, y, z} definition insert (x : X) (a : set X) : set X := {y : X | y = x ∨ y ∈ a} notation `'{`:max a:(foldr `,` (x b, insert x b) ∅) `}`:0 := a /- filter -/ theorem eq_filter_of_subset {s t : set X} (ssubt : s ⊆ t) : s = {x ∈ t | x ∈ s} := setext (take x, iff.intro (suppose x ∈ s, and.intro (ssubt this) this) (suppose x ∈ {x ∈ t | x ∈ s}, and.right this)) /- set difference -/ definition diff (s t : set X) : set X := {x ∈ s | x ∉ t} infix `\`:70 := diff theorem mem_of_mem_diff {s t : set X} {x : X} (H : x ∈ s \ t) : x ∈ s := and.left H theorem not_mem_of_mem_diff {s t : set X} {x : X} (H : x ∈ s \ t) : x ∉ t := and.right H theorem mem_diff {s t : set X} {x : X} (H1 : x ∈ s) (H2 : x ∉ t) : x ∈ s \ t := and.intro H1 H2 theorem diff_eq (s t : set X) : s \ t = {x ∈ s | x ∉ t} := rfl theorem mem_diff_iff (s t : set X) (x : X) : x ∈ s \ t ↔ x ∈ s ∧ x ∉ t := !iff.refl theorem mem_diff_eq (s t : set X) (x : X) : x ∈ s \ t = (x ∈ s ∧ x ∉ t) := rfl /- large unions -/ section variables {I : Type} variable a : set I variable b : I → set X variable C : set (set X) definition Inter : set X := {x : X | ∀i, x ∈ b i} definition bInter : set X := {x : X | ∀₀ i ∈ a, x ∈ b i} definition sInter : set X := {x : X | ∀₀ c ∈ C, x ∈ c} definition Union : set X := {x : X | ∃i, x ∈ b i} definition bUnion : set X := {x : X | ∃₀ i ∈ a, x ∈ b i} definition sUnion : set X := {x : X | ∃₀ c ∈ C, x ∈ c} -- TODO: need notation for these end end set