/- Copyright (c) 2015 Floris van Doorn. All rights reserved. Released under Apache 2.0 license as described in the file LICENSE. Authors: Floris van Doorn, Egbert Rijke -/ import ..move_to_lib types.fin types.trunc open nat eq equiv sigma sigma.ops is_equiv is_trunc trunc prod fiber function is_conn namespace seq_colim definition seq_diagram [reducible] (A : ℕ → Type) : Type := Π⦃n⦄, A n → A (succ n) structure Seq_diagram : Type := (carrier : ℕ → Type) (struct : seq_diagram carrier) definition is_equiseq [reducible] {A : ℕ → Type} (f : seq_diagram A) : Type := forall (n : ℕ), is_equiv (@f n) structure Equi_seq : Type := (carrier : ℕ → Type) (maps : seq_diagram carrier) (prop : is_equiseq maps) protected abbreviation Mk [constructor] := Seq_diagram.mk attribute Seq_diagram.carrier [coercion] attribute Seq_diagram.struct [coercion] variables {A A' : ℕ → Type} (f : seq_diagram A) (f' : seq_diagram A') {n m k : ℕ} include f definition lrep {n m : ℕ} (H : n ≤ m) : A n → A m := begin induction H with m H fs, { exact id }, { exact @f m ∘ fs } end definition lrep_irrel_pathover {n m m' : ℕ} (H₁ : n ≤ m) (H₂ : n ≤ m') (p : m = m') (a : A n) : lrep f H₁ a =[p] lrep f H₂ a := apo (λm H, lrep f H a) !is_prop.elimo definition lrep_irrel {n m : ℕ} (H₁ H₂ : n ≤ m) (a : A n) : lrep f H₁ a = lrep f H₂ a := ap (λH, lrep f H a) !is_prop.elim definition lrep_eq_transport {n m : ℕ} (H : n ≤ m) (p : n = m) (a : A n) : lrep f H a = transport A p a := begin induction p, exact lrep_irrel f H (nat.le_refl n) a end definition lrep_irrel2 {n m : ℕ} (H₁ H₂ : n ≤ m) (a : A n) : lrep_irrel f (le.step H₁) (le.step H₂) a = ap (@f m) (lrep_irrel f H₁ H₂ a) := begin have H₁ = H₂, from !is_prop.elim, induction this, refine ap02 _ !is_prop_elim_self ⬝ _ ⬝ ap02 _(ap02 _ !is_prop_elim_self⁻¹), reflexivity end definition lrep_eq_lrep_irrel {n m m' : ℕ} (H₁ : n ≤ m) (H₂ : n ≤ m') (a₁ a₂ : A n) (p : m = m') : (lrep f H₁ a₁ = lrep f H₁ a₂) ≃ (lrep f H₂ a₁ = lrep f H₂ a₂) := equiv_apd011 (λm H, lrep f H a₁ = lrep f H a₂) (is_prop.elimo p H₁ H₂) definition lrep_eq_lrep_irrel_natural {n m m' : ℕ} {H₁ : n ≤ m} (H₂ : n ≤ m') {a₁ a₂ : A n} (p : m = m') (q : lrep f H₁ a₁ = lrep f H₁ a₂) : lrep_eq_lrep_irrel f (le.step H₁) (le.step H₂) a₁ a₂ (ap succ p) (ap (@f m) q) = ap (@f m') (lrep_eq_lrep_irrel f H₁ H₂ a₁ a₂ p q) := begin esimp [lrep_eq_lrep_irrel], symmetry, refine fn_tro_eq_tro_fn2 _ (λa₁ a₂, ap (@f _)) q ⬝ _, refine ap (λx, x ▸o _) (@is_prop.elim _ _ _ _), apply is_trunc_pathover end definition is_equiv_lrep [constructor] [Hf : is_equiseq f] {n m : ℕ} (H : n ≤ m) : is_equiv (lrep f H) := begin induction H with m H Hlrepf, { apply is_equiv_id }, { exact is_equiv_compose (@f _) (lrep f H) _ _ }, end local attribute is_equiv_lrep [instance] definition lrep_back [reducible] [Hf : is_equiseq f] {n m : ℕ} (H : n ≤ m) : A m → A n := (lrep f H)⁻¹ᶠ section generalized_lrep -- definition lrep_pathover_lrep0 [reducible] (k : ℕ) (a : A 0) : lrep f k a =[nat.zero_add k] lrep0 f k a := -- begin induction k with k p, constructor, exact pathover_ap A succ (apo f p) end /- lreplace le_of_succ_le with this -/ definition lrep_f {n m : ℕ} (H : succ n ≤ m) (a : A n) : lrep f H (f a) = lrep f (le_of_succ_le H) a := begin induction H with m H p, { reflexivity }, { exact ap (@f m) p } end definition lrep_lrep {n m k : ℕ} (H1 : n ≤ m) (H2 : m ≤ k) (a : A n) : lrep f H2 (lrep f H1 a) = lrep f (nat.le_trans H1 H2) a := begin induction H2 with k H2 p, { reflexivity }, { exact ap (@f k) p } end -- -- this should be a squareover -- definition lrep_lrep_succ (k l : ℕ) (a : A n) : -- lrep_lrep f k (succ l) a = change_path (idontwanttoprovethis n l k) -- (lrep_f f k (lrep f l a) ⬝o -- lrep_lrep f (succ k) l a ⬝o -- pathover_ap _ (λz, n + z) (apd (λz, lrep f z a) (succ_add l k)⁻¹ᵖ)) := -- begin -- induction k with k IH, -- { constructor}, -- { exact sorry} -- end definition f_lrep {n m : ℕ} (H : n ≤ m) (a : A n) : f (lrep f H a) = lrep f (le.step H) a := idp definition rep (m : ℕ) (a : A n) : A (n + m) := lrep f (le_add_right n m) a definition rep0 (m : ℕ) (a : A 0) : A m := lrep f (zero_le m) a definition rep_pathover_rep0 {n : ℕ} (a : A 0) : rep f n a =[nat.zero_add n] rep0 f n a := !lrep_irrel_pathover -- definition old_rep (m : ℕ) (a : A n) : A (n + m) := -- by induction m with m y; exact a; exact f y -- definition rep_eq_old_rep (m : ℕ) (a : A n) : rep f m a = old_rep f m a := -- by induction m with m p; reflexivity; exact ap (@f _) p definition rep_f (k : ℕ) (a : A n) : pathover A (rep f k (f a)) (succ_add n k) (rep f (succ k) a) := begin induction k with k IH, { constructor }, { unfold [succ_add], apply pathover_ap, exact apo f IH} end definition rep_rep (k l : ℕ) (a : A n) : pathover A (rep f k (rep f l a)) (nat.add_assoc n l k) (rep f (l + k) a) := begin induction k with k IH, { constructor}, { apply pathover_ap, exact apo f IH} end variables {f f'} definition is_trunc_fun_lrep (k : ℕ₋₂) (H : n ≤ m) (H2 : Πn, is_trunc_fun k (@f n)) : is_trunc_fun k (lrep f H) := begin induction H with m H IH, apply is_trunc_fun_id, exact is_trunc_fun_compose k (H2 m) IH end definition is_conn_fun_lrep (k : ℕ₋₂) (H : n ≤ m) (H2 : Πn, is_conn_fun k (@f n)) : is_conn_fun k (lrep f H) := begin induction H with m H IH, apply is_conn_fun_id, exact is_conn_fun_compose k (H2 m) IH end definition lrep_natural (τ : Π⦃n⦄, A n → A' n) (p : Π⦃n⦄ (a : A n), τ (f a) = f' (τ a)) {n m : ℕ} (H : n ≤ m) (a : A n) : τ (lrep f H a) = lrep f' H (τ a) := begin induction H with m H IH, reflexivity, exact p (lrep f H a) ⬝ ap (@f' m) IH end definition rep_natural (τ : Π⦃n⦄, A n → A' n) (p : Π⦃n⦄ (a : A n), τ (f a) = f' (τ a)) {n : ℕ} (k : ℕ) (a : A n) : τ (rep f k a) = rep f' k (τ a) := lrep_natural τ p _ a definition rep0_natural (τ : Π⦃n⦄, A n → A' n) (p : Π⦃n⦄ (a : A n), τ (f a) = f' (τ a)) (k : ℕ) (a : A 0) : τ (rep0 f k a) = rep0 f' k (τ a) := lrep_natural τ p _ a variables (f f') end generalized_lrep section shift definition shift_diag [unfold_full] : seq_diagram (λn, A (succ n)) := λn a, f a definition kshift_diag [unfold_full] (k : ℕ) : seq_diagram (λn, A (k + n)) := λn a, f a definition kshift_diag' [unfold_full] (k : ℕ) : seq_diagram (λn, A (n + k)) := λn a, transport A (succ_add n k)⁻¹ (f a) definition lrep_kshift_diag {n m k : ℕ} (H : m ≤ k) (a : A (n + m)) : lrep (kshift_diag f n) H a = lrep f (nat.add_le_add_left2 H n) a := by induction H with k H p; reflexivity; exact ap (@f _) p end shift section constructions omit f definition constant_seq (X : Type) : seq_diagram (λ n, X) := λ n x, x definition seq_diagram_arrow_left [unfold_full] (X : Type) : seq_diagram (λn, X → A n) := λn g x, f (g x) definition seq_diagram_prod [unfold_full] : seq_diagram (λn, A n × A' n) := λn, prod_functor (@f n) (@f' n) open fin definition seq_diagram_fin [unfold_full] : seq_diagram fin := lift_succ definition id0_seq [unfold_full] (a₁ a₂ : A 0) : ℕ → Type := λ k, rep0 f k a₁ = rep0 f k a₂ definition id0_seq_diagram [unfold_full] (a₁ a₂ : A 0) : seq_diagram (id0_seq f a₁ a₂) := λ (k : ℕ) (p : rep0 f k a₁ = rep0 f k a₂), ap (@f k) p definition id_seq [unfold_full] (n : ℕ) (a₁ a₂ : A n) : ℕ → Type := λ k, rep f k a₁ = rep f k a₂ definition id_seq_diagram [unfold_full] (n : ℕ) (a₁ a₂ : A n) : seq_diagram (id_seq f n a₁ a₂) := λ (k : ℕ) (p : rep f k a₁ = rep f k a₂), ap (@f (n + k)) p definition trunc_diagram [unfold_full] (k : ℕ₋₂) (f : seq_diagram A) : seq_diagram (λn, trunc k (A n)) := λn, trunc_functor k (@f n) end constructions section over variable {A} variable (P : Π⦃n⦄, A n → Type) definition seq_diagram_over : Type := Π⦃n⦄ {a : A n}, P a → P (f a) definition weakened_sequence [unfold_full] : seq_diagram_over f (λn a, A' n) := λn a a', f' a' definition id0_seq_diagram_over [unfold_full] (a₀ : A 0) : seq_diagram_over f (λn a, rep0 f n a₀ = a) := λn a p, ap (@f n) p variable (g : seq_diagram_over f P) variables {f P} definition seq_diagram_of_over [unfold_full] {n : ℕ} (a : A n) : seq_diagram (λk, P (rep f k a)) := λk p, g p definition seq_diagram_sigma [unfold 6] : seq_diagram (λn, Σ(x : A n), P x) := λn v, ⟨f v.1, g v.2⟩ variables (f P) theorem rep_f_equiv [constructor] (a : A n) (k : ℕ) : P (lrep f (le_add_right (succ n) k) (f a)) ≃ P (lrep f (le_add_right n (succ k)) a) := equiv_apd011 P (rep_f f k a) theorem rep_rep_equiv [constructor] (a : A n) (k l : ℕ) : P (rep f (l + k) a) ≃ P (rep f k (rep f l a)) := (equiv_apd011 P (rep_rep f k l a))⁻¹ᵉ end over omit f -- do we need to generalize this to the case where the bottom sequence consists of equivalences? definition seq_diagram_pi {X : Type} {A : X → ℕ → Type} (g : Π⦃x n⦄, A x n → A x (succ n)) : seq_diagram (λn, Πx, A x n) := λn f x, g (f x) variables {f f'} definition seq_diagram_over_fiber (g : Π⦃n⦄, A' n → A n) (p : Π⦃n⦄ (a : A' n), g (f' a) = f (g a)) : seq_diagram_over f (λn, fiber (@g n)) := λk a, fiber_functor (@f' k) (@f k) (@p k) idp definition seq_diagram_fiber (g : Π⦃n⦄, A' n → A n) (p : Π⦃n⦄ (a : A' n), g (f' a) = f (g a)) {n : ℕ} (a : A n) : seq_diagram (λk, fiber (@g (n + k)) (rep f k a)) := seq_diagram_of_over (seq_diagram_over_fiber g p) a definition seq_diagram_fiber0 (g : Π⦃n⦄, A' n → A n) (p : Π⦃n⦄ (a : A' n), g (f' a) = f (g a)) (a : A 0) : seq_diagram (λk, fiber (@g k) (rep0 f k a)) := λk, fiber_functor (@f' k) (@f k) (@p k) idp end seq_colim