lean2/hott/homotopy/cellcomplex.hlean

50 lines
1.7 KiB
Text
Raw Normal View History

/-
Copyright (c) 2015 Ulrik Buchholtz. All rights reserved.
Released under Apache 2.0 license as described in the file LICENSE.
Authors: Ulrik Buchholtz
-/
import types.trunc homotopy.sphere hit.pushout
open eq is_trunc is_equiv nat equiv trunc prod pushout sigma unit pointed
-- where should this be?
definition family : Type := ΣX, X → Type
namespace cellcomplex
/-
define by recursion on
both the type of fdccs of dimension n
and the realization map fdcc n → Type
in other words, we define a function
fdcc : → family
an alternative to the approach here (perhaps necessary) is to
define relative cell complexes relative to a type A, and then use
spherical indexing, so a -1-dimensional relative cell complex is
just star : unit with realization A
-/
definition fdcc_family [reducible] : → family :=
nat.rec
-- a zero-dimensional cell complex is just an set
-- with realization the identity map
⟨Set , λA, trunctype.carrier A⟩
(λn fdcc_family_n, -- sigma.rec (λ fdcc_n realize_n,
/- a (succ n)-dimensional cell complex is a triple of
an n-dimensional cell complex X, an set of (succ n)-cells A,
and an attaching map f : A × sphere n → |X| -/
⟨Σ X : pr1 fdcc_family_n , Σ A : Set, A × sphere n → pr2 fdcc_family_n X ,
/- the realization of such is the pushout of f with
canonical map A × sphere n → unit -/
sigma.rec (λX , sigma.rec (λA f, pushout (λx , star) f))
⟩)
definition fdcc (n : ) : Type := pr1 (fdcc_family n)
definition cell : Πn, fdcc n → Set :=
nat.cases (λA, A) (λn T, pr1 (pr2 T))
end cellcomplex