Library UniMath.CategoryTheory.limits.graphs.kernels

Kernels defined in terms of limits

Contents

  • Definition coincides with the direct definition
Require Import UniMath.Foundations.PartD.
Require Import UniMath.Foundations.Propositions.
Require Import UniMath.Foundations.Sets.

Require Import UniMath.CategoryTheory.Core.Categories.

Require Import UniMath.CategoryTheory.limits.graphs.limits.
Require Import UniMath.CategoryTheory.limits.graphs.colimits.
Local Open Scope cat.
Require Import UniMath.CategoryTheory.limits.graphs.zero.
Require Import UniMath.CategoryTheory.limits.graphs.equalizers.
Require Import UniMath.CategoryTheory.limits.kernels.

Definition of kernels in terms of limits

Section def_kernels.

  Variable C : precategory.
  Variable hs: has_homsets C.
  Variable Z : Zero C.

  Definition Kernel {a b : C} (f : C⟦a, b⟧) := Equalizer C f (ZeroArrow Z a b).

Maps between Kernels as limits and direct definition.

  Lemma equiv_Kernel1_eq {a b : C} (f : C⟦a, b⟧) (K : limits.kernels.Kernel (equiv_Zero2 Z) f) :
    KernelArrow K · f = KernelArrow K · ZeroArrow Z a b.
  Show proof.
    rewrite precomp_with_ZeroArrow. rewrite <- equiv_ZeroArrow. apply KernelCompZero.

  Lemma equiv_Kernel1_isEqualizer {a b : C} (f : C⟦a, b⟧)
        (K : limits.kernels.Kernel (equiv_Zero2 Z) f) :
    isEqualizer C f (ZeroArrow Z a b) K (KernelArrow K) (equiv_Kernel1_eq f K).
  Show proof.
    use (mk_isEqualizer _ hs).
    intros e h' H'.
    use unique_exists.
    - use KernelIn.
      + exact h'.
      + rewrite precomp_with_ZeroArrow in H'.
        use (pathscomp0 _ (!(equiv_ZeroArrow e b Z))).
        exact H'.
    - cbn. use KernelCommutes.
    - intros y. apply hs.
    - intros y X. cbn in X.
      use limits.kernels.KernelInsEq. unfold KernelArrow.
      use (pathscomp0 X). apply pathsinv0.
      use limits.kernels.KernelCommutes.

  Definition equiv_Kernel1 {a b : C} (f : C⟦a, b⟧) (K : limits.kernels.Kernel (equiv_Zero2 Z) f) :
    Kernel f.
  Show proof.
    use mk_Equalizer.
    - exact K.
    - exact (KernelArrow K).
    - exact (equiv_Kernel1_eq f K).
    - exact (equiv_Kernel1_isEqualizer f K).


  Lemma equiv_Kernel2_eq {a b : C} (f : C⟦a, b⟧) (K : Kernel f) :
    EqualizerArrow C K · f = limits.zero.ZeroArrow (equiv_Zero2 Z) (EqualizerObject C K) b.
  Show proof.
    rewrite (EqualizerArrowEq C K). rewrite equiv_ZeroArrow.
    rewrite precomp_with_ZeroArrow. apply idpath.

  Lemma equiv_Kernel2_isEqualizer {a b : C} (f : C⟦a, b⟧) (K : Kernel f) :
    isKernel (equiv_Zero2 Z) (EqualizerArrow C K) f (equiv_Kernel2_eq f K).
  Show proof.
    use (mk_isKernel hs).
    intros w h H.
    use unique_exists.
    - use EqualizerIn.
      + exact h.
      + rewrite H.
        rewrite precomp_with_ZeroArrow.
        apply equiv_ZeroArrow.
    - use EqualizerArrowComm.
    - intros y. apply hs.
    - intros y T. cbn in T.
      use EqualizerInUnique. exact T.

  Definition equiv_Kernel2 {a b : C} (f : C⟦a, b⟧) (K : Kernel f) :
    limits.kernels.Kernel (equiv_Zero2 Z) f.
  Show proof.
    use mk_Kernel.
    - exact (EqualizerObject C K).
    - exact (EqualizerArrow C K).
    - exact (equiv_Kernel2_eq f K).
    - exact (equiv_Kernel2_isEqualizer f K).

End def_kernels.