open import Cat.Instances.Shape.Terminal
open import Cat.Functor.Coherence
open import Cat.Instances.Functor
open import Cat.Prelude

import Cat.Functor.Reasoning as Func
import Cat.Reasoning as Cat
module Cat.Functor.Kan.Base where
private
  variable
    o ℓ : Level
    C C' D E : Precategory o ℓ
  kan-lvl :  {o ℓ o' ℓ' o'' ℓ''} {C : Precategory o ℓ} {C' : Precategory o' ℓ'} {D : Precategory o'' ℓ''}
           Functor C D  Functor C C'  Level
  kan-lvl {a} {b} {c} {d} {e} {f} _ _ = a ⊔ b ⊔ c ⊔ d ⊔ e ⊔ f

open _=>_

Left Kan extensions🔗

Suppose we have a functor and a functor — perhaps to be thought of as a full subcategory inclusion, where is a completion of but the situation applies just as well to any pair of functors — which naturally fit into a commutative diagram

but as we can see this is a particularly sad commutative diagram; it’s crying out for a third edge

extending to a functor If there exists an universal such extension (we’ll define what “universal” means in just a second), we call it the left Kan extension of along and denote it Such extensions do not come for free (in a sense they’re pretty hard to come by), but concept of Kan extension can be used to rephrase the definition of both limit and adjoint functor.

A left Kan extension is equipped with a natural transformation witnessing the (“directed”) commutativity of the triangle (so that it need not commute on-the-nose) which is universal among such transformations; Meaning that if is another functor with a transformation there is a unique natural transformation which commutes with

Note that in general the triangle commutes “weakly”, but when is fully faithful and is cocomplete, genuinely extends in that is a natural isomorphism.

record
  is-lan (p : Functor C C') (F : Functor C D) (L : Functor C' D) (eta : F => L F∘ p)
    : Type (kan-lvl p F) where
  field

Universality of eta is witnessed by the following fields, which essentially say that, in the diagram below (assuming has a natural transformation witnessing the same “directed commutativity” that does for the 2-cell exists and is unique.

    σ : {M : Functor C' D} (α : F => M F∘ p)  L => M
    σ-comm : {M : Functor C' D} {α : F => M F∘ p}  (σ α ◂ p) ∘nt eta ≡ α
    σ-uniq : {M : Functor C' D} {α : F => M F∘ p} {σ' : L => M}
            α ≡ (σ' ◂ p) ∘nt eta
            σ α ≡ σ'

  σ-uniq₂
    : {M : Functor C' D} (α : F => M F∘ p) {σ₁' σ₂' : L => M}
     α ≡ (σ₁' ◂ p) ∘nt eta
     α ≡ (σ₂' ◂ p) ∘nt eta
     σ₁' ≡ σ₂'
  σ-uniq₂ β p q = sym (σ-uniq p) ∙ σ-uniq q

  σ-uniqp
    :  {M₁ M₂ : Functor C' D}
     {α₁ : F => M₁ F∘ p} {α₂ : F => M₂ F∘ p}
     (q : M₁ ≡ M₂)
     PathP  i  F => q i F∘ p) α₁ α₂
     PathP  i  L => q i) (σ α₁) (σ α₂)
  σ-uniqp q r = Nat-pathp refl q λ c' i 
    σ {M = q i} (r i) .η c'

  open _=>_ eta

We also provide a bundled form of this data.

record Lan (p : Functor C C') (F : Functor C D) : Type (kan-lvl p F) where
  field
    Ext     : Functor C' D
    eta     : F => Ext F∘ p
    has-lan : is-lan p F Ext eta

  module Ext = Func Ext
  open is-lan has-lan public

Right Kan extensions🔗

Our choice of universal property in the section above isn’t the only choice; we could instead require a terminal solution to the lifting problem, instead of an initial one. We can picture the terminal situation using the following diagram.

Note the same warnings about “weak, directed” commutativity as for left Kan extensions apply here, too. Rather than either of the triangles commuting on the nose, we have natural transformations witnessing their commutativity.

record is-ran
  (p : Functor C C') (F : Functor C D) (Ext : Functor C' D)
  (eps : Ext F∘ p => F)
  : Type (kan-lvl p F) where
  no-eta-equality

  field
    σ : {M : Functor C' D} (α : M F∘ p => F)  M => Ext
    σ-comm : {M : Functor C' D} {β : M F∘ p => F}  eps ∘nt (σ β ◂ p) ≡ β
    σ-uniq : {M : Functor C' D} {β : M F∘ p => F} {σ' : M => Ext}
            β ≡ eps ∘nt (σ' ◂ p)
            σ β ≡ σ'

  open _=>_ eps renaming (η to ε)

  σ-uniq₂
    : {M : Functor C' D} (β : M F∘ p => F) {σ₁' σ₂' : M => Ext}
     β ≡ eps ∘nt (σ₁' ◂ p)
     β ≡ eps ∘nt (σ₂' ◂ p)
     σ₁' ≡ σ₂'
  σ-uniq₂ β p q = sym (σ-uniq p) ∙ σ-uniq q

record Ran (p : Functor C C') (F : Functor C D) : Type (kan-lvl p F) where
  no-eta-equality
  field
    Ext     : Functor C' D
    eps     : Ext F∘ p => F
    has-ran : is-ran p F Ext eps

  module Ext = Func Ext
  open is-ran has-ran public
module _ {p : Functor C C'} {F : Functor C D} {G : Functor C' D} {eta : F => G F∘ p} where
  is-lan-is-prop : is-prop (is-lan p F G eta)
  is-lan-is-prop a b = path where
    private
      module a = is-lan a
      module b = is-lan b

    σ≡ : {M : Functor _ _} (α : F => M F∘ p)  a.σ α ≡ b.σ α
    σ≡ α = ext (a.σ-uniq (sym b.σ-comm) ηₚ_)

    open is-lan
    path : a ≡ b
    path i .σ α = σ≡ α i
    path i .σ-comm {α = α} =
      is-prop→pathp  i  Nat-is-set ((σ≡ α i ◂ p) ∘nt eta) α)
        (a.σ-comm {α = α}) (b.σ-comm {α = α})
        i
    path i .σ-uniq {α = α} β =
      is-prop→pathp  i  Nat-is-set (σ≡ α i) _)
        (a.σ-uniq β) (b.σ-uniq β)
        i

  instance
    H-Level-is-lan :  {k}  H-Level (is-lan p F G eta) (suc k)
    H-Level-is-lan = prop-instance is-lan-is-prop

module _ {p : Functor C C'} {F : Functor C D} {G : Functor C' D} {eps : G F∘ p => F} where
  is-ran-is-prop : is-prop (is-ran p F G eps)
  is-ran-is-prop a b = path where
    private
      module a = is-ran a
      module b = is-ran b

    σ≡ : {M : Functor _ _} (α : M F∘ p => F)  a.σ α ≡ b.σ α
    σ≡ α = ext (a.σ-uniq (sym b.σ-comm) ηₚ_)

    open is-ran
    path : a ≡ b
    path i .σ α = σ≡ α i
    path i .σ-comm {β = α} =
      is-prop→pathp  i  Nat-is-set (eps ∘nt (σ≡ α i ◂ p)) α)
        (a.σ-comm {β = α}) (b.σ-comm {β = α})
        i
    path i .σ-uniq {β = α} γ =
      is-prop→pathp  i  Nat-is-set (σ≡ α i) _)
        (a.σ-uniq γ) (b.σ-uniq γ)
        i

  instance
    H-Level-is-ran :  {k}  H-Level (is-ran p F G eps) (suc k)
    H-Level-is-ran = prop-instance is-ran-is-prop

Preservation and reflection of Kan extensions🔗

Let be the left Kan extension of along and suppose that is a functor. We can “apply” to all the data of the Kan extension, obtaining the following diagram.

This looks like yet another Kan extension diagram, but it may not be universal! If this diagram is a left Kan extension, we say that preserves

module _
  {p : Functor C C'} {F : Functor C D} {G : Functor C' D} {eta : F => G F∘ p} where
  preserves-lan : (H : Functor D E)  is-lan p F G eta  Type _
  preserves-lan H _ =
    is-lan p (H F∘ F) (H F∘ G) (nat-assoc-to (H ▸ eta))

In the diagram above, the 2-cell is simply the whiskering Unfortunately, proof assistants; our definition of whiskering lands in but we require a natural transformation to

We say that a Kan extension is absolute if it is preserved by all functors out of An important class of examples is given by adjoint functors.

  is-absolute-lan : is-lan p F G eta  Typeω
  is-absolute-lan lan =
    {o ℓ : Level} {E : Precategory o ℓ} (H : Functor D E)  preserves-lan H lan

It may also be the case that is already a left kan extension of along We say that reflects this Kan extension if is a also a left extension of along

  reflects-lan
    : (H : Functor D E)
     is-lan p (H F∘ F) (H F∘ G) (nat-assoc-to (H ▸ eta))
     Type _
  reflects-lan _ _ =
    is-lan p F G eta
module _
  {p : Functor C C'} {F : Functor C D} {G : Functor C' D} {eps : G F∘ p => F} where

We can define dual notions for right Kan extensions as well.

  preserves-ran : (H : Functor D E)  is-ran p F G eps  Type _
  preserves-ran H _ =
    is-ran p (H F∘ F) (H F∘ G) (nat-assoc-from (H ▸ eps))

  is-absolute-ran : is-ran p F G eps  Typeω
  is-absolute-ran ran =
    {o ℓ : Level} {E : Precategory o ℓ} (H : Functor D E)  preserves-ran H ran

  reflects-ran
    : (H : Functor D E)
     is-ran p (H F∘ F) (H F∘ G) (nat-assoc-from (H ▸ eps))
     Type _
  reflects-ran _ _ =
    is-ran p F G eps
to-lan
  :  {p : Functor C C'} {F : Functor C D} {L : Functor C' D} {eta : F => L F∘ p}
   is-lan p F L eta
   Lan p F
to-lan {L = L} lan .Lan.Ext = L
to-lan {eta = eta} lan .Lan.eta = eta
to-lan lan .Lan.has-lan = lan