nLab weak factorization system

Redirected from "weak factorization systems".
Contents

Contents

Idea

A weak factorization system on a category is a pair (,)(\mathcal{L},\mathcal{R}) of classes of morphisms (“projective morphisms” and “injective morphisms”) such that 1) every morphism of the category factors as the composite of one in \mathcal{L} followed by one in \mathcal{R}, and 2) \mathcal{L} and \mathcal{R} are closed under having the lifting property against each other.

If the liftings here are unique, then one speaks of an orthogonal factorization system. A classical example of an orthogonal factorization system is the (epi,mono)-factorization system on the category Set or in fact on any topos.

Non-orthogonal weak factorization systems are the key ingredient in model categories, which by definition carry a weak factorization system called (=\mathcal{L} = cofibrations,=\mathcal{R} = acyclic fibrations) and another one called (=\mathcal{L} = acyclic cofibrations, =\mathcal{R} = fibrations). Indeed most examples of non-orthogonal weak factorization systems arise in the context of model category theory. A key tool for constructing these, or verifying their existence, is the small object argument.

There are other properties which one may find or impose on a weak factorization system, for instance functorial factorization. There is also extra structure which one may find or impose, such as for algebraic weak factorization systems. For more variants see at factorization system.

Definition

Weak factorization systems

Definition

A weak factorization system (WFS) on a category 𝒞\mathcal{C} is a pair (,)(\mathcal{L},\mathcal{R}) of classes of morphisms of 𝒞\mathcal{C} such that

  1. Every morphism f:XYf \colon X\to Y of 𝒞\mathcal{C} may be factored as the composition of a morphism in \mathcal{L} followed by one in \mathcal{R}

    f:XZY. f\;\colon\; X \overset{\in \mathcal{L}}{\longrightarrow} Z \overset{\in \mathcal{R}}{\longrightarrow} Y \,.
  2. The classes are closed under having the lifting property against each other:

    1. \mathcal{L} is precisely the class of morphisms having the left lifting property against every morphism in \mathcal{R};

    2. \mathcal{R} is precisely the class of morphisms having the right lifting property against every morphism in \mathcal{L}.

Definition

For 𝒞\mathcal{C} a category, a functorial factorization of the morphisms in 𝒞\mathcal{C} is a functor

fact:𝒞 Δ[1]𝒞 Δ[2] fact \;\colon\; \mathcal{C}^{\Delta[1]} \longrightarrow \mathcal{C}^{\Delta[2]}

which is a section of the composition functor d 1:𝒞 Δ[2]𝒞 Δ[1]d_1 \;\colon \;\mathcal{C}^{\Delta[2]}\to \mathcal{C}^{\Delta[1]}.

Remark

In def. we are using the following notation, see at simplex category and at nerve of a category:

Write Δ[1]={01}\Delta[1] = \{0 \to 1\} and Δ[2]={012}\Delta[2] = \{0 \to 1 \to 2\} for the ordinal numbers, regarded as posets and hence as categories. The arrow category Arr(𝒞)Arr(\mathcal{C}) is equivalently the functor category 𝒞 Δ[1]Funct(Δ[1],𝒞)\mathcal{C}^{\Delta[1]} \coloneqq Funct(\Delta[1], \mathcal{C}), while 𝒞 Δ[2]Funct(Δ[2],𝒞)\mathcal{C}^{\Delta[2]}\coloneqq Funct(\Delta[2], \mathcal{C}) has as objects pairs of composable morphisms in 𝒞\mathcal{C}. There are three injective functors δ i:[1][2]\delta_i \colon [1] \rightarrow [2], where δ i\delta_i omits the index ii in its image. By precomposition, this induces functors d i:𝒞 Δ[2]𝒞 Δ[1]d_i \colon \mathcal{C}^{\Delta[2]} \longrightarrow \mathcal{C}^{\Delta[1]}. Here

  • d 1d_1 sends a pair of composable morphisms to their composition;

  • d 2d_2 sends a pair of composable morphisms to the first morphism;

  • d 0d_0 sends a pair of composable morphisms to the second morphism.

Definition

A weak factorization system, def. , is called a functorial weak factorization system if the factorization of morphisms may be chosen to be a functorial factorization factfact, def. , i.e. such that d 2factd_2 \circ fact lands in \mathcal{L} and d 0factd_0\circ fact in \mathcal{R}.

Remark

Not all weak factorization systems are functorial, although most (including those produced by the small object argument, with due care) are. But all orthogonal factorization systems, def. , automatically are functorial. An example of a weak factorization that is not functorial can be found in Isaksen 2001.

Orthogonal factorization systems

Definition

An orthogonal factorization system (OFS) is a weak factorization system (,)(\mathcal{L},\mathcal{R}), def. such that the lifts of elements in \mathcal{L} against elements in \mathcal{R} are unique.

Remark

While every OFS (def. ) is a WFS (def. ), the primary examples of each are different:

A “basic example” of an OFS is (epi,mono)-factorization in Set (meaning LL is the collection of epimorphisms and RR that of monomorphisms), while a “basic example” of a WFS is (mono, epi) in SetSet. The superficial similarity of these two examples masks the fact that they generalize in very different ways.

The OFS (epi, mono) generalizes to any topos or pretopos, and in fact to any regular category if we replace “epi” with regular epi. Likewise it generalizes to any quasitopos if we instead replace “mono” with regular mono.

On the other hand, saying that (mono,epi) is a WFS in Set is equivalent to the axiom of choice. A less loaded statement is that (L,R)(L,R) is a WFS, where LL is the class of inclusions AABA\hookrightarrow A\sqcup B into a binary coproduct and RR is the class of split epis. In this form the statement generalizes to any extensive category; see also weak factorization system on Set.

Algebraic weak factorization systems

An algebraic weak factorization system enhances the properties of lifting and factorization to algebraic structure.

Accessible weak factorization systems

An accessible weak factorization system is a wfs on a locally presentable category whose factorization is given by an accessible functor.

Properties

Closure properties

Proposition

Let (,)(\mathcal{L},\mathcal{R}) be a weak factorization system, def. on some category 𝒞\mathcal{C}. Then

  1. Both classes contain the class of isomorphism of 𝒞\mathcal{C}.

  2. Both classes are closed under composition in 𝒞\mathcal{C}.

    \mathcal{L} is also closed under transfinite composition.

  3. Both classes are closed under forming retracts in the arrow category 𝒞 Δ[1]\mathcal{C}^{\Delta[1]} (see remark ).

  4. \mathcal{L} is closed under forming pushouts of morphisms in 𝒞\mathcal{C} (“cobase change”).

    \mathcal{R} is closed under forming pullback of morphisms in 𝒞\mathcal{C} (“base change”).

  5. \mathcal{L} is closed under forming coproducts in 𝒞 Δ[1]\mathcal{C}^{\Delta[1]}.

    \mathcal{R} is closed under forming products in 𝒞 Δ[1]\mathcal{C}^{\Delta[1]}.

Proof

We go through each item in turn.

containing isomorphisms

Given a commuting square

A f X Iso i p B g Y \array{ A &\overset{f}{\rightarrow}& X \\ {}_{\mathllap{\in Iso}}^{\mathllap{i}}\downarrow && \downarrow^{\mathrlap{p}} \\ B &\underset{g}{\longrightarrow}& Y }

with the left morphism an isomorphism, then a lift is given by using the inverse of this isomorphism fi 1{}^{{f \circ i^{-1}}}\nearrow. Hence in particular there is a lift when pp \in \mathcal{R} and so ii \in \mathcal{L}. The other case is formally dual.

closure under composition

Given a commuting square of the form

A X p 1 i p 2 B Y \array{ A &\longrightarrow& X \\ \downarrow && \downarrow^{\mathrlap{p_1}}_{\mathrlap{\in \mathcal{R}}} \\ {}^{\mathllap{i}}_{\mathllap{\in \mathcal{L}}}\downarrow && \downarrow^{\mathrlap{p_2}}_{\mathrlap{\in \mathcal{R}}} \\ B &\longrightarrow& Y }

consider its pasting decomposition as

A X p 1 i p 2 B Y. \array{ A &\longrightarrow& X \\ \downarrow &\searrow& \downarrow^{\mathrlap{p_1}}_{\mathrlap{\in \mathcal{R}}} \\ {}^{\mathllap{i}}_{\mathllap{\in \mathcal{L}}}\downarrow && \downarrow^{\mathrlap{p_2}}_{\mathrlap{\in \mathcal{R}}} \\ B &\longrightarrow& Y } \,.

Now the bottom commuting square has a lift, by assumption. This yields another pasting decomposition

A X i p 1 p 2 B Y \array{ A &\longrightarrow& X \\ {}^{\mathllap{i}}_{\mathllap{\in \mathcal{L}}}\downarrow && \downarrow^{\mathrlap{p_1}}_{\mathrlap{\in \mathcal{R}}} \\ \downarrow &\nearrow& \downarrow^{\mathrlap{p_2}}_{\mathrlap{\in \mathcal{R}}} \\ B &\longrightarrow& Y }

and now the top commuting square has a lift by assumption. This is now equivalently a lift in the total diagram, showing that p 1p 1p_1\circ p_1 has the right lifting property against \mathcal{L} and is hence in \mathcal{R}. The case of composing two morphisms in \mathcal{L} is formally dual. From this the closure of \mathcal{L} under transfinite composition follows since the latter is given by colimits of sequential composition and successive lifts against the underlying sequence as above constitutes a cocone, whence the extension of the lift to the colimit follows by its universal property.

closure under retracts

Let jj be the retract of an ii \in \mathcal{L}, i.e. let there be a commuting diagram of the form.

id A: A C A j i j id B: B D B. \array{ id_A \colon & A &\longrightarrow& C &\longrightarrow& A \\ & \downarrow^{\mathrlap{j}} && \downarrow^{\mathrlap{i}}_{\mathrlap{\in \mathcal{L}}} && \downarrow^{\mathrlap{j}} \\ id_B \colon & B &\longrightarrow& D &\longrightarrow& B } \,.

Then for

A X j f B Y \array{ A &\longrightarrow& X \\ {}^{\mathllap{j}}\downarrow && \downarrow^{\mathrlap{f}}_{\mathrlap{\in \mathcal{R}}} \\ B &\longrightarrow& Y }

a commuting square, it is equivalent to its pasting composite with that retract diagram

A C A X j i j f B D B Y. \array{ A &\longrightarrow& C &\longrightarrow& A &\longrightarrow& X \\ \downarrow^{\mathrlap{j}} && \downarrow^{\mathrlap{i}}_{\mathrlap{\in \mathcal{L}}} && \downarrow^{\mathrlap{j}} && \downarrow^{\mathrlap{f}}_{\mathrlap{\in \mathcal{R}}} \\ B &\longrightarrow& D &\longrightarrow& B &\longrightarrow & Y } \,.

Now the pasting composite of the two squares on the right has a lift, by assumption,

A C A X j i f B D B Y. \array{ A &\longrightarrow& C &\longrightarrow& A &\longrightarrow& X \\ \downarrow^{\mathrlap{j}} && \downarrow^{\mathrlap{i}}_{\mathrlap{\in \mathcal{L}}} && \nearrow && \downarrow^{\mathrlap{f}}_{\mathrlap{\in \mathcal{R}}} \\ B &\longrightarrow& D &\longrightarrow& B &\longrightarrow & Y } \,.

By composition, this is also a lift in the total outer rectangle, hence in the original square. Hence jj has the left lifting property against all pp \in \mathcal{R} and hence is in \mathcal{L}. The other case is formally dual.

closure under pushout and pullback

Let pp \in \mathcal{R} and and let

Z× fX X f *p p Z f Y \array{ Z \times_f X &\longrightarrow& X \\ {}^{\mathllap{{f^* p}}}\downarrow && \downarrow^{\mathrlap{p}} \\ Z &\stackrel{f}{\longrightarrow} & Y }

be a pullback diagram in 𝒞\mathcal{C}. We need to show that f *pf^* p has the right lifting property with respect to all ii \in \mathcal{L}. So let

A Z× fX i f *p B g Z \array{ A &\longrightarrow& Z \times_f X \\ {}^{\mathllap{i}}_{\mathllap{\in \mathcal{L}}}\downarrow && \downarrow^{\mathrlap{\mathrlap{f^* p}}} \\ B &\stackrel{g}{\longrightarrow}& Z }

be a commuting square. We need to construct a diagonal lift of that square. To that end, first consider the pasting composite with the pullback square from above to obtain the commuting diagram

A Z× fX X i f *p p B g Z f Y. \array{ A &\longrightarrow& Z \times_f X &\longrightarrow& X \\ {}^{\mathllap{i}}\downarrow && \downarrow^{\mathrlap{f^* p}} && \downarrow^{\mathrlap{p}} \\ B &\stackrel{g}{\longrightarrow}& Z &\stackrel{f}{\longrightarrow}& Y } \,.

By the right lifting property of pp, there is a diagonal lift of the total outer diagram

A X i (fg)^ p B fg Y. \array{ A &\longrightarrow& X \\ \downarrow^{\mathrlap{i}} &{}^{\hat {(f g)}}\nearrow& \downarrow^{\mathrlap{p}} \\ B &\stackrel{f g}{\longrightarrow}& Y } \,.

By the universal property of the pullback this gives rise to the lift g^\hat g in

Z× fX X g^ f *p p B g Z f Y. \array{ && Z \times_f X &\longrightarrow& X \\ &{}^{\hat g} \nearrow& \downarrow^{\mathrlap{f^* p}} && \downarrow^{\mathrlap{p}} \\ B &\stackrel{g}{\longrightarrow}& Z &\stackrel{f}{\longrightarrow}& Y } \,.

In order for g^\hat g to qualify as the intended lift of the total diagram, it remains to show that

A Z× fX i g^ B \array{ A &\longrightarrow& Z \times_f X \\ \downarrow^{\mathrlap{i}} & {}^{\hat g}\nearrow \\ B }

commutes. To do so we notice that we obtain two cones with tip AA:

  • one is given by the morphisms

    1. AZ× fXXA \to Z \times_f X \to X
    2. AiBgZA \stackrel{i}{\to} B \stackrel{g}{\to} Z

    with universal morphism into the pullback being

    • AZ× fXA \to Z \times_f X
  • the other by

    1. AiBg^Z× fXXA \stackrel{i}{\to} B \stackrel{\hat g}{\to} Z \times_f X \to X
    2. AiBgZA \stackrel{i}{\to} B \stackrel{g}{\to} Z.

    with universal morphism into the pullback being

    • AiBg^Z× fXA \stackrel{i}{\to} B \stackrel{\hat g}{\to} Z \times_f X.

The commutativity of the diagrams that we have established so far shows that the first and second morphisms here equal each other, respectively. By the fact that the universal morphism into a pullback diagram is unique this implies the required identity of morphisms.

The other case is formally dual.

closure under (co-)products

Let {(A si sB s)} sS\{(A_s \overset{i_s}{\to} B_s) \in \mathcal{L}\}_{s \in S} be a set of elements of \mathcal{L}. Since colimits in the presheaf category 𝒞 Δ[1]\mathcal{C}^{\Delta[1]} are computed componentwise, their coproduct in this arrow category is the universal morphism out of the coproduct of objects sSA s\underset{s \in S}{\coprod} A_s induced via its universal property by the set of morphisms i si_s:

sSA s(i s) sSsSB s. \underset{s \in S}{\sqcup} A_s \overset{(i_s)_{s\in S}}{\longrightarrow} \underset{s \in S}{\sqcup} B_s \,.

Now let

sSA s X (i s) sS f sSB s Y \array{ \underset{s \in S}{\sqcup} A_s &\longrightarrow& X \\ {}^{\mathllap{(i_s)_{s\in S}}}\downarrow && \downarrow^{\mathrlap{f}}_{\mathrlap{\in \mathcal{R}}} \\ \underset{s \in S}{\sqcup} B_s &\longrightarrow& Y }

be a commuting square. This is in particular a cocone under the coproduct of objects, hence by the universal property of the coproduct, this is equivalent to a set of commuting diagrams

{A s X i s f B s Y} sS. \left\{ \;\;\;\; \array{ A_s &\longrightarrow& X \\ {}^{\mathllap{i_s}}_{\mathllap{\in \mathcal{L}}}\downarrow && \downarrow^{\mathrlap{f}}_{\mathrlap{\in \mathcal{R}}} \\ B_s &\longrightarrow& Y } \;\;\;\; \right\}_{s\in S} \,.

By assumption, each of these has a lift s\ell_s. The collection of these lifts

{A s X i s s f B s Y} sS \left\{ \;\;\;\; \array{ A_s &\longrightarrow& X \\ {}^{\mathllap{i_s}}_{\mathllap{\in \mathcal{L}}}\downarrow &{}^{\ell_s}\nearrow& \downarrow^{\mathrlap{f}}_{\mathrlap{\in \mathcal{R}}} \\ B_s &\longrightarrow& Y } \;\;\;\; \right\}_{s\in S}

is now itself a compatible cocone, and so once more by the universal property of the coproduct, this is equivalent to a lift ( s) sS(\ell_s)_{s\in S} in the original square

sSA s X (i s) sS ( s) sS f sSB s Y. \array{ \underset{s \in S}{\sqcup} A_s &\longrightarrow& X \\ {}^{\mathllap{(i_s)_{s\in S}}}\downarrow &{}^{(\ell_s)_{s\in S}}\nearrow& \downarrow^{\mathrlap{f}}_{\mathrlap{\in \mathcal{R}}} \\ \underset{s \in S}{\sqcup} B_s &\longrightarrow& Y } \,.

This shows that the coproduct of the i si_s has the left lifting property against all ff\in \mathcal{R} and is hence in \mathcal{L}. The other case is formally dual.

Remark

Beware, in the situation of prop. , that \mathcal{L} is not in general closed under all colimits in 𝒞 Δ[1]\mathcal{C}^{\Delta[1]}, and similarly \mathcal{R} is not in general closed under all limits in 𝒞 Δ[1]\mathcal{C}^{\Delta[1]}. Also \mathcal{L} is not in general closed under forming coequalizers in 𝒞\mathcal{C}, and \mathcal{R} is not in general closed under forming equalizers in 𝒞\mathcal{C}. However, if (,)(\mathcal{L},\mathcal{R}) is an orthogonal factorization system, def. , then \mathcal{L} is closed under all colimits and \mathcal{R} is closed under all limits.

Remark

Here by a retract of a morphism XfYX \stackrel{f}{\longrightarrow} Y in some category 𝒞\mathcal{C} is meant a retract of ff as an object in the arrow category 𝒞 Δ[1]\mathcal{C}^{\Delta[1]}, hence a morphism AgBA \stackrel{g}{\longrightarrow} B such that in 𝒞 Δ[1]\mathcal{C}^{\Delta[1]} there is a factorization of the identity on gg through ff

id g:gfg. id_g \;\colon\; g \longrightarrow f \longrightarrow g \,.

This means equivalently that in 𝒞\mathcal{C} there is a commuting diagram of the form

id A: A X A g f g id B: B Y B. \array{ id_A \colon & A &\longrightarrow& X &\longrightarrow& A \\ & \downarrow^{\mathrlap{g}} && \downarrow^{\mathrlap{f}} && \downarrow^{\mathrlap{g}} \\ id_B \colon & B &\longrightarrow& Y &\longrightarrow& B } \,.

Retract argument

Lemma

(retract argument)

Consider a composite morphism

f:XiApY. f \;\colon\; X\stackrel{i}{\longrightarrow} A \stackrel{p}{\longrightarrow} Y \,.

Then:

  1. If ff has the left lifting property against pp, then ff is a retract of ii.

  2. If ff has the right lifting property against ii, then ff is a retract of pp.

Proof

We discuss the first statement, the second is formally dual.

Write the factorization of ff as a commuting square of the form

X i A f p Y = Y. \array{ X &\stackrel{i}{\longrightarrow}& A \\ {}^{\mathllap{f}}\downarrow && \downarrow^{\mathrlap{p}} \\ Y &= & Y } \,.

By the assumed lifting property of ff against pp there exists a diagonal filler gg making a commuting diagram of the form

X i A f g p Y = Y. \array{ X &\stackrel{i}{\longrightarrow}& A \\ {}^{\mathllap{f}}\downarrow &{}^{\mathllap{g}}\nearrow& \downarrow^{\mathrlap{p}} \\ Y &= & Y } \,.

By rearranging this diagram a little, it is equivalent to

X = X f i id Y: Y g A p Y. \array{ & X &=& X \\ & {}^{\mathllap{f}}\downarrow && {}^{\mathllap{i}}\downarrow \\ id_Y \colon & Y &\underset{g}{\longrightarrow}& A &\underset{p}{\longrightarrow}& Y } \,.

Completing this to the right, this yields a diagram exhibiting the required retract according to remark :

id X: X = X = X f i f id Y: Y g A p Y. \array{ id_X \colon & X &=& X &=& X \\ & {}^{\mathllap{f}}\downarrow && {}^{\mathllap{i}}\downarrow && {}^{\mathllap{f}}\downarrow \\ id_Y \colon & Y &\underset{g}{\longrightarrow}& A &\underset{p}{\longrightarrow}& Y } \,.

Examples

  • Model categories provide many examples of weak factorization systems. In fact, most applications of WFS involve model categories or model-categorical ideas.

  • The existence of certain WFS on Set is related to the axiom of choice.

  • See the Catlab for more examples.

Algebraic model structures: Quillen model structures, mainly on locally presentable categories, and their constituent categories with weak equivalences and weak factorization systems, that can be equipped with further algebraic structure and “freely generated” by small data.

structuresmall-set-generatedsmall-category-generatedalgebraicized
weak factorization systemcombinatorial wfsaccessible wfsalgebraic wfs
model categorycombinatorial model categoryaccessible model categoryalgebraic model category
construction methodsmall object argumentsame as \toalgebraic small object argument

References

Introductory texts:

The following dissertation section is entirely written after learning of Riehl (2008) above, but has complementary examples and may dive deeper into some proofs:

  • Andreas Nuyts, Contributions to Multimode and Presheaf Type Theory, section 2.4: Factorization Systems, PhD thesis, KU Leuven, Belgium, 2020

Last revised on April 17, 2024 at 16:39:09. See the history of this page for a list of all contributions to it.