Skip to content
Open
Show file tree
Hide file tree
Changes from all commits
Commits
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
1 change: 1 addition & 0 deletions databases/catdat/data/categories/FI.yaml
Original file line number Diff line number Diff line change
Expand Up @@ -14,6 +14,7 @@ related_categories:
- B
- FS
- FinSet
- Mono

satisfied_properties:
- property: locally small
Expand Down
184 changes: 184 additions & 0 deletions databases/catdat/data/categories/Mono.yaml
Original file line number Diff line number Diff line change
@@ -0,0 +1,184 @@
id: Mono
name: category of injective set functions and commutative diagrams
notation: $\Mono$
objects: >-
triples $(X, Y, f)$ where $X$ and $Y$ are sets, and $f : X \hookrightarrow Y$ is an injective function
morphisms: >-
a morphism $(X, Y, f) \to (X', Y', f')$ is a pair of functions $\ell : X \to X'$ and $r : Y \to Y'$ making a commutative square
$$\begin{CD}
X @>{f}>> Y \\
@V{\ell}VV @VV{r}V \\
X' @>>{f'}> Y'
\end{CD}$$
description: >-
This is the full subcategory of $\Set^{\rightarrow}$ consisting of the injective set functions. It is also equivalent to the category of pairs of a set $X$ and a subset $S \subseteq X$, with morphisms $(X, S) \to (Y, T)$ being the functions $f : X \to Y$ such that $f(S) \subseteq T$.
nlab_link: https://ncatlab.org/nlab/show/Mono

Copy link
Copy Markdown
Owner

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

The link doesn't seem to fit.


tags:
- set theory

related_categories:
- Set_arrow
- FI

satisfied_properties:
- property: locally small
proof: This is easy.

- property: semi-strongly connected
proof: This is immediate from the fact that $\Mono$ is a full subcategory of $\Set^{\rightarrow}$, and the latter is semi-strongly connected.

Copy link
Copy Markdown
Owner

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Please add a link to the category page. This way, the category and hence the proof can be found without scrolling anywhere else first.

<a href="/category/Set_arrow">$\Set^{\rightarrow}$</a>


- property: generator
proof: >-
The object $0 \hookrightarrow 1$ is a generator. This is because it represents the codomain functor taking an object $f : X \hookrightarrow Y$ to $Y$, and taking a morphism $(\ell, r)$ to $r$. That functor is faithful because once $r$ is fixed, $\ell$ is uniquely determined since $f'$ is a monomorphism in $\Set$.

Copy link
Copy Markdown
Owner

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Please add what f' is. I think one can write (l,r) : f \to f'


- property: cogenerator
proof: >-
The object $\{ \top, \bot \} \xhookrightarrow{\id} \{ \top, \bot \}$. To see this, suppose we have two morphisms $(\ell_1, r_1), (\ell_2, r_2) : (X, Y, f) \rightrightarrows (X', Y', g)$ such that for every morphism $(\ell', r') : (X', Y', g) \to (\{ \top, \bot \}, \{ \top, \bot \})$, $(\ell', r') \circ (\ell_1, r_1) = (\ell', r') \circ (\ell_2, r_2)$. Then morphisms $(X', Y', g) \to (\{ \top, \bot \}, \{ \top, \bot \}, \id)$ are equivalent to functions $Y' \to \{ \top, \bot \}$. Since $\{ \top, \bot \}$ is a cogenerator in $\Set$, this implies $r_1 = r_2$; and since $g$ is a monomorphism in $\Set$, we automatically get $\ell_1 = \ell_2$ as well.

Copy link
Copy Markdown
Owner

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

The first sentence seems to be incomplete.

Copy link
Copy Markdown
Owner

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Isn't this an adjunction argument? Can we reuse something?

Copy link
Copy Markdown
Owner

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I think the proof is easier with the description of set-subsets pairs.


- property: complete
proof: The component-wise limit of injective functions is again an injective function, and it forms the limit in $\Mono$.
check_redundancy: false

- property: cocomplete
proof: 'We have that $\Mono$ is a reflective subcategory of $\Set^{\rightarrow}$, with the reflector taking a function $f : X \to Y$ to the inclusion map $\im(f) \hookrightarrow Y$. Therefore, the result follows from the fact that $\Set^{\rightarrow}$ is cocomplete.'

Copy link
Copy Markdown
Owner

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Let's add also here a link to the category page (last sentence).

check_redundancy: false

- property: disjoint coproducts
proof: This follows from the fact that coproducts in $\Mono$ can be computed component-wise, along with the fact that $\Set$ has disjoint coproducts.

- property: locally cartesian closed
proof: >-
For this construction, we will work in the equivalent category of pairs $(X, X')$ where $X' \subseteq X$. Now, for any object $(D, D')$ of $\Mono$, the reflector functor $\Set^{\rightarrow} \to \Mono$ induces a reflector functor of slices $\Set^{\rightarrow} / (D' \hookrightarrow D)$ to $\Mono / (D, D')$. We claim that this reflector preserves binary products, i.e. it transforms pullbacks over $D' \hookrightarrow D$ in $\Set^{\rightarrow}$ to pullbacks over $(D, D')$ in $\Mono$. Thus, suppose we have two objects $f, g$ of $\Set^{\rightarrow}$ with degree morphisms to $D' \hookrightarrow D$:

Copy link
Copy Markdown
Owner

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

pairs $(X, X')$ where $X' \subseteq X$

Let's try to find a consistent notation. It was (X,S) in the introduction.

Copy link
Copy Markdown
Owner

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

What are "degree morphisms"?

$$\begin{CD}
X @> \deg_X >> D' @< \deg_Y << Y\\
@V f VV @VVV @VV g V\\
Z @>> \deg_Z > D @<< \deg_W < W.
\end{CD}$$
Then the reflector turns $f$ into $(Z, \im(f))$ and $g$ into $(W, \im(g))$. The pullback in $\Mono$ is $(Z \times_D W, \im(f) \times_{D'} \im(g))$. On the other hand, taking the pullback in $\Set^{\rightarrow}$ first, and then applying the reflector, gives $(Z \times_D W, \im(X \times_{D'} Y \to Z \times_D W))$. Thus, we must show $\im(f) \times_{D'} \im(g) = \im(X \times_{D'} Y \to Z \times_D W)$.

We easily see that $\im(X \times_{D'} Y \to Z \times_D W) \subseteq \im(f) \times_{D'} \im(g)$ (this is the inclusion that we automatically get from the canonical comparison morphism). Conversely, suppose we have $z\in \im(f)$ and $w\in \im(g)$ with the same degree (which must be in $D'$). Then there exist $x\in X$ such that $f(x) = z$ and $y\in Y$ such that $g(y) = w$. Since $\deg_X(x) = \deg_Z(z) = \deg_W(w) = \deg_Y(y) \in D'$, we have $(x, y) \in X \times_{D'} Y$, and the image of $(x, y)$ in $Z \times_D W$ is exactly $(z, w)$, finishing the proof of the claim.

Therefore, from Lemma 5 <a href="/content/subcategories">here</a>, we conclude that each slice of $\Mono$ is cartesian closed, i.e. that $\Mono$ is locally cartesian closed.

- property: regular subobject classifier
proof: 'The object $\{ \top, \bot \} \xhookrightarrow{\id} $\{ \top, \bot \}$ is a regular subobject classifier, with $\top : (1, 1, \id) \to (\{ \top, \bot \}, \{ \top, \bot \}, \id)$ consisting of the maps with image $\top$. This follows from the description of regular monomorphisms below.'

Copy link
Copy Markdown
Owner

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

LaTeX issue: a dollar needs to be removed


- property: locally finitely presentable
proof: >-

Copy link
Copy Markdown
Owner

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Let's add a short alternative proof like: "Alternatively, we already know that the category is cocomplete, and the objects of the form ... are finitely presentable and generate the category under filtered colimits".

It is equivalent to the category of models of the finite limit sketch with a single limit cone
$$\begin{CD}
X @> \id >> X \\
@V \id VV @VV f VV \\
X @>> f > Y.
\end{CD}$$

- property: coregular
proof: >-
For this proof we will work in the equivalent category of pairs $(X, X')$ where $X' \subseteq X$. It suffices to show regular monomorphisms are stable under pushouts. Thus, suppose we have two morphisms $f : (X, X') \to (Y, Y')$ and $g : (X, X') \to (Z, Z')$ where $f$ is a regular monomorphism, which means $f$ is injective and $X' = f^{-1}(Y')$. Then we can construct a pushout diagram
$$\begin{CD}
(X, X') @> f >> (Y, Y') \\
@V g VV @VV i_1 VV \\
(Z, Z') @>> i_2 > ((Y \setminus f(X)) \sqcup Z, i_1(Y') \cup i_2(Z')).
\end{CD}$$
Here $i_1$ sends $f(x) \mapsto g(x) \in Z$ for $x \in X$, and otherwise it sends $y \mapsto y$ for $y \in Y \setminus f(X)$; and $i_2$ is the inclusion of $Z$. We certainly have $i_2$ is injective; it remains to show $i_2^{-1}(i_1(Y') \cup i_2(Z')) = Z'$. However, we have
$$i_2^{-1}(i_1(Y') \cup i_2(Z')) = i_2^{-1}(i_1(Y')) \cup i_2^{-1}(i_2(Z'))$$
and obviously $i_2^{-1}(i_2(Z')) = Z'$. Therefore, it suffices to show $i_2^{-1}(i_1(Y')) \subseteq Z'$. Thus, suppose we have $z \in Z$ such that $i_2(z) \in i_1(Y')$. That means there exists $y' \in Y'$ such that $z = i_1(y')$. But in order for $i_1(y')$ to land in the $Z$ portion of $(Y \setminus f(X)) \sqcup Z$, we must have $i_1(y') \in f(X)$. That means there exists $x \in X$ such that $f(x) = y'$; and by the assumption on $f$, in fact $x \in X'$. Therefore, $i_2(g(x)) = i_1(f(x)) = y' = i_2(z)$, so $z = g(x) \in Z'$.

Copy link
Copy Markdown
Owner

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Typo:

Suggested change
and obviously $i_2^{-1}(i_2(Z')) = Z'$. Therefore, it suffices to show $i_2^{-1}(i_1(Y')) \subseteq Z'$. Thus, suppose we have $z \in Z$ such that $i_2(z) \in i_1(Y')$. That means there exists $y' \in Y'$ such that $z = i_1(y')$. But in order for $i_1(y')$ to land in the $Z$ portion of $(Y \setminus f(X)) \sqcup Z$, we must have $i_1(y') \in f(X)$. That means there exists $x \in X$ such that $f(x) = y'$; and by the assumption on $f$, in fact $x \in X'$. Therefore, $i_2(g(x)) = i_1(f(x)) = y' = i_2(z)$, so $z = g(x) \in Z'$.
and obviously $i_2^{-1}(i_2(Z')) = Z'$. Therefore, it suffices to show $i_2^{-1}(i_1(Y')) \subseteq Z'$. Thus, suppose we have $z \in Z$ such that $i_2(z) \in i_1(Y')$. That means there exists $y' \in Y'$ such that $z = i_1(y')$. But in order for $i_1(y')$ to land in the $Z$ portion of $(Y \setminus f(X)) \sqcup Z$, we must have $y' \in f(X)$. That means there exists $x \in X$ such that $f(x) = y'$; and by the assumption on $f$, in fact $x \in X'$. Therefore, $i_2(g(x)) = i_1(f(x)) = y' = i_2(z)$, so $z = g(x) \in Z'$.

Copy link
Copy Markdown
Owner

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I think the last sentence

Therefore, $i_2(g(x)) = i_1(f(x)) = y' = i_2(z)$, so $z = g(x) \in Z'$.

is not quite correct. Instead:

Therefore, $i_2(g(x)) = i_1(f(x)) = i_1(y') = z = i_2(z)$, so $z = g(x) \in Z'$.


- property: co-Malcev
proof: >-
For this proof we will work in the equivalent category of pairs $(X, X')$ where $X' \subseteq X$. Thus, suppose we have a coreflexive corelation $p : (X \sqcup X, X' \sqcup X') \twoheadrightarrow (E, E')$ with coreflexivity morphism $r : (E, E') \to (X, X')$. From the assumption that $p$ is an epimorphism, we have that $p : X \sqcup X \to E$ is surjective. Since $\Set$ is co-Malcev, it follows that $E \simeq X \sqcup_Y X$ for some subset $Y \subseteq X$. It remains to show that $E' = i_1(X') \cup i_2(X') \subseteq X \sqcup_Y X$. Certainly, since we have a morphism $(X \sqcup_Y X, i_1(X') \cup i_2(X')) \to (E, E')$ induced by $p$, we must have $i_1(X') \cup i_2(X') \subseteq E'$. On the other hand, any element of $E'$ is equal to either $i_1(x)$ or $i_2(x)$ for $x \in X$. In the first case, we must have $r(i_1(x)) = x \in X'$, so $i_1(x) \in i_1(X')$; and similarly for the second case.

Copy link
Copy Markdown
Owner

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Let's add a link to the page for Set.

Copy link
Copy Markdown
Owner

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Let's use \cong for isomorphism, not \simeq.

Copy link
Copy Markdown
Owner

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

On the other hand, any element of $E'$ is equal to either $i_1(x)$ or $i_2(x)$ for $x \in X$.

Maybe we can add something like any element of $E'$ is an element of $E$ and hence ...

Copy link
Copy Markdown
Owner

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

`In the first case, we must have $r(i_1(x)) = x \in X'$``

When I read a chain of relations like this $a = b \in T$, I always read it like: We know $a=b$ and $b \in T$, hence $a \in T$. But here, it's different: We know $a \in T$ and $a=b$, hence $b \in T$. Thus, I think it is better to write it as $b = a \in T$.

So here: $x = r(i_1(x)) \in X'$.


- property: effective cocongruences
proof: 'See the proof that $\Mono$ is co-Malcev: It shows that in fact any coreflexive corelation is equivalent to an effective cocongruence $X \sqcup X \to X \sqcup_Y X$.'

@ScriptRaccoon ScriptRaccoon Jul 5, 2026

Copy link
Copy Markdown
Owner

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Do you think it's worth adding a property "every reflexive relation is effective" to the database? (Not in this PR of course.)

I think we have seen this a couple of times already (Haus, ...).


unsatisfied_properties:
- property: skeletal
proof: Consider the objects $X \xhookrightarrow{\id} X$ and $Y \xhookrightarrow{\id} Y$ for isomorphic but non-equal sets $X$ and $Y$.

- property: balanced
proof: The unique morphism from $0 \hookrightarrow 1$ to $1 \hookrightarrow 1$ is both a monomorphism and an epimorphism, but not an isomorphism (see descriptions below).

- property: cofiltered-limit-stable epimorphisms
proof: >-

Copy link
Copy Markdown
Owner

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Isn't this an adjunction argument? Can we reuse something?

Consider the sequence of epimorphisms
$$\begin{CD}
\IN_{\ge n} @> \id >> \IN_{\ge n} \\
@V ! VV @VV ! V \\
1 @>> \id > 1.
\end{CD}$$
The limit is
$$\begin{CD}
0 @> \id >> 0 \\
@V ! VV @VV ! V \\
1 @>> \id > 1
\end{CD}$$
which is not an epimorphism.

- property: Malcev
proof: This is clear since $\Set$ is not Malcev, and it has a left exact embedding in $\Mono$ via the functor sending a set $X$ to $X \xhookrightarrow{\id} X$.

special_objects:
initial object:
description: the unique function $0 \to 0$
terminal object:
description: the unique function $1 \to 1$
coproducts:
description: component-wise defined disjoint union, equipped with the disjoint union of the functions
products:
description: component-wise defined cartesian product, equipped with the product function

special_morphisms:
isomorphisms:
description: pairs $(\ell, r)$ where $\ell$ and $r$ are both bijections
proof: This is easy.
monomorphisms:
description: pairs $(\ell, r)$ where $r$ is an injective function, from which it automatically follows that $\ell$ is also injective
proof: For the non-trivial direction, use the fact that the codomain functor taking an object $(X, Y, f)$ to $Y$ and a morphism $(\ell, r)$ to $r$ is representable by $0 \hookrightarrow 1$.
epimorphisms:
description: pairs $(\ell, r)$ where $r$ is a surjective function
proof: >-
Suppose we have a morphism $(\ell, r) : (X, Y, f) \to (X', Y', g)$ such that $r$ is surjective. To see it is an epimorphism, suppose we have two morphisms $(\ell_1, r_1), (\ell_2, r_2) : (X', Y', g) \rightrightarrows (T, U, h)$ such that $(\ell_1, r_1) \circ (\ell, r) = (\ell_2, r_2) \circ (\ell, r)$. Then $r_1 \circ r = r_2 \circ r$ implies $r_1 = r_2$. It automatically follows that $\ell_1 = \ell_2$ since $h$ is a monomorphism in $\Set$.

Conversely, suppose we have an epimorphism $(\ell, r) : (X, Y, f) \to (X', Y', g)$. Then for any set $T$, the morphisms $(X', Y', g) \to (T, T, \id_T)$ are equivalent to the functions $Y' \to T$; therefore, $r$ is an epimorphism in $\Set$.

Copy link
Copy Markdown
Owner

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Isn't this an adjunction argument?

regular monomorphisms:
description: >-
pairs $(\ell, r)$ where $r$ is an injective function and moreover the commutative square
$$\begin{CD}
X @> f >> Y \\
@V \ell VV @VV r V \\
X' @>> g > Y'
\end{CD}$$
is in fact a cartesian square
proof: >-

Copy link
Copy Markdown
Owner

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I think the proof will be better if we decide to either

  1. work with injective functions, as in the definition, or
  2. work with the equivalent category of set-subset pairs

The current approach is somewhere in between and I find it a bit confusing, but also inaesthetic.

I think 2. simplifies the proof, and also the formulation of the regular monos. (This also applies to the classification of monos and epis from above.)

On the other hand, from the perspective of ETCS, there is no difference at all between 1. and 2.

To show any regular monomorphism satisfies this condition, suppose we have an equalizer diagram
$$\begin{CD}
X @> f >> Y \\
@V \ell VV @VV r V \\
X' @> g >> Y' \\
@V \ell_1 V \ell_2 V @V r_1 V r_2 V \\
X'' @>> h > Y''.
\end{CD}$$
Without loss of generality, we may assume $f$, $g$, $h$, $\ell$, and $r$ are all inclusion maps. Now suppose we have $y' \in Y'$ which is in the intersection of $X'$ and $Y$. Then $y' \in Y$ implies $r_1(y') = r_2(y')$. Therefore, since $h$ is injective, we can conclude $\ell_1(y') = \ell_2(y')$, so in fact $y' \in X$. This shows $X' \cap Y \subseteq X$; the other inclusion is automatic from the commutativity of the diagram.

To show any morphism satisfying this condition is a regular monomorphism, we may again assume $f$, $g$, $\ell$, and $r$ are all inclusion maps. We can then construct an equalizer diagram
$$\begin{CD}
X @> f >> Y \\
@V \ell VV @VV r V \\
X' @> g >> Y' \\
@V \chi_X V \top_{X'} V @V \chi_Y V \top_{Y'} V \\
\{ \top, \bot \} @> \id >> \{ \top, \bot \}.
\end{CD}$$
Here $\top_{X'}$ represents the constant function on $X'$ with value $\top$, and similarly for $\top_{Y'}$. Here, the assumption $X = X' \cap Y$ is used to establish the commutativity of the characteristic morphism $(\chi_X, \chi_Y) : (X', Y', g) \to (\{ \top, \bot \}, \{ \top, \bot \}, \id)$.
regular epimorphisms:
description: pairs $(\ell, r)$ where both $\ell$ and $r$ are surjective functions
proof: >-
To show any regular epimorphism satisfies this condition, suppose we have a parallel pair
$$\begin{CD}
X @> f >> Y \\
@V \ell_1 V \ell_2 V @V r_1 V r_2 V \\
X' @> g >> Y'.

Copy link
Copy Markdown
Owner

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Suggested change
X' @> g >> Y'.
X' @>> g > Y'.

\end{CD}$$
To form the coequalizer in $\Mono$, first we may form the coequalizer in $\Set^{\rightarrow}$, $\coeq(\ell_1, \ell_2) \xrightarrow{\bar g} \coeq(r_1, r_2)$. Note this is not injective in general; however, we may apply the reflector to get the coequalizer $\im(\bar g) \hookrightarrow \coeq(r_1, r_2)$ in $\Mono$. Here, clearly the map $Y' \to \coeq(r_1, r_2)$ is surjective. Likewise, both the functions $X' \to \coeq(\ell_1, \ell_2)$ and $coeq(\ell_1, \ell_2) \to \im(\bar g)$ are surjective, so their composition $X' \to \coeq(\ell_1, \ell_2)$ is surjective as well.

Copy link
Copy Markdown
Owner

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

$coeq(\ell_1, \ell_2)

Here \ is missing.


To show any morphism $(\ell, r) : (X, Y, f) \to (X', Y', g)$ satisfying this condition is a regular epimorphism, note that the kernel pair of $(\ell, r)$ can be computed component-wise. Therefore, the coequalizer in $\Set^{\rightarrow}$ of this kernel pair is exactly $(X', Y', g)$, so applying the reflector does nothing and $(X', Y', g)$ is also the coequalizer in $\Mono$.
1 change: 1 addition & 0 deletions databases/catdat/data/categories/Set_arrow.yaml
Original file line number Diff line number Diff line change
Expand Up @@ -20,6 +20,7 @@ related_categories:
- Set
- SetxSet
- Sh(X)
- Mono

satisfied_properties:
- property: locally small
Expand Down
1 change: 1 addition & 0 deletions databases/catdat/data/macros.yaml
Original file line number Diff line number Diff line change
Expand Up @@ -122,3 +122,4 @@
\Pair: \mathbf{Pair}
\Span: \mathbf{Span}
\Arr: \mathbf{Arr}
\Mono: \mathbf{Mono}