Finite support of multiset and distribution functor

47 Views Asked by At

It is unclear to me why in the distribution functor as well as in the multiset functor we require finite support. Below, I provided the definition from "Introduction to Coalgebra" by Bart Jacobs.

In several other sources, this is explained in a similar fashion, not at all. It is usually just stated that it is necessary for it to be finite, or to avoid infinte sums. As I do not have a background in math, I am assuming I must be missing something obvious.

Def: For a commutative monoid $M = (M,+,0)$ we define the multiset functor $M_M: Sets \to Sets$ on a set $X$ as $$ M_M(X) = \{ \varphi: X \to M \mid supp(\varphi) \text{ is finite} \} $$ where $supp(\varphi) = \{ x \in X \mid \varphi(x) \neq 0\}$ is called the support of $\varphi$.

On a function $f: X \to Y$ one has $M_M(f): M_M(X) \to M_M(Y)$ via $$M_M(f)(\varphi)(y) = \sum_{x \in f^{-1}(y)} \varphi(x) = \sum\{\varphi(x) \mid x \in supp(\varphi) \text{ with } f(x) = y\} $$