Diagonal functor

from Wikipedia, the free encyclopedia

In the mathematical subfield of category theory , the diagonal functor is a functor that allows a category to be embedded in the category of functors for any non-empty ( small ) category . The name comes from the fact that for a discrete with two elements the diagonal functor is just the mapping .

Definition and functoriality

Be a category and a small category. Then the diagonal functor is defined as a mapping that assigns a natural transformation to each morphism , given that they assign each object and thus each morphism to the morphism . For an object there is obviously a functor. In order to see that is actually a functor, consider the concatenation of natural transformations for morphisms and from the category and , by definition, this results in the following commutative diagram for each :

Diagonal functor.svg

This is nothing more than:

Diagonal functor (replaced) .svg

This corresponds to the natural transformation , which proves that . For non-empty is obviously injective , i.e. embedded in the corresponding functor category . Under a certain condition is also full : be natural transformation, i. i.e. that for each in the diagram

Fullness of a diagonal functor.svg

commutes (because and ). Which means nothing else than that whenever a morphism exists between and . If the category as a graph construed weakly connected is, is so constant and thus the image of , which is full. For example, this is true for an arrow category or, more generally, for with a start or end object, but not for a product for discrete with at least two elements.

Connection with limits

A cone with respect to a functor is nothing more than an object in provided with a natural transformation from to . A Limes of is a special cone, namely a - universal solution for . Dual to this is a Kolimes of a special Kokegel, namely a -universal solution for . If has a right adjoint functor , then it is completely up with respect to limits , the converse also applies. This adjoint functor is precisely the Limes functor . Correspondingly, the Kolimes functor (if it exists) is left adjoint to the diagonal functor.

The diagonal functor is continuous , i.e. i.e., it receives all limits that exist in. He also receives all Kolimites.

literature

Web links

Individual evidence

  1. Pumplün, pp. 105-106
  2. ^ Mac Lane, p. 233
  3. Pumplün, p. 169