Take the 2-minute tour ×
Mathematics Stack Exchange is a question and answer site for people studying math at any level and professionals in related fields. It's 100% free, no registration required.

A very stupid question from a programmer here.

There is a theorem in Pierce's "Basic category theory for computer scientists" that sounds as follows:

Let $K$ be a category, and $F$ be an endofunctor on this category. If $F$-Algebra ($A$, $a:F(A)\rightarrow A$) is initial in $F$-$Alg$ category, then $a$ is isomorphism in $K$.

Now, I can easily show that there is an arrow $h:A\rightarrow F(A)$, which is a unique $F$-homomorphism $(A, a)\rightarrow (F(A), F(a))$, and that $a\circ h = id_A$, because $a\circ h$ is a $F$-homomorphism $(A, a)\rightarrow (A, a)$, which can only be $id_A$ due to initiality of $(A, a)$.

But when it comes to showing that $h\circ a = id_{F(A)}$, I get somewhat stuck. It's obvious that I'm missing something simple, but could anybody help with finding what exactly it would be?

share|improve this question
add comment

1 Answer 1

up vote 2 down vote accepted

From $a \circ h = id_A$ you can deduce from the functoriality of $F$ that $F(a) \circ F(h) = id_{F(a)}$. But by definition of $h$ you have a commutative square: $$\begin{array}{ccc} F(A) & \xrightarrow{a} & A \\ \downarrow F(h) & & \downarrow h \\ F(F(A)) & \xrightarrow{F(a)} & F(A) \end{array}$$ Therefore $h \circ a = F(a) \circ F(h) = id_{F(A)}$.

share|improve this answer
    
I'll disobey the Big Brother. Thank you for the answer :) –  Dmitriy Korolevich Apr 9 at 12:26
add comment

Your Answer

 
discard

By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.