4
$\begingroup$

Let me lay out a definition first.

Definition: A non-empty open connected subset of $\mathbb{C}$ is called a domain.

I am currently self-studying Complex Analysis and have been referring to multiple books. Right now, I am at the point of trying to understand the definition of holomorphic functions.

What got me confused is that in some books, they define holomorphic functions on a domain (cf. Lecture Notes by Ivan F. Wilde) while on some books, they define holomorphic functions on open subsets of $\mathbb{C}$ (cf. Stein-Shakarchi Complex Analysis, Priestley's Intro to Complex Analysis, Lang's Complex Analysis).

So, my question is, why is this so? Would there be major consequences from this slight difference of definition? Thanks in advance.

$\endgroup$
5
$\begingroup$

Every open set can be partitioned into at most countably many domains (the connected components of the open set). A function is holomorphic on an open set if and only if it is holomorphic on every connected component thereof (since being holomorphic is a local property). This basically tells you everything about the relationship: a holomorphic function on an open set is just a completely independent collection of holomorphic functions on domains. Otherwise said: if you understand everything about holomorphic functions domains, you immediately also know everything about open sets - and books using open sets as the default will explicitly say when they want connectivity.

You would need connectivity for the identity theorem. You wouldn't need it for any explicitly local theorems such as if you derived the Cauchy-Riemann equations from complex differentiability or noted the invariance of line integrals under homotopy of the curve. Sometimes you'll even want things to be simply connected, meaning that every curve is homotopic to the identity - for instance, for getting an antiderivative to a holomorphic function. There's a few flavors of these connectivity requirements - but it's not so meaningful which is used as the default.

| cite | improve this answer | |
$\endgroup$
  • $\begingroup$ This answer basically answered all the questions in my mind regarding this issue. Thank you! $\endgroup$ – devianceee Mar 29 at 14:03
5
$\begingroup$

Yes, there are big differences. For example if $U$ and $V$ are two disjoint open sets then $f(z)=1$ for $z \in U$ and $f(z)=0$ for $z \in V$ gives a holomorphic function on $\Omega =U \cup V$. Its zeros have a limit point but the function is not identically $0$. This cannot happen when $\Omega$ is a domain. Any holomorphic function whose zeros have a limti point is identically $0$ in this case.

| cite | improve this answer | |
$\endgroup$
  • $\begingroup$ Thank you for the counterexample, very illuminating indeed! $\endgroup$ – devianceee Mar 29 at 14:03

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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