First, see this link on the alternative characterizations of analytic functions. I want to prove a version of 3) for complex-analytic functions. In particular:

If $f$ is a complex-analytic function on an open set $D \subset \mathbb{C}$, then for every compact set $K \subset D$ there exists a constant $C$ such that for every $x \in K$ and every non-negative integer $k$ the following estimate holds: $$ \left| \frac{d^k f}{dx^k}(x) \right| \leq C^{k+1} k!$$

Using Cauchy's estimates on individual points gives a "local" version of the estimate. How do I choose a single $C$ that will work for all $x \in K$? I'm pretty sure the compactness of $K$ will play a vital role. Can I surround $K$ by a precompact neighborhood $N$ contained in $D$, then choose $C$ as the reciprocal of the distance between $K$ and the boundary of $N$?

I believe the proof of the real-analytic case might give me some insights but I haven't seen a satisfactory proof yet for the real-analytic case; if you can point me to some reference, I'll appreciate it.

I found a "proof" of the real-analytic case here but I wasn't satisfied (see my comments underneath the answer). Basically, I don't understand the transition from the Cauchy's estimates being true for centers of the sets in the finite cover to being true for all of $x$ in the compact set.


There exists $r>0$ such that $\cup_{a\in K}B(a,r)$ is contained in a larger compact set $K_1\subset D.$ Let $M= \sup_{K_1}|f|.$ Then for all $a\in K,$

$$|f^{(k)}(a)| \le k!\frac{M}{r^k} = k!\left( \frac{M^{1/k}}{r}\right)^{k}$$

by Cauchy's estimates. This will give your result.

  • $\begingroup$ This is the solution I had in mind but I wasn't sure about the existence of such an $r$. I would appreciate it if you gave more details on how to choose or why there is such an $r$. Then I can accept your answer. Thanks. $\endgroup$ – Alex Strife Jul 14 '15 at 1:33
  • $\begingroup$ The distance from $K$ to $D^c$ is positive. Any $r$ less than this distance will work. $\endgroup$ – zhw. Jul 14 '15 at 1:42
  • $\begingroup$ In the case $D = \mathbb{C}$, $D^C$ is empty... so I believe we should first surround $K$ by a bounded open set contained in $D$ and consider the complement of that instead. $\endgroup$ – Alex Strife Jul 23 '15 at 5:28
  • $\begingroup$ I have already completed the solution though. Thanks! $\endgroup$ – Alex Strife Jul 23 '15 at 5:30

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.