1
$\begingroup$

Let $f: \overline{\mathbb{D}} \to \mathbb{R}^+$ be a real (positive) valued function on the closed unit disc that is bounded and analytic on $\mathbb{D}$ (open unit disc) and $$\lim_{|z| \to 1}f(z) = 1.$$ Can we conclude that $f(z) \leq 1$ in the whole disc?

I'm aware of a Blaschke product type argument that says that a function that is of unit modulus on the boundary, and bounded and analytic in the interior of the disc must have finitely many zeroes and hence must be a finite Blaschke product - however, can we ensure that a Blaschke product is always real valued? And if so, what does this say about the bound on $f$ in the interior of the disc?

EDIT: An alternative formulation to the above, would be to consider the function $|f|$ where $f$ is an analytic, bounded function on the interior of the disc and continuous and bounded on the boundary with $$\lim_{|z| \to 1} |f(z)| =1.$$ Similar to the above, the question would now concern the bound on $|f|$.

| cite | improve this question | |
$\endgroup$
  • 1
    $\begingroup$ Edit: Reading the question again, the only analytic function with a purely real image would be a constant so yes, that inequality is trivially true because that means $f=1$ $\endgroup$ – Ninad Munshi Sep 13 '19 at 11:44
  • $\begingroup$ Thank you @NinadMunshi! I have actually added an edit to the question in response to your previous comment which adds a bit more of a substantial query. Though, your answer may well be enough for my problem! $\endgroup$ – CS1994 Sep 13 '19 at 11:50
  • $\begingroup$ @OlivierRoche No that is not analytic. One definition of analytic is that $f\in \text{ker}\left(\frac{\partial}{\partial \bar{z}}\right)$ but for your function, $\frac{\partial f}{\partial \bar{z}} \neq 0$ $\endgroup$ – Ninad Munshi Sep 13 '19 at 11:58
  • 1
    $\begingroup$ I hoped the function $z \mapsto 2 - |z|$ would give a counterexample but this fails since it isn't continuously differentiable in $0$. I think the function $f : z \mapsto 2 - \cos (|z|\cdot \frac{\pi}{2})$ might do the trick, but I'm too lazy to check if $f$ is analytic. :D $\endgroup$ – Olivier Roche Sep 13 '19 at 12:03
1
$\begingroup$

The answer to the edited question is YES.

Let $\epsilon >0$. Then there exists $\delta >0$ such that $|f(z)| \leq 1+\epsilon$ for $|z| \geq 1-\delta$. By MMP applied to the disk of radius $1-\delta$ we get $|f(z)| \leq 1+\epsilon$ whenever $|z| \leq 1-\delta$. Can you finish the proof now?

| cite | improve this answer | |
$\endgroup$
  • $\begingroup$ I'm sorry, what is MMP? (My complex analysis is abysmal) $\endgroup$ – CS1994 Sep 13 '19 at 12:13
  • $\begingroup$ MMP stands for Maximum Modulus Principle. According to this theorem if $|f(z)| \leq M$ for $|z|=R$ then $|f(z)| \leq M$ for $|z|\leq R$. $\endgroup$ – Kavi Rama Murthy Sep 13 '19 at 12:14
  • $\begingroup$ Ah okay, so this immediately just gives that since $|f(z)| \leq1$ on $|z|=1$, then $|f(z)| \leq 1$ also for $|z|\leq 1$? $\endgroup$ – CS1994 Sep 13 '19 at 12:16
  • $\begingroup$ Unfortunately you have to use a more complicated argument. This is because $f$ is only given to be analytic in the open set $(|z|<1)$ and the circle $|z|=1$ is not inside this open set , so MMP cannot be applied directly. @CS1994 $\endgroup$ – Kavi Rama Murthy Sep 13 '19 at 12:20
  • $\begingroup$ Could the Blaschke product argument not work in this scenario? Here we have an analytic function $f$ in the open disc that has the limit as stated in the question. This implies that $f$ must have finite number of zeroes and be of the form of a finite Blaschke product (see Fatou for the well-known result). A property of a finite Blaschke product is that its modulus is at most 1 in the interior of the disc. The proof of Fatou's result looks similar to the logic you are using here. @kaviramamurthy $\endgroup$ – CS1994 Sep 13 '19 at 12:27

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.