Complex Variables - Max Modulus Inequality

Click For Summary

Discussion Overview

The discussion revolves around the Max Modulus Inequality in the context of complex variables, specifically focusing on analytic functions defined on the unit disk. Participants explore the implications of the inequality under certain conditions, including the behavior of functions at specific points and the application of the Schwarz lemma.

Discussion Character

  • Technical explanation
  • Mathematical reasoning
  • Debate/contested

Main Points Raised

  • One participant proposes a function transformation to show that if \( f \) is analytic and bounded, then \( |f(z)| \leq M | \frac{z - \alpha}{1 - \overline{\alpha} z} | \) holds for \( |z| < 1 \).
  • Another participant challenges the initial transformation, suggesting that the mapping may not be well-defined for all \( z \) in the unit disk and proposes using the Schwarz lemma instead.
  • Further contributions reiterate the use of the Schwarz lemma to establish bounds on the transformed function \( g(z) \) and its implications for \( f(z) \).
  • Participants discuss a related problem involving the conditions on derivatives of \( f \) at the origin and whether the previously established inequalities can be used to prove a new bound on \( |f(z)| \).
  • One participant confirms that the proofs presented are correct and suggests that a standard proof similar to the Schwarz lemma can be applied to the new problem without needing the earlier inequalities.

Areas of Agreement / Disagreement

There is no consensus on the initial transformation proposed by the first participant, as it is challenged by others regarding its validity. However, there is agreement on the applicability of the Schwarz lemma to derive bounds for \( f(z) \). The discussion on the new problem remains unresolved, with differing opinions on the necessity of prior inequalities.

Contextual Notes

Participants note that the validity of certain transformations depends on the specific values of \( \alpha \) and \( z \), and there are unresolved assumptions regarding the behavior of analytic functions under these transformations.

Who May Find This Useful

This discussion may be useful for students and researchers interested in complex analysis, particularly those studying properties of analytic functions and inequalities in the context of the unit disk.

joypav
Messages
149
Reaction score
0
Suppose that f is analytic on the disc $\vert{z}\vert<1$ and satisfies $\vert{f(z)}\vert\le{M}$ if $\vert{z}\vert<1$. If $f(\alpha)=0$ for some $\alpha, \vert{\alpha}\vert<1$. Show that,
$$\vert{f(z)}\vert\le{M\vert{\frac{z-\alpha}{1-\overline{\alpha}z}}\vert}$$

What I have:
Let $g(z)=f((1-\overline{\alpha}z)z+\alpha)$. Then g is analytic on the disc $\vert{z}\vert<1$ and $g(0)=f(\alpha)=0$.
If $\vert{f(z)}\vert\le{M}$ if $\vert{z}\vert<1$, then $\vert{g(z)}\vert\le{M}$ if $\vert{z}\vert<1$

1. $\vert{f((1-\overline{\alpha}z)z+\alpha)}\vert=\vert{g(z)}\vert\le{M\vert{z}\vert} \rightarrow \vert{f(z)}\vert\le{M\vert{\frac{z-\alpha}{1-\overline{\alpha}z}}}\vert$
(By substituting $z=\frac{z-\alpha}{1-\overline{\alpha}z}$)

2. Show that $g$ does satisfy $\vert{g(z)}\vert\le{M\vert{z}\vert}$.
Let $h(z)=\frac{g(z)}{z}$. Then h(z) is analytic on $\vert{z}\vert<1$, (removable singularity at z=0). Then by the Max Modulus Theorem, $\vert{h(z)}\vert$ attains its max somewhere on the boundary of $\vert{z}\vert<1$. Call this max M. Then,
$$\forall{z}, \vert{h(z)}\vert\le{M} \rightarrow \frac{\vert{g(z)}\vert}{\vert{z}\vert}\le{M} \rightarrow \vert{g(z)}\vert\le{M\vert{z}\vert}$$

Then, show that
$$M\vert{\frac{f(z)-f(z_0)}{M^2-\overline{f(z_0)}f(z)}}\vert \le{\vert{\frac{z-z_0}{1-\overline{z_0}z}}}\vert$$
for all $z, z_0$ in $w:\vert{w}\vert<1$.I'm just wondering if my first part is correct. If so, I assume I'll utilize the first to prove the second?
 
Physics news on Phys.org
Hi joypav,

In order for $g$ to be well-defined, $(1 - \overline{\alpha}z)z + \alpha$ has to be in $\Bbb D$ for all $z\in \Bbb D$, which is not generally true (consider $\alpha = z = \frac{1}{\sqrt{2}}$). Instead, consider that the function $\phi_\alpha : z\mapsto \frac{z - \alpha}{1 - \overline{\alpha}z}$ is an analytic mapping of $\Bbb D$ onto $\Bbb D$, so we may define $g = \frac{1}{M}f\circ \phi_{\alpha}^{-1}$. Then $g$ is an analytic mapping $\Bbb D\to \overline{\Bbb D}$ with $g(0) = 0$, so by the Schwarz lemma $\lvert g(z) \rvert \le \lvert z\rvert$ for all $z\in \Bbb D$. Thus $\lvert Mg(\phi_\alpha(z))\rvert \le M\lvert \phi_\alpha(z)\rvert$, or, $\lvert f(z)\rvert \le M\lvert\frac{z - \alpha}{1 - \overline{\alpha}z}\rvert$ for all $z\in \Bbb D$.
 
For the first part...
Consider the function
$$\Psi_\alpha:z\rightarrow\frac{z-\alpha}{1-\overline{\alpha}z}$$
an analytic map from D onto D, where D is $\vert{z}\vert<1$.
Then define
$$g=\frac{1}{M}(f\circ\Psi_\alpha^{-1})$$
$$g(0)=\frac{1}{M}f(\alpha)=0$$
Then, by Schwarz Lemma,
$$\forall{z}\in{D}, \vert{g(z)}\vert\leq\vert{z}\vert.$$
Take $z=\Psi_\alpha(z)$ and multiply M to both sides.
$$\vert{Mg(\Psi_\alpha(z))}\vert\leq{M}\vert{\Psi_\alpha(z)}\vert$$
$\rightarrow\vert{f(z)}\vert\leq{M}\vert{\frac{z-\alpha}{1-\overline{\alpha}z}}\vert$, for all $z\in{D}$.

For the second part...

Consider the function
$$g(z)=\frac{f(z)}{M}$$
$\rightarrow\vert{g(z)}\vert\leq{1}$ in $\vert{z}\vert<{1}$.
Define
$\Psi_{\alpha_0}:z\rightarrow\frac{z-\alpha_0}{1-\overline{\alpha_0}z}$ and $h(z)=\Psi_{\alpha_0}(g(z))$
Then $\vert{h(z)}\vert\leq1$ and $h(z_0)=\Psi_{\alpha_0}(g(z_0))=\Psi_{\alpha_0}(\alpha_0)=0$.
Now we can apply the previous inequality to the function h.

$$\vert{h(z)}\vert\leq{1\vert{\frac{z-z_0}{1-\overline{z_0}z}}\vert}$$

$$\rightarrow\vert{(\Psi_{\alpha_0}\circ{g})(z)}\vert\leq{\vert{\frac{z-z_0}{1-\overline{z_0}z}}\vert}$$

$$\rightarrow\vert{\Psi_{\alpha_0}(\frac{f(z)}{M})}\vert\leq{\vert{\frac{z-z_0}{1-\overline{z_0}z}}\vert}$$

$$\rightarrow\vert{\frac{\frac{f(z)}{M}-\alpha_0}{1-\overline{\alpha_0}\frac{f(z)}{M}}}\vert\leq{\vert{\frac{z-z_0}{1-\overline{z_0}z}}\vert}$$

Sub in $\alpha_0=g(z_0)\rightarrow\alpha_0=\frac{f(z_0)}{M}$.

$$\rightarrow\vert{\frac{\frac{f(z)}{M}-\frac{f(z_0)}{M}}{1-\frac{\overline{f(z_0)}}{M}\frac{f(z)}{M}}}\vert\leq{\vert{\frac{z-z_0}{1-\overline{z_0}z}}\vert}$$

$$\rightarrow\vert{\frac{\frac{1}{M}(f(z)-f(z_0))}{\frac{1}{M^2}(M^2-\overline{f(z_0)}f(z))}}\vert\leq{\vert{\frac{z-z_0}{1-\overline{z_0}z}}\vert}$$

$$\rightarrow{M\vert{\frac{f(z)-f(z_0)}{M^2-\overline{f(z_0)}f(z)}}\vert\leq{\vert{\frac{z-z_0}{1-\overline{z_0}z}}\vert}}$$

- - - Updated - - -

Additionally, there is a problem...

Suppose that $f^{(k)}(0)=0$ for k=0,...,N. Show that
$\vert{f(z)}\vert\leq{M\vert{z}\vert^{N+1}}$ for all $z, \vert{z}\vert<1$.

Are those inequalities used to prove this?
 
Your proofs look good. (Yes)

As for your latest question, you don't need those inequalities. A proof similar to the standard proof of the Schwarz lemma will suffice. The conditions on $f$ imply that that there is an analytic function $g$ on $\Bbb D$ such that $f(z) = z^{N+1} g(z)$ for all $z\in \Bbb D$. Since $\lvert f(z)\rvert \le M$ for all $\lvert z\rvert < 1$, then on every circle $\lvert z\rvert = \rho < 1$, $\lvert g(z)\rvert \le M\rho^{-N-1}$. Use the maximum principle, then take the limit as $\rho \to 1^{-}$ to obtain $\lvert g(z)\vert \le M$ for every $z\in \Bbb D$, i.e., $\lvert f(z)\rvert \le M \lvert z\rvert^{N+1}$ for all $z\in \Bbb D$.
 
Euge said:
Your proofs look good. (Yes)

As for your latest question, you don't need those inequalities. A proof similar to the standard proof of the Schwarz lemma will suffice. The conditions on $f$ imply that that there is an analytic function $g$ on $\Bbb D$ such that $f(z) = z^{N+1} g(z)$ for all $z\in \Bbb D$. Since $\lvert f(z)\rvert \le M$ for all $\lvert z\rvert < 1$, then on every circle $\lvert z\rvert = \rho < 1$, $\lvert g(z)\rvert \le M\rho^{-N-1}$. Use the maximum principle, then take the limit as $\rho \to 1^{-}$ to obtain $\lvert g(z)\vert \le M$ for every $z\in \Bbb D$, i.e., $\lvert f(z)\rvert \le M \lvert z\rvert^{N+1}$ for all $z\in \Bbb D$.

Oh okay, it is almost exactly like that proof. Thanks!
 

Similar threads

  • · Replies 8 ·
Replies
8
Views
2K
Replies
1
Views
1K
  • · Replies 2 ·
Replies
2
Views
2K
  • · Replies 7 ·
Replies
7
Views
3K
  • · Replies 1 ·
Replies
1
Views
3K
  • · Replies 8 ·
Replies
8
Views
2K
  • · Replies 14 ·
Replies
14
Views
4K
  • · Replies 2 ·
Replies
2
Views
3K
  • · Replies 7 ·
Replies
7
Views
3K
  • · Replies 14 ·
Replies
14
Views
3K