Show that there exists a unique $z_0\in B(0,1)$ s.t $f(z_0)=z_0$.

323 Views Asked by At

Let $\varepsilon >0$ and let $f:B(0,1+\varepsilon )\rightarrow B(0,1)$ be a holomorphic function. Show that there exists a unique $z_0\in B(0,1)$ s.t $f(z_0)=z_0$.

For this problem, I don't want to use the Earle-Hamilton fixed-point Theorem because my complex-analysis textbook doesn't have it, which implies the uniqueness of $z_0$ straightforwardly.

However, what I'm considering instead is the maximum modulus principle (or Schwarz Lemma or Rouche's Theorem).

Here's my idea.

Since $B(0,1)$ is bounded, $\exists \alpha \in B(0,1+\varepsilon )$ s.t $|f(z)| \leq |f(\alpha)| <1$ for any $z \in B(0,1+\varepsilon )$.

Then, by the Maximum modulus principle, $f$ is constant.

So, $f(z)=z_0$ for any $z\in B(0,1+\varepsilon )$ and for some $z_0 \in B(0,1)$.

Since, $z_0 \in B(0,1) \subset B(0,1\varepsilon )$, $f(z_0)=z_0$.

Lastly, the uniqueness of $z_0$ is obvious because for any $z \in B(0,1) \setminus \{z_0\}$, $z \neq z_0$. $\blacksquare $

How does this sound like?

I have a little doubt due to the fact that $f$ is constant.

Another question is this. If we want to use the Schwarz Lemma for this problem, then we would need the condition $f(0)=0$. But, since $f$ might not satisfy this, I wonder how we should modify this Lemma? I think we can let $g(z)=f(z)-f(0)$ so that the Lemma can be used. Would this method work?

Update:

I'm also trying to use Rouche's Theorem, comparing $f(z)$ and $g(z)=z$ on $\partial B(0,1)$.

But, I am stuck at showing $|f(z_0)-z_0| \neq |f(z_0)|+1$ for any $z_0 \in \partial B(0,1)$. How this could be shown?

Update:

I finally could get a simple solution to this problem using Rouche's Theorem.

Note that $\overline{B(0,1)}\subseteq B(0,1+\varepsilon )$.

Now, we are going to compare $|f(z)|$ and $|-z|$ on $\partial B(0,1)$.

As $f(z) \in B(0,1)$, $|f(z)|<1$ for any $z \in \partial B(0,1)$. Also, $|-z|=1$ for any $z \in \partial B(0,1)$.

So, $|f(z)|<|-z|$ on $\partial B(0,1)$.

By a Corollary derived from Rouche's Theorem, we can say that $-z$ and $f(z)-z$ have the same number of zeros in $B(0,1)$.

Then, this implies that $f(z)-z = 0$ at a unique $z_0 \in B(0,1)$.

Hence, we conclude that there exists unique $z_0 \in B(0,1)$ such that $f(z_0)=z_0$.

However, I still want to know if it is possible to use Schwarz Lemma for this problem.

Any help would be really appreciated.