For $n\in\mathbb{N}$ let $p_n$ be a polynomial of degree $n$. Suppose there exists $c>0$ such that

$\bullet$ if $z\in\mathbb{C}$ is a zero of a $p_n$, then $|z^2+c|\leq c$ (note that in particular the $p_n$'s have no real positive zeros)

$\bullet$ $p_n(x)\xrightarrow[n\to\infty]{}f(x)$ for all $x>0$, where $f$ is an analytic function

Can we conclude that $f$ is never zero on $\mathbb{R^+}$?

I think Vitali's theorem should play a role, but I don't see precisely how.

  • $\begingroup$ Is $c$ dependent on $n$? $\endgroup$ – 23rd Apr 11 '13 at 18:02
  • $\begingroup$ yes $c$ is the same for every $n$ $\endgroup$ – qwertyuio Apr 11 '13 at 18:25
  • $\begingroup$ If so, I think $f\equiv 0$. $\endgroup$ – 23rd Apr 11 '13 at 18:28
  • $\begingroup$ why? $p_n$ is never zero on $\mathbb{R}^+$ $\endgroup$ – qwertyuio Apr 11 '13 at 18:37
  • $\begingroup$ Would you mind if I post an answer for proving $f\equiv 0$? $\endgroup$ – 23rd Apr 11 '13 at 18:41

Let $R=\sqrt{2c}>0$. Write $p_n$ as $p_n(z)=a_n\cdot\prod_{i=1}^n(z-z_{n,i})$. Then for $i=1,\dots, n$, from $|z_{n,i}^2+c|\le c$ we know that $|z_{n,i}|\le R$. Fix an integer $k\ge 1$. It follows that when $|z|\le kR$, $|p_n(z)|\le |a_n|\cdot[(k+1)R]^n$; when $|z|\ge (2k+3)R$, $|p_n(z)|\ge |a_n|\cdot[2(k+1)R]^n$. In particular, if $0<x\le kR<(2k+3)R\le y$, then $|p_n(x)|\le \frac{|p_n(y)|}{2^n}$. Letting $n\to\infty$, it follows that $f(x)=0$ for every $0<x\le kR$. Since $k$ is arbitrary, $f\equiv 0$ on $\mathbb{R}^+$.

  • $\begingroup$ I read your argument and I don't find an error, but I'have a counter-example! $\endgroup$ – qwertyuio Apr 11 '13 at 19:48
  • $\begingroup$ @qwertyuio: Could you please show me your example? $\endgroup$ – 23rd Apr 11 '13 at 19:57
  • $\begingroup$ It's a bit complicated because it descends from a theorem and so it would require some effort to put it in an explicit form. So I prefer to read more carefully your proof $\endgroup$ – qwertyuio Apr 11 '13 at 20:01
  • $\begingroup$ @qwertyuio: OK. $\endgroup$ – 23rd Apr 11 '13 at 20:06
  • $\begingroup$ I'm terribly sorry. Your proof is correct. My functions $p_n$ are not polynomials but rational functions.. $\endgroup$ – qwertyuio Apr 11 '13 at 20:07

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.