2
$\begingroup$

Hadamard's Factorization Theorem. Let $f$ be an entire function of order $ρ=1$. Then $f(s)=P(s)exp(Q(s))$, where $P(s)$ is a canonical product with the same zeros as $f$ and $Q$ is a polynomial of degree less than or equal to $ρ=1$. In this case we have $|f(s)|≤_{ε}exp(|s|^{ρ+ε})$ for all $ε>0$.

We have

$f(s)=exp(Q(s))∏_{k=1}^{∞}(((s_{k}-s)/(s_{k})))exp( (s/(s_{k})))$

where $s_{k}$ are the roots of $f$.

Now, consider $h(s)=f(1-2t)$ where $s=\sigma+it$ and $0<t<1$, let Ω=(0,1). The equation $f(1-2t)=0$ implies that $t∈Ω₁$ where $Ω₁$ is a discrete set since the roots are isolated. For the case of $f$ we cannot have infinitely many zeros in $(0,1)$. The main reason here is that a nonzero analytic function can have infinitely many zeros but they must not have an accumulation point in $ℂ$. Now, if there are infinitely many zeros of $f(1-2t)=0$ in $(0,1)$ they do have an accumulation point in $[0,1]$ and since $f(1-2t)$ is extendable to an analytic function to $ℂ$ containing $[0,1]$ this must not happen.

My question is: What is the Hadamard's Factorization of $h$ taking in acaunt that it has a finite number of zeros.

$\endgroup$
5
$\begingroup$

"$h(s)=f(1-2t)$." I take it you are using the standard convention $s=\sigma+it$ with $\sigma$ and $t$ real.

"$f(1-2t)$ is extendable to an analytic function." Since this analytic function agrees with $f(1-2s)$ on an interval, it must be $f(1-2s)$, right? So, $h(s)$ is an entire function defined everywhere by $h(s)=f(1-2s)$, right?

Now if $h(s)$ has finitely many zeros, it must be a polynomial times an entire function with no zeros, right? So $h(s)=P(s)\exp(Q(s))$ where $P$ is a polynomial and $Q$ is entire. And since we are talking order $1$, we must have $$h(s)=P(s)e^{a+bs}$$ for some polynomial $P$ and some constants $a$ and $b$.

EDIT: OP has made some major changes in the question. The function $h$ is now only defined on the interval $(0,1)$. Hadamard factorization is a concept in complex analysis, and does not apply to a function defined only on some interval of real numbers.

No doubt OP will change the question again, but I have had enough.

$\endgroup$

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.