29

Am I correct or not? I think that a ring of holomorphic functions in one variable is not a UFD, because there are holomorphic functions with an infinite number of $0$'s, and hence it will have an infinite number of irreducible factors! But I am unable to get a concrete example. Please give some example.

glS
  • 7,963
Myshkin
  • 36,898
  • 28
  • 168
  • 346

3 Answers3

35

You are perfectly right: the ring of entire functions $\mathcal O(\mathbb C)$ is not a UFD. Here is why:

In a UFD a non-zero element has only finitely many irreducible (=prime) divisors and this does not hold for our ring $\mathcal O(\mathbb C)$.
Indeed the only primes in $\mathcal O(\mathbb C)$ are the affine functions $z-a$ and on the other hand the function $\sin(z)$ is divided by the infinitely many primes $z-k\pi\; (k\in \mathbb Z) $.

The same proof shows that for an arbitrary domain $D$ the ring $\mathcal O(D)$ is not a UFD, once you know Weierstrass's theorem which implies that there exist non identically zero holomorphic functions in $D$ with infinitely many zeros.

NB
It is sufficient for the proof above to show that the $z-a$ are irreducible in $\mathcal O(\mathbb C)$ (you don't need that there are no other irreducibles). And that is easy: if $a=0$ for example, write $z=fg$ and you will see that $f$ (say) has no zero and is thus a unit $f\in \mathcal O(\mathbb C)^*$.

Georges Elencwajg
  • 156,622
13

Consider $f(z):=\sin(\pi z)=\pi z\prod_{j=1}^{+\infty}\left(1-\frac{z^2}{j^2}\right)$. It's an holomophic function, and the zeroes are the integers. If $f$ could be written as a product of a unit and a finite product of irreducible elements of $\mathcal O(\Bbb C)$, then one of these irreducible elements would have infinitely many roots. It's not possible, for example if $r_n$ are the roots of this elements $g_1$, then $g_1(z)=(z-r_1)h_1(z)$, where $h_1(z)=\prod_{j=2}^{+\infty}(z-r_j)a_j$ and none of these two elements is an unit.

Davide Giraudo
  • 181,608
  • Could you please tell me which theorem allow to write $sin(\pi z)$ is like that?and how that infinite product is equal to $sin(\pi z)$? – Myshkin Jun 04 '12 at 19:55
  • 4
    You can check that this infinite product is uniformly convergent on compact subsets. Write $\sin z=\frac{e^{iz}-e^{-iz}}{2i}$, then use the fact that $e^u=\lim_{n\to \infty}\left(1+\frac un\right)^n$, and factorize the obtained expression: $a^n-b^n=(a-b)\cdots$ – Davide Giraudo Jun 04 '12 at 20:03
  • aha! thanks a lot..pleased. – Myshkin Jun 04 '12 at 20:06
4

Look at $f(z)=\sin(z)$ for example...

Malik Younsi
  • 3,968