22
$\begingroup$

$$\sin x = x\prod_{n=1}^\infty \left[1-\frac{x^2}{n^2\pi^2}\right]$$

If you apply $\log$ to both sides and derivate:

$$\cot x = \frac{1}{x} - \sum_{n=1}^\infty \left[\frac{2x}{n^2\pi^2} \frac{1}{1-\frac{x^2}{n^2\pi^2}}\right]$$

I have to make the expansion:

$$\frac{1}{1-\frac{x^2}{n^2\pi^2}} = 1 + {\left(\frac{x^2}{n^2\pi^2}\right)} + \left(\frac{x^2}{n^2\pi^2}\right)^2 + \left(\frac{x^2}{n^2\pi^2}\right)^3 + \cdots \tag{1}$$

But for this, I'm gonna expand the infinite sum:

$$\cot x = \frac{1}{x} - \left(\frac{2x}{1^2\pi^2} \frac{1}{1-\frac{x^2}{1^2\pi^2}} + \frac{2x}{2^2\pi^2} \frac{1}{1-\frac{x^2}{2^2\pi^2}} + \frac{2x}{3^2\pi^2} \frac{1}{1-\frac{x^2}{3^2\pi^2}}+\cdots\right)$$

So I can understand the boundaries of this expansion (cause the series $\frac{1}{1-x}=1 + x + x^2 + x^3 + \cdots$ as long as $|x|<1$.

So, am I right in saying that I can do the expansion $(1)$ as long as $|x|<\pi$? Because for $x=\pi$ we have: $$\frac{1}{1-\frac{\pi^2}{1^2\pi^2}} = \frac{1}{1-1}$$ And if $|x|<\pi$ the other terms like:

$$\frac{1}{1-\frac{\pi^2}{2^2\pi^2}}$$ Can be expanded by $(1)$ too.

So:

$$\cot x = \frac{1}{x} - \sum_{n=1}^\infty \left[\frac{2x}{n^2\pi^2} \left(1 + {\left(\frac{x^2}{n^2\pi^2}\right)} + \left(\frac{x^2}{n^2\pi^2}\right)^2 + \left(\frac{x^2}{n^2\pi^2}\right)^3 + \cdots \right)\right]\tag{$|x|<\pi$}$$ $$x \cot x = 1 - 2\sum_{n=1}^\infty \left[\frac{x^2}{n^2\pi^2} + \frac{x^4}{n^4\pi^4} + \frac{x^6}{n^6\pi^6} + \frac{x^8}{n^8\pi^8} + \cdots\right]$$

And finally the famous: $$x \cot x = 1 - 2\sum_{n=1}^{\infty} \left[\zeta(2n)\frac{x^{2n}}{\pi^{2n}}\right]\tag{$|x|<\pi$}$$

Am I right with the boundaries for $x$?

|cite|improve this question
$\endgroup$
  • 6
    $\begingroup$ Yes, you're right. For a meromorphic function, the radius of convergence is always given by the distance from the closest singularity, $\pm \pi$ in your case. $\endgroup$ – Jack D'Aurizio Aug 8 '14 at 17:04
  • $\begingroup$ (+1) This is the same as $(1)$ in this answer. $\endgroup$ – robjohn♦ Aug 8 '14 at 22:50
  • $\begingroup$ Is this basically the way to compute every even zeta function? Very nice $\endgroup$ – Yuriy S Mar 13 '16 at 2:45
1
$\begingroup$

Actualy (see [1] pg. 75) $$ \cot(z)=\frac{1}{z}-\sum^{\infty}_{n=1}\frac{(-1)^{n-1}2^{2n}B_{2n}}{(2n)!}z^{2n-1}\textrm{, }|z|<\pi $$ and ([1] pg.807) $$ \zeta(2n)=\frac{(2\pi)^{2n}(-1)^{n-1}}{2(2n)!}B_{2n}\textrm{, }n\in\textbf{N} $$ where $B_{2n}$ are the Bernoulli numbers i.e. $$ \frac{x}{e^x-1}=\sum^{\infty}_{n=0}\frac{B_n}{n!}x^n\textrm{, }|x|<2\pi $$ Your expansion is very right. But unfortunately this is quite known. In [1] you can find more of these results. For example $$ \log\left(\frac{\sin(z)}{z}\right)=-\sum^{\infty}_{n=1}\frac{\zeta(2n)}{n}\frac{z^{2n}}{\pi^{2n}} $$

Actualy for an analytic function in $(-1,1)$ which is also absolutly convergent at 1 we have $$ \sum^{\infty}_{k=1}\left(f\left(\frac{x}{2\pi i k}\right)+f\left(\frac{-x}{2\pi i k}\right)-2f(0)\right)=2\sum^{\infty}_{k=1}\frac{f^{(2k)}(0)}{(2k)!}\frac{B_{2k}}{(2k)!}x^{2k}\textrm{, }|x|<2\pi $$

[1]: Milton Abramowitz and Irene A. Stegun. "Handbook of Mathematical functions". Dover publications, inc., New York. (1972)

|cite|improve this answer
$\endgroup$
  • 1
    $\begingroup$ With $\frac{B_k}{k!}$ defined as the coefficients of $\frac{x}{e^x-1}$ note that $\zeta(2k)=\frac{(2\pi)^{2k}(-1)^{k-1}}{2(2k)!}B_{2k}$ implies $x \cot x = 1 - 2\sum_{k=1}^{\infty} \left[\zeta(2k)\frac{x^{2k}}{\pi^{2k}}\right] = 1- x\sum_{n=1}^\infty \left[\frac{2x}{n^2\pi^2} \frac{1}{1-\frac{x^2}{n^2\pi^2}}\right]$ and $\sin x = x\prod_{n=1}^\infty \left[1-\frac{x^2}{n^2\pi^2}\right]$ $\endgroup$ – reuns Jul 21 at 7:10
  • $\begingroup$ @ renus. Yes indeed. $\endgroup$ – Nikos Bagis Jul 22 at 8:41

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.