2
$\begingroup$

Recently I have obtained the following:

$$ \prod_{k=1}^{\infty} \left(1 + \frac 1{4\cdot k^2\cdot(4\cdot k-3)}\right)\,\ =\,\ \frac 4\pi $$ or, equivalently, $$ \prod_{k=1}^{\infty} \left(1 - \frac 1{(2\cdot k-1)^2\cdot(4\cdot k+1))}\right)\,\ =\,\ \frac \pi 4 $$

where I have used the Wallis product. (These equations sound so classical that they must be well-known; I'd appreciate a reference.)

QUESTION   Do you know or can you provide simple proofs of the above equation which do not use the Wallis product?

REMARK 1   Perhaps different proofs of Wallis theorem may lead to different proofs of the above formulas (e.g. via Fourier analysis?).

REMARK 2   In general, products $\ \prod_{k=1}^n (1+a_k)\ $ (where $\ a_k\ $ are small) can be treated by the Euler's method by studying $\ \sum_{k=1}^n\log(a_k),\ $ etc.

$\endgroup$
1
  • $\begingroup$ This question got downvoted a couple days ago or so. I am curious what would be the reason? $\endgroup$ – Włodzimierz Holsztyński Jan 19 '17 at 3:21
4
$\begingroup$

Let's consider the partial product $P_n:=\prod_{k=1}^n\frac{(4k+1)(2k-1)^2}{(2k)^2(4k-3)}$. Then, convert to factorials: \begin{align} P_n&=\prod_{k=1}^n\frac{4k+1}{4k-3}\cdot\prod_{k=1}^n\frac{(2k)^2(2k-1)^2}{(2k)^4} \\ &=\frac{4n+1}{2^{4n}}\binom{2n}n^2=(4n+1)\left(\frac{(2n)!}{2^{2n}n!^2}\right)^2. \end{align} Applying Stirling's approximation $n!\sim \sqrt{2\pi n}\,\left(\frac{n}e\right)^n$, we find that \begin{align} P_n \,\,&\sim\,\, (4n+1)\left(\frac{\sqrt{4\pi n}\,\left(\frac{2n}e\right)^{2n}}{2^{2n}(2\pi n)\,\left(\frac{n}e\right)^{2n}} \right)^2 \\ &=\frac{4n+1}{\pi n} \\ &\rightarrow \frac4{\pi}, \end{align} as $n\rightarrow\infty$. The proof follows. Note: this does not accelerate the product any more than the Wallis formula does.

If you like to know a bit more about this and related topics, you may look into this paper:

Tewodros Amdeberhan, Olivier R. Espinosa, Victor H. Moll, Armin Straub, Wallis–Ramanujan–Schur–Feynman, American Mathematical Monthly 117 (2010) pp 618–632, doi:10.4169/000298910X496741 , arXiv:1004.2453 (pdf)

$\endgroup$
3
  • $\begingroup$ T.A., thank you for your answer. Using Stirling's approximation seems less elementary than using Wallis product--in fact, a popular way to obtain the Stirling's theorem (the version which you use) is by applying the Wallis product, then it's simple. *** Thank you for your link too, I'll have a look. $\endgroup$ – Włodzimierz Holsztyński Dec 17 '16 at 8:45
  • $\begingroup$ Nice title of that paper, by the way! $\endgroup$ – theHigherGeometer Aug 20 '19 at 23:57
  • $\begingroup$ @DavidRoberts Thank you! $\endgroup$ – T. Amdeberhan Aug 21 '19 at 14:23

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.