2

My question is about this proof here about Stirling's formula:

A proof I found a while ago entirely relies on creative telescoping. Since $\frac{1}{n^2}-\frac{1}{n(n+1)}=\frac{1}{n^2(n+1)}$,

$$\begin{eqnarray*} \sum_{n\geq m}\frac{1}{n^2}&=&\sum_{n\geq m}\left(\frac{1}{n}-\frac{1}{(n+1)}\right)+\frac{1}{2}\sum_{n\geq m}\left(\frac{1}{n^2}-\frac{1}{(n+1)^2}\right)\\&+&\frac{1}{6}\sum_{n\geq m}\left(\frac{1}{n^3}-\frac{1}{(n+1)^3}\right)-\frac{1}{6}\sum_{n\geq m}\frac{1}{n^3(n+1)^3}\tag{1}\end{eqnarray*} $$ hence: $$ \psi'(m)=\sum_{n\geq m}\frac{1}{n^2}\leq \frac{1}{m}+\frac{1}{2m^2}+\frac{1}{6m^3}\tag{2}$$ and in a similar fashion: $$ \psi'(m) \geq \frac{1}{m}+\frac{1}{2m^2}+\frac{1}{6m^3}-\frac{1}{30m^5}.\tag{3}$$ Integrating twice, we have that $\log\Gamma(m)$ behaves like: $$ \log\Gamma(m)\approx\left(m-\frac{1}{2}\right)\log(m)-\color{red}{\alpha} > m+\color{blue}{\beta}+\frac{1}{12m}\tag{4}$$ where $\color{red}{\alpha=1}$ follows from $\log\Gamma(m+1)-\log\Gamma(m)=\log m$.

That gives Stirling's inequality up to a multiplicative constant.

$\color{blue}{\beta=\log\sqrt{2\pi}}$ then follows from Legendre's duplication formula and the well-known identity:

$$ \Gamma\left(\frac{1}{2}\right)=2 \int_{0}^{+\infty}e^{-x^2}\,dx = \sqrt{\pi}.\tag{5}$$

I understand how they got to $\frac{1}{m}+\frac{1}{2m^2}+\frac{1}{6m^3}-\frac{1}{30m^5}\leq\psi′\leq\frac{1}{m}+\frac{1}{2m^2}+\frac{1}{6m^3}$, but how do I proceed from there? Do I integrate all 3 sides, or do I approximate $\psi′\approx\frac{1}{m}+\frac{1}{2m^2}+\frac{1}{6m^3}$ and integrate that?

And isn't one of the constants $\alpha$ or $\beta$ supposed to be multiplied by $m$ because of the double integration? I also don't quite understand how they got the values of $\alpha$ and $\beta$.

Thanks in advance.

D.R.
  • 8,691
  • 4
  • 22
  • 52
  • It is pretty much the same thing. Once you know that $$\psi'(m)=\frac{1}{m}+\frac{1}{2m^2}+\frac{1}{6m^3}+O\left(\frac{1}{m^5}\right)$$, you can just integrate both sides of such identity twice. And the values of $\alpha$ and $\beta$ are derived... exactly like it is written. – Jack D'Aurizio Sep 04 '17 at 19:20
  • Integrating $\mathcal{O}(x^{-5})$ isn't entirely obvious, as $\int_x^\infty \mathcal{O}(y^{-5})dy=\mathcal{O}(\int_x^\infty y^{-5}dy) = \mathcal{O}(x^{-4})$ but $\int_1^x \mathcal{O}(y^{-5})dy = \mathcal{O}(1)$ @JackD'Aurizio – reuns Sep 04 '17 at 23:01
  • @reuns: but the integral to perform is $\int_{M}^{M+1}(\ldots),dm$ or $\int_{M}^{+\infty}(\ldots),dm$, so there is no issue. – Jack D'Aurizio Sep 04 '17 at 23:07
  • @JackD'Aurizio If we can just write $\psi'=\frac{1}{m}+\frac{1}{2m^2}+\frac{1}{6m^3}+O\left(\frac{1}{m^5}\right)$, why did you write the inequalities above? What are they for? – D.R. Sep 05 '17 at 00:11
  • To ensure you are allowed to use that O symbol, of course. – Jack D'Aurizio Sep 05 '17 at 01:36

0 Answers0