5
$\begingroup$

I am trying to prove that $$f(n)=4^n\frac{\dbinom{2n}{n}}{\dbinom{4n}{2n}}$$ converges as $n\rightarrow\infty$.

I have already tried to use the fact that, if $n, k \in\mathbb{N}, n\geq k\geq1,$ then $(\frac{n}{k})^k\leq\binom{n}{k}\leq e^k(\frac{n}{k})^k$. However, this has been to no avail.

Could anybody suggest how I could approach this proof?

$\endgroup$
  • 3
    $\begingroup$ You could use Stirling's Approximation to estimate the central binomial coefficient. $\endgroup$ – r9m Apr 9 '15 at 11:47
6
$\begingroup$

This can be done directly, without Stirling. Note that after the simplification of the factorials the sequence becomes $$ f_n = \frac{2n+2}{2n+1}\frac{2n+4}{2n+3}...\frac{4n}{4n-1} $$

Taking $\log$ we see that $$\log f_n = \log\left(1+\frac{1}{2n+1}\right)+...+\log\left(1+\frac{1}{4n-1}\right)$$ Using the fact that $\log (1+x) \simeq x$ for $x$ close to $0$, we have

$$ \log f_n \simeq \frac{1}{2n+1}+...+\frac{1}{4n-1}$$

Now denote $a_n = \displaystyle \frac{1}{n+1}+...+\frac{1}{2n}$. Using partial sums of a Riemann integral, or the logarithm approximation, we find that $a_n \to \ln 2$.

Now just note that $\displaystyle \frac{1}{2n+1}+...+\frac{1}{4n-1} = a_{2n}-\frac{a_n}{2}$. Therefore $$ \log f_n \to \frac{\ln 2}{2}$$ which implies that $f_n \to \sqrt{2}$.

$\endgroup$
  • 2
    $\begingroup$ (+1) 'without stirling' .. its just another application of partial summation formula on $H_n$. $\endgroup$ – r9m Apr 9 '15 at 12:49
5
$\begingroup$

Hint: Use the asymptotic properties of the Central Binomial Coefficient to find $\displaystyle\lim_{n\to\infty} f(n)$

$$n\to\infty\implies \binom{2n}{n}\sim \frac{4^n}{\sqrt{\pi n}}\quad\textrm{and}\quad\binom{4n}{2n}\sim\frac{4^{2n}}{\sqrt{2\pi n}}$$

The limit will come out as $\sqrt 2$ proving that the limit converges.

$\endgroup$
  • 1
    $\begingroup$ I suppose that you have a typo in the first approximation : the $2$ should not be present in the radical. $\endgroup$ – Claude Leibovici Apr 9 '15 at 12:01
  • $\begingroup$ @ClaudeLeibovici, It was a typo actually. Fixed it now. Thanks for pointing it out :) $\endgroup$ – Prasun Biswas Apr 9 '15 at 12:02
  • $\begingroup$ ...which is the direct consequence of using Stirling's formula $\endgroup$ – Alex Apr 9 '15 at 13:20
  • $\begingroup$ @Alex, yes, I know that. I just wrote it like that since this is also a pretty well-known result. $\endgroup$ – Prasun Biswas Apr 9 '15 at 13:21
3
$\begingroup$

Inequality $(9)$ from this answer shows that $$ \frac{4^n}{\sqrt{\pi(n+\frac13)}}\le\binom{2n}{n}\le\frac{4^n}{\sqrt{\pi(n+\frac14)}} $$ Therefore, $$ \sqrt{\frac{2n+\frac14}{n+\frac13}}\le4^n\frac{\dbinom{2n}{n}}{\dbinom{4n}{2n}}\le\sqrt{\frac{2n+\frac13}{n+\frac14}} $$ The Squeeze Theorem says that $$ \lim_{n\to\infty}4^n\frac{\dbinom{2n}{n}}{\dbinom{4n}{2n}}=\sqrt2 $$

$\endgroup$

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.