I have to deal a lot with functions which are $_3F_2$ with unit argument, and I need to find their behavior for large values of the parameters. One example is


where I'm interested in the behavior of large $n$. Now, I have played around with this function numerically a bit and it looks to me that asymptotically (for $n \to \infty$) it behaves as

$$\frac{1}{2\sqrt{\pi}}4^n n^{1/2-x} \Gamma(x)$$

However, this is of course very dirty and I would like to have a clean way to get this kind of asymptotics. Is there a way to do this or are there some references with similar results?


The formula is correct if $\pi$ is changed to $\sqrt \pi$. For this particular example, we have $${_3\hspace{-1px}F_2}(n, n, n; 2 n, n + x; 1) = \frac {\Gamma(x) \Gamma(2 n)} {\Gamma(n) \Gamma(n + x)} \,{_3\hspace{-1px}F_2}(n, x, x; n + x, n + x; 1).$$ Now the $k$th term in the hypergeometric sum is of order $n^{-k}$. For $n \to \infty$ and $x$ fixed, $${_3\hspace{-1px}F_2}(n, x, x; n + x, n + x; 1) \sim 1, \\ \frac {\Gamma(x) \Gamma(2 n)} {\Gamma(n) \Gamma(n + x)} \sim \frac {\Gamma(x) n^{1/2 - x} 2^{2 n - 1}} {\sqrt \pi}.$$

  • $\begingroup$ yes the missing square root was a typo. Thanks a lot! $\endgroup$ – bnado Oct 12 '18 at 8:50

Your Answer

By clicking "Post Your Answer", you acknowledge that you have read our updated terms of service, privacy policy and cookie policy, and that your continued use of the website is subject to these policies.

Not the answer you're looking for? Browse other questions tagged or ask your own question.