Hypergeometric representation degenerate cases

75 Views Asked by At

In Concrete Mathematics, section 5.5, the book discussed degenerate cases for hypergeometric series. But slightly later on, after establishing the gamma function, $\Gamma(z+1)=z!$, $(-z)!\Gamma(z)=\frac{\pi}{sin (\pi z)}$, generalised rising and falling powers, $z^{\underline{w}}=$ and the limit definition for binomial coefficients $\binom{z}{w}=\lim_{\zeta \rightarrow z}\lim_{\omega \rightarrow w}\frac{\zeta !}{\omega !(\zeta - \omega)!}$ (to be used whenever $\frac{z!}{w!(z-w)!}$ is undefined), the book goes on to use factorial cancellation liberally without considering any degenerate cases.

Factorial canceling on expansion of binomial coefficients on Concrete Mathematics (This links to a related qn)

To calculate a hypergeometric representation of a series $\sum_{k\geq 0}t_k$, we consider the term ratio $\frac{t_{k+1}}{t_k}$, so my definition for a degenerate case would be:

  1. Whenever $\binom{z}{w}\neq \frac{z!}{w!(z-w)!}$, i.e. both numerator and denominator $\rightarrow \infty$, in this case we need to use the limit definition.

  2. Whenever $t_{k+1} = \infty$ and $\frac{1}{t_k} = 0$, this can occur when there is a negative factorial in both expressions.

The book goes on to transform Vandermonde's convolution into a hypergeometric. $$\begin{align}&\sum_k\binom{r}{k}\binom{s}{n-k}=\binom{r+s}{n}\\ &t_k=\frac{r!}{(r-k)!k!}\frac{s!}{(s-n+k)!(n-k)!}\\ &\frac{t_{k+1}}{t_k}=\frac{k-r}{k+1}\frac{k-n}{k+s-n+1}\\ &\binom{s}{n}F(-r,-n;s-n+1|1)=\binom{r+s}{n},\, n\in \mathbb{Z}^+ \end{align}$$ The book seems to imply that the last equality holds for any real $r,s$, which seems problematic to me. For e.g. when integer $s<n$, the hypergeometric formula isn't even defined. The book spoke previously of considering the limit $$\lim_{\epsilon \rightarrow 0}F(-r,-n;s-n+1+\epsilon|1)$$ in such cases, but in this section, the book didn't make any mention of degenerate cases. There are also values of $r,s$ that lead to degenerate situations of the 2 points I raised above.

So how should I interpret this portion of the book?