I have to prove that $$ \forall n \in \mathbb{N}_{0}, ~ \forall x \in \mathbb{R} \setminus \mathbb{N}_{0}: \qquad \sum_{k = 0}^{n} \frac{(-1)^{k} \binom{n}{k}}{x + k} = \frac{n!}{(x + 0) (x + 1) \cdots (x + n)}. $$ However, I was unable to find a proof. I have tried to use the binomial expansion of $ (1 + x)^{n} $ to get the l.h.s., by performing a suitable multiplication followed by integration, but I was unable to obtain the required form. Please help me out with the proof. Thanks in advance.

  • 1
    $\begingroup$ I believe the right side should be $n! / [x(x+1) \cdots (x+n)]$. This actually does hold for $n = 0$. This appears to be a generalization of the partial fraction expansion $1/x(x+1) = 1/x - 1/(x+1)$. $\endgroup$ – Yakov Shklarov Sep 3 '16 at 2:26
  • $\begingroup$ You need another factor of $x$ in the denominator of the righthand side. $\endgroup$ – Brian M. Scott Sep 3 '16 at 2:27
  • $\begingroup$ (For the record, I typed out my answer before I saw @Winther's comment.) $\endgroup$ – arctic tern Sep 3 '16 at 2:51
  • 1
    $\begingroup$ Note that the domain of both expressions is $ \mathbb{R} \setminus \mathbb{N}_{0} $, not $ \mathbb{R} $. $\endgroup$ – Transcendental Sep 4 '16 at 0:22

Write down the partial fraction decomposition

$$\frac{1}{x(x+1)\cdots(x+n)}=\sum_{k=0}^n \frac{c_k}{x+k}. \tag{$\circ$}$$

To determine the coefficient $c_k$, multiply by $x+k$ then compute the limit $x\to -k$, getting

$$c_k=\frac{1}{(-k)(1-k)(2-k)\cdots (-2)\cdot (-1)\cdot 1\cdot2\cdots(n-k)}=\frac{(-1)^k}{k!(n-k)!}$$

Substitute this into $(\circ)$, muliply both sides by $n!$, then use $\binom{n}{k}=\frac{n!}{k!(n-k)!}$ to conclude.


\begin{align*} (1-t)^n = \sum_{k=0}^n (-1)^k \binom{n}{k} t^k \end{align*} Multiplying by $t^{x-1}$ and integrating between 0 and 1, we get \begin{align*} \int_0^1 t^{x-1}(1-t)^n dt = \sum_{k=0}^n (-1)^k \binom{n}{k}\frac{1}{x+k} \end{align*} and the left hand side is \begin{align*} \beta(x,n+1) &= \frac{\Gamma(x)\Gamma(n+1)}{\Gamma(x+n+1)} \end{align*} This simplifies to \begin{align*} \frac{n!}{x(x+1)(x+2)\ldots(x+n)} \end{align*}

  • $\begingroup$ @YakovShklarov, Yes, you are correct. $\endgroup$ – user348749 Sep 3 '16 at 2:35

An induction on $n$ will work. Let


and for the induction step suppose that



$$\begin{align*} \frac{(n+1)!}{x(x+1)\ldots(x+n+1)}&=\frac{n+1}{x+n+1}f(x,n)\\ &=\frac{n+1}{x+n+1}\sum_{k=0}^n\frac{(-1)^k}{x+k}\binom{n}k\\ &=(n+1)\sum_{k=0}^n(-1)^k\binom{n}k\frac1{(x+k)(x+n+1)}\\ &=(n+1)\sum_{k=0}^n\frac{(-1)^k}{n+1-k}\binom{n}k\left(\frac1{x+k}-\frac1{x+n+1}\right)\\ &=(n+1)\sum_{k=0}^n\frac{(-1)^k}{n+1}\binom{n+1}k\frac1{x+k}\\ &\qquad-\frac{n+1}{x+n+1}\sum_{k=0}^n\frac{(-1)^k}{n+1}\binom{n+1}k\\ &=\sum_{k=0}^n\frac{(-1)^k}{x+k}\binom{n+1}k-\frac1{x+n+1}\color{red}{\sum_{k=0}^n(-1)^k\binom{n+1}k}\\ &=f(x,n+1)-\frac{(-1)^{n+1}}{x+n+1}-\frac1{x+n+1}\color{red}{\left(0-(-1)^{n+1}\right)}\\ &=f(x,n+1)\;, \end{align*}$$

as desired.

  • $\begingroup$ There is a question at this MSE link that would appear to be a match to your activity profile. $\endgroup$ – Marko Riedel Sep 3 '16 at 23:02

I thought it might be instructive to present a proof by induction.

First, we establish a base case. For $n=0$, it is straightforward to show that the expression holds.

Second, we assume that for some $n\geq 0$, we have

$$\sum_{k=0}^n \binom{n}{k}\frac{(-1)^k}{x+k}=n!\prod_{k=0}^n\frac{1}{x+k}$$

Third, we analyze the sum $\sum_{k=0}^{n+1} \binom{n+1}{k}\frac{(-1)^k}{x+k}$. Clearly, we can write

$$\begin{align} \sum_{k=0}^{n+1} \binom{n+1}{k}\frac{(-1)^k}{x+k}&=\frac1x+\sum_{k=1}^{n} \binom{n+1}{k}\frac{(-1)^k}{x+k}+\frac{(-1)^{n+1}}{x+n+1}\\\\ &=\color{green}{\frac1x}+\sum_{k=1}^{n} \left(\color{green}{\binom{n}{k}}+\color{blue}{\binom{n}{k-1}}\right)\frac{(-1)^k}{x+k}+\color{red}{\frac{(-1)^{n+1}}{x+n+1}}\\\\ &=\color{green}{n!\prod_{k=0}^n\frac{1}{x+k}}+\color{blue}{\sum_{k=0}^{n-1} \binom{n}{k}\frac{(-1)^{k+1}}{(x+1)+k}}+\color{red}{\frac{(-1)^{n+1}}{x+n+1}}\\\\ &=\color{green}{n!\prod_{k=0}^n\frac{1}{x+k}}+\color{blue}{-n!\prod_{k=0}^n\frac{1}{x+1+k}-\frac{(-1)^{n+1}}{x+n+1}}+\color{red}{\frac{(-1)^{n+1}}{x+n+1}}\\\\ &=(n+1)!\prod_{k=0}^{n+1}\frac{1}{x+k} \end{align}$$

And we are done!


Remark. What follows is an exact duplicate of an earlier post of mine which I am unable to locate despite making a considerable effort.

We seek to verify that

$$\sum_{k=0}^n \frac{1}{x+k} (-1)^k {n\choose k} = n! \times \prod_{q=0}^n \frac{1}{x+q}.$$

Consider the function

$$f(z) = n! \frac{1}{x+z} \prod_{q=0}^n \frac{1}{z-q}$$

where clearly $x$ is not equal to $0,-1,-2,\ldots -n$ or the LHS of the target formula becomes singular.

We compute the residues of $f(z).$ We get for the poles at $p\in [0,n]$ the residue

$$\mathrm{Res}_{z=p} f(z) = n!\frac{1}{x+p} \prod_{q=0}^{p-1} \frac{1}{p-q} \prod_{q=p+1}^n \frac{1}{p-q} \\ = n! \frac{1}{x+p} \frac{1}{p!} \frac{(-1)^{n-p}}{(n-p)!} = \frac{1}{x+p} (-1)^{n-p} {n\choose p}.$$

The residue at $z=-x$ yields

$$\mathrm{Res}_{z=-x} f(z) = n! \prod_{q=0}^n \frac{1}{-x-q} = n! (-1)^{n+1} \times \prod_{q=0}^n \frac{1}{x+q}.$$

The residue at infinity is

$$\mathrm{Res}_{z=\infty} f(z) = - \mathrm{Res}_{z=0} \frac{1}{z^2} f(1/z) \\ = - \mathrm{Res}_{z=0} \frac{1}{z^2} n! \frac{1}{x+1/z} \prod_{q=0}^n \frac{1}{1/z-q} \\ = - \mathrm{Res}_{z=0} \frac{1}{z^2} n! \frac{z}{zx+1} \prod_{q=0}^n \frac{z}{1-qz} \\ = - \mathrm{Res}_{z=0} \frac{z^{n+2}}{z^2} n! \frac{1}{zx+1} \prod_{q=0}^n \frac{1}{1-qz} \\ = - \mathrm{Res}_{z=0} z^n n! \frac{1}{zx+1} \prod_{q=0}^n \frac{1}{1-qz} = 0.$$

To conclude observe that the residues sum to zero and collecting everything we obtain

$$\sum_{p=0}^n\frac{1}{x+p} (-1)^{n-p} {n\choose p} + n! (-1)^{n+1} \times \prod_{q=0}^n \frac{1}{x+q} = 0.$$

This is

$$\sum_{p=0}^n\frac{1}{x+p} (-1)^{p} {n\choose p} - n! \times \prod_{q=0}^n \frac{1}{x+q} = 0.$$


$$\sum_{p=0}^n\frac{1}{x+p} (-1)^{p} {n\choose p} = n! \times \prod_{q=0}^n \frac{1}{x+q}$$

which is the claim.


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.