2
$\begingroup$

Let $m,n,r\in\mathbb{N}\cup\{0\}.$ I am interested in finding a closed form for the sum

$$\sum_{i=0}^m{{n+i}\choose{r+i}}.$$

Let $f(m,n,r)$ denote the above sum.

We may make a few trivial observations. For example, $f(0,n,r)={{n}\choose{r}}.$ We also have stuff like $f(m,n,r)=0$ if $n<r$ and $f(m,n,r)=m+1$ if $n=r.$

Is there a closed form for this? By "closed form" I mean I'm looking for an expression for $f(m,n,r)$ in terms of $m,n,$ and $r$ that uses something like binomial coefficients, or the like.

Many times such a closed form doesn't exist, but I think one might just exist here. That's because we have closed forms for stuff like $\sum_{i=0}^n{{n}\choose{i}}$ and $\sum_{i=0}^n{{i}\choose{r}},$ and my sum is just a "combination" of these two.

Maybe there's some slick combinatorial argument for this? Even if there isn't a combinatorial argument that gives a closed form, I'd be interested in a combinatorial interpretation of the above sum.

$\endgroup$
2
  • 1
    $\begingroup$ In this book there are algorithms to both determine or proof that no closed form exists, for certain precise definition of closed form, that likely covers what you intend. These algorithms are implemented in many software. For example this one From the result it looks like if you multiply by $n-r+1$, you can turn it into a telescopic sum of $(r+k+1)\binom{n+k+1}{r+k+1}-(r+k)\binom{n+k}{r+k}$. $\endgroup$
    – ameg
    Commented May 28 at 17:01
  • 3
    $\begingroup$ Use the hockey stick identity to call this $\binom{n+m+1}{r+m} - \binom{n}{r-1}$ by recognizing your sum merely as a "hockeystick" missing the top of the handle which itself could be represented as its own "hockeystick" $\endgroup$
    – JMoravitz
    Commented May 28 at 17:08

1 Answer 1

5
$\begingroup$

It's always a good idea to try writing $\binom{a+i}{b+i}$ as $\binom{a+i}{b-a}$ in a sum over $i$; that way, only one parameter of the binomial coefficient varies with $i$. By doing that here, and then re-indexing the sum, we can put it into one of the forms mentioned in the question: $$\sum_{i=0}^m \binom{n+i}{r+i} = \sum_{i=0}^m \binom{n+i}{n-r} = \sum_{j=n}^{n+m} \binom j{n-r}.$$ This is not the same as the sum as $j$ goes from $0$ to $n+m$, which would be $\binom{n+m+1}{n-r+1}$. However, we can write an arbitrary sum like this as a difference of two partial sums starting at $0$: $$ \sum_{j=n}^{n+m} \binom{j}{n-r} = \sum_{j=0}^{n+m} \binom{j}{n-r} - \sum_{j=0}^{n-1} \binom j{n-r} = \binom{n+m+1}{n-r+1} - \binom{n}{n-r+1}. $$

$\endgroup$

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .