# Partitions of an integer into k parts.

I am interested in knowing whether an exact formula (analogous to the Hardy-Ramanujan-Rademacher formula for $p(n)$) for the number of partitions of a positive integer into k parts is known. I tried searching on the internet but could not find such a formula. Can someone confirm whether such a formula is known and provide a link to its proof if possible.

Thanks

-
You want $k$ equal parts? What about $[k\mid n]$ (using the Iverson bracket). –  Marc van Leeuwen Nov 19 '12 at 9:26
No I do not form not equal parts. I am looking for the analogous formula for number of compositions in k parts which is $\binom{n-1}{k-1}$. –  Shahab Nov 19 '12 at 9:31
You have a cryptic way of expressing yourself. Saying you "do not form not equal parts" means you do form equal parts only. I dont't think that is what you mean, but it is what you say. And the question says equal parts too. –  Marc van Leeuwen Nov 19 '12 at 9:37
I'm sorry. I do not want equal parts. That was a mistake in typing. –  Shahab Nov 19 '12 at 9:45

In Chapter 6 of Andrews and Eriksson, Integer Partitions, formulas are given for $k=1,2,3,4,5$. Full proofs are given for $k=1,2,3,4$, while $k=5$ is left as an exercise.
Actually, the function studied in that text is a bit different from what you want, but it's easy to get back and forth between the two functions. The number of partitions into (exactly) $k$ parts equals the number of partitions with greatest part $k$. The text studies the number of partitions with greatest part at most $k$. But partitions of $n$ with greatest part $k$ correspond to partitions of $n-k$ with greatest part at most $k$.
Thank you. Is a formula for a general $k$ known on the lines of the analogous formula for compositions $\binom{n-1}{k-1}$? –  Shahab Nov 22 '12 at 2:51
Well, it seems unlikely. They have to use different methods for different $k$, and the formulas look very different, and they don't say anything about there being a general formula. I deduce from that that there isn't a general formula, but I recognize that's not a proof. –  Gerry Myerson Nov 22 '12 at 11:51