(2) I'm trying to give a combinatorial description of the $$ n^{\mathrm{th}} $$ of

$$ \left(\displaystyle\frac{1}{1-x}\right)\left(\displaystyle\frac{1}{1-x^2}\right)\left(\displaystyle\frac{1}{1-x^3}\right)\cdots \left(\displaystyle\frac{1}{1-x^n}\right) $$

(the coefficient that corresponds to the x^nth term.

Also, I don't even see the purpose of what 1/(1-x) does.

  • $\begingroup$ Your title is missing some words, it seems. $\endgroup$ – Mariano Suárez-Álvarez Mar 27 '12 at 0:55
  • $\begingroup$ I've edited the title. mary, edit it some more if my edit wasn't right. $\endgroup$ – Gerry Myerson Mar 27 '12 at 0:59

Each factor in the product is equal to an infinite geometric series: $$\frac1{1-x^k} = 1 + x^k + x^{2k} + x^{3k} + \cdots$$ Therefore $$\prod \frac1{1-x^j} = (1 + x + x^2 + \cdots)(1 + x^2 + x^4 + \cdots)(1 + x^3 + x^6 + \cdots) \cdots $$ So the coefficient of $x^n$ in the expansion is the number of ways to pick a nonnegative integer, then a nonnegative multiple of two, then a nonnegative multiple of three, and so on, to add up to $n$ (note that it doesn't matter for this purpose that we stop at multiples of $n$). Another way of thinking about "adding a multiple of $k$" is "adding some copies of $k$". Then the statement becomes just the number of ways to write $n$ as a sum of positive integers, up to permutation. For example, with $n = 7$, we could pick $1$ from the first bracket (zero 1s), $x^4$ from the second bracket (two 2s), and $x^3$ from the third bracket (one 3), and $1$ from the rest (no more numbers). This would correspond to the sum $$7 = 2 + 2 + 3.$$ That is, the coefficient of $x^n$ is the number of partitions of $n$.

  • $\begingroup$ and still would be if the product is take up to $\infty$ rather than up to $n$ $\endgroup$ – Henry Mar 27 '12 at 7:50

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.