Mathematics Stack Exchange is a question and answer site for people studying math at any level and professionals in related fields. Join them; it only takes a minute:

Sign up
Here's how it works:
  1. Anybody can ask a question
  2. Anybody can answer
  3. The best answers are voted up and rise to the top

Let $\{ r_1, \dots, r_n \}$ and $\{ s_1, \dots, s_m \}$ be two sets of positive integers. If $\sum_{i = 1}^{n} r_{i} = \sum_{i = 1}^{m} s_{i}$ and $\prod_{i = 1}^{n}(r_i + 1) = \prod_{i = 1}^{m} (s_i + 1)$, does it necessarily follow that $n = m$? More generally, given the equalities, does it follow that there is a permutation $\pi \in S_{n}$ such that $\pi( \{ r_1, \dots, r_n \}) = \{ s_1, \dots, s_m \}$?

Perhaps symmetric functions play a non-trivial role here.

Update: As stated, neither question above can be answered in the affirmative (See Prof. Israel's counter-example). Let $e_k(x_1, \dots, x_n)$ denote the $k^{\text{th}}$-elementary symmetric polynomial of $n$ indeterminates.

Is there a positive integer $l < n$ such that if $e_{k}(r_1, \dots, r_n) = e_{k}(s_1, \dots, s_m)$ for $0 \leqslant k \leqslant l$ and $\prod_{i = 1}^{n}(r_i + 1) = \prod_{i = 1}^{m} (s_i + 1)$, then $n = m$ and $e_{k}(r_1, \dots, r_n) = e_{k}(s_1, \dots, s_m)$ for $0 \leqslant k \leqslant n$, or more generally, there is a permutation $\pi \in S_{n}$ such that $\pi( \{ r_1, \dots, r_n \}) = \{ s_1, \dots, s_m \}$?

share|cite|improve this question

No, it's not true. For example, try $\{1, 2, 14\}$ and $\{8, 9\}$

share|cite|improve this answer
Thank you for the counter-example. I posted a new follow-up question. – user02138 Aug 14 '12 at 4:55

For any $\ell$, there are disjoint sets $\{{r_1,\dots,r_n\}}$ and $\{{s_1,\dots,s_n\}}$ of positive integers such that $$\sum r_i^k=\sum s_i^k,\quad 1\le k\le\ell$$ Information can be found under the headings "multigrade equations", or "Tarry-Escott problem". It follows that the polynomials $\prod(x-r_i)$ and $\prod(x-s_i)$ will have identical coefficients from $x^n$ down to $x^{n-\ell}$, which is to say, the first $\ell$ elementary symmetric functions will agree. It's conjectured that one can take $n$ as small as $\ell+1$, but only proved that one can take $n$ to be on the order of $\ell^2$.

I know this doesn't account for the condition on $\prod(r_i+1)$, but there are so many solutions to the multigrade equation I suspect that at worst this would push $n$ up by 1.

share|cite|improve this answer

Your Answer


By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.