What is the number of equivalence classes of $\{0,\ldots,n\}^m/\mathrm S_m$?
where $\mathrm S_m$ is the symmetric group and the equivalence relation is (obviously?) defined as
$$\alpha\sim\beta\iff \alpha\in\beta\cdot\mathrm S_m$$
Im trying to calculate it but it seems hard. Suppose that $n+1\ge m$.
Then if Im not wrong for some subset $A\subset\{0,\ldots,n\}$ such that $|A|=m$ then if $\alpha\in\{0,\ldots,m\}^m$ then
$$|A^m/\mathrm S_m|=\sum_{|\alpha|=m}\binom{m}{\alpha}=m^m$$
where $|\alpha|:=\sum \alpha_j$, in other words the above is the sum of all multinomial coefficients for $m$ positions and $m$ possibly different elements.
If $n+1>m$ then exist $\binom{n+1}{m}$ subsets of cardinality $m$, then the final answer should be
$$|\{0,\ldots,n\}^m/\mathrm S_m|=\binom{n+1}{m}\sum_{|\alpha|=m} \binom m \alpha =\binom{n+1}{m}m^m$$
But if $n+1<m$ then the multinomial sum will be not complete and we can calculate it in a more complicated way and where is not a closed form.
Im very unsure of my calculations. Can someone confirm or correct it? Thank you in advance.
Thank to both answers, both are correct, at least to this point:
$$|\{0,\ldots,n\}^m/\mathrm S_m|=\sum_{k=1}^{n+1}\binom{n+1}{k}\binom{m-1}{k-1}=\binom{m+n}{m}$$
- The Vandermonde's identity is key to show the equality.
Elements of $\{0,\ldots,n\}^m$ consist of $m$-tuples and I am assuming $S_m$ acts by place permutation. Since we are only interested in the orbits under the action of $S_m$, we can choose a nice representative: $$(\underbrace{0,\ldots,0}_{m_0},\underbrace{1,\ldots,1}_{m_1},\ldots,\underbrace{n,\ldots,n}_{m_n}).$$ Note that size of the orbit of this element is $$\displaystyle{m\choose m_0,\ldots,m_n}=\frac{m!}{m_0!m_1!\cdots m_n!}$$ and the number of such orbits is equal to the number of compositions of $m$ into at most $n+1$ parts, with parts labelled by an increasing sequence of integers between $0$ and $n$. Alternatively, one can count the number of orbits as the number of partitions of $m$ with at most $n+1$ parts, with parts labelled by distinct integers (in any order) between $0$ and $n$. The number of partitions of $m$ with exactly $k$ parts is the coefficient of $x^m$ in the power series expansion of $$\frac{x^k}{(1-x)(1-x^2)\cdots(1-x^{k})}.$$ It follows that the number of labelled partitions of $m$ with exactly $k$ parts is the coefficient of $x^m$ in power series expansion of $$\frac{{n+1\choose k}x^k}{(1-x)(1-x^2)\cdots(1-x^{k})}$$ and, therefore, the number of orbits is the coefficient of $x^m$ in the power series expansion of $$\sum_{k=0}^{n+1}\frac{{n+1\choose k}x^k}{(1-x)(1-x^2)\cdots(1-x^{k})}.$$