Is the result of this inner product correct?

58 Views Asked by At

Let $\mathscr{H}$ be a Hilbert space with orthonormal basis $\{e_{1},e_{2},...\}$ and let $\sigma$ be a permutation of the set $\{1,...,n\}$ for some $n \ge 1$. Consider the following object: $$v = e_{i_{1}}\otimes \cdots \otimes e_{i_{n}}$$ Notice that the indices of $v$ may repeat. For each $i = 1,2,...$, let $n_{i}$ the number of times that the index $i$ appears in $v$.

On page 24 of these lecture notes, it is stated that:

$$\|\sum_{\sigma}e_{i_{\sigma(1)}}\otimes \cdots \otimes e_{i_{\sigma(n)}}\|^{2} = n! \prod_{i=1}^{\infty}n_{i}!$$

However, according to my calculations, this should be: $$\|\sum_{\sigma}e_{i_{\sigma(1)}}\otimes \cdots \otimes e_{i_{\sigma(n)}}\|^{2} = \frac{n!}{ \prod_{i=1}^{\infty}n_{i}!}$$ instead, because expanding $\|\cdot\|^{2}$ one would get: $$\sum_{\sigma,\rho}\delta_{i_{\sigma(1)},i_{\rho(1)}}\cdots \delta_{i_{\sigma(n)},i_{\rho(n)}}$$ where $\delta_{i,j}$ denotes the Kronecker delta. This would be equal to the number of permutations $\sigma,\rho$ which agree with $n_{1}$ repetitions of the $i=1$ index, $n_{2}$ repetitions of the $i=2$ index and so on. Am I doing something wrong or the result of the text is wrong?

1

There are 1 best solutions below

0
On BEST ANSWER

The formule in the book is correct! Indeed, to count the number of permutations, you could start by choosing your $\sigma$ (you have $n!$ choices). After that you need to choose $\rho$ such that $i_{\sigma(j)} = i_{\rho(j)}$ for all $j$. So $\rho(1)$ you have $n_{i_{\sigma(1)}}$ choices possible, for the next one you have $n_{i_{\sigma(1)}} -1$ choices and so one and so forth. You will have $\prod_{i=1}^\infty n_i!$ permutations $\rho$.