Let X be a descret variable over a variable-space of size k ( ).
And let f(X) be a guassian distributionof X's .
I am randomly choosing an X using disribution f(X), and repeat this processes N times, until I get a vector of size N. (N > k, and can be N >> k if it will make the problem easyer..).
My question is:
Is there a way to estimate how many different such vectors I can get from this process?