New type of distribution based on empirical data

[RomainBes]

I was wondering if it would be possible, and relatively easy, to add a new type of distribution for parameters based on empirical distribution that would rely on the random library?

import random
random.choices(np.arange(0, 4), weights=np.arange(0, 4), k=10)

[raphaeljolivet]

It could instead.
But I guess semantically it would be better to introduce a new type of "integer" parameters then.
That's something I had in mind.