OrdinalLogistic - wrong support?

[svilupp]

I was getting an odd indexing error (using negative indices) when trying to get the pdf of the OrdinalLogistic, which led me to check the support of the distribution.

Given an Ordinal Logistic (PO model) with K categories and K-1 cutpoints, shouldn't we have support from 1:K?

In the following line, we seem to define minimum support to be 0?

I'm happy to open the PR, but I wanted to make sure I'm not missing something.