The Student T distribution is described as being the “Non-central Student’s T log-likelihood.” But either I’m thoroughly confused, or it’s not really the non-central T as described on wikipedia and in spicy.stats.nct. The pymc3 implementation has a location parameter, but not a non-centrality parameter, and random samples are always symmetric.
Am I right or confused?
That indeed seems like a typo. I think what was supposed to be written was “Non-centered”. Want to do a PR with the fix (can also just remove “non-central”).