Mathias Lerbech Jeppesen

Thanks for the quick response. I was messing around with doing some likelihood calculation "manually" while doing some MCMC sampling in connection with some bayesian GLM. In that task I...

Hmm I guess an `as_distributional` function with a family method could still be useful - just in conjunction with some "custom" code. By reparametrisation I mean that if I want...

Ah okay, I didn't notice the existence of the `log_likelihood` function, I'll use that together with something like the above. Thanks for engaging with my problem! I'll leave it up...