bgoodri

I think so. Stan changing the RNG is probably going to cause a lot of unit test failures in packages that test whether the average of a finite number of...

I'll look at it. The progress code is very dependent on both the operating system and whether the person is using RStudio. Positron didn't exist at the time and is...

@avehtari knows the model. In any event, all the post-estimation functions in rstanarm work with `stan_clogit` output, so I think it is mostly a matter of synchronizing with what projpred...

A chatbot seems to think that `project` could be made to support clogit models by calling `posterior_predict(stan_clogit_output)` to get 4000 or so outcome vectors and then calling `survival::clogit()` on each...

Augmented data projection is an interesting possibility for clogit models because the denominator of the [likelihood](https://en.wikipedia.org/wiki/Conditional_logistic_regression#Conditional_likelihood) function already requires evaluating the probability of observing _all_ possible configurations of the outcome...

Let me see if I can summarize what the chatbot is telling me. In a clogit model, the strata are conditionally independent, but the outcomes within a stratum are not...