Paul-Christian Bürkner

Ok, I would leave it open. It can be used as a starting point for dicussion on more principled default priors. Thank you for taking the time to think about...

Thank you! I think I do see that point now better thanks to your examples. I will think about it a little more but will probably implement your, or a...

Thanks for opening this issue. Refers to discussion https://discourse.mc-stan.org/t/predict-with-a-brms-phylogenetic-model-for-a-new-species-with-known-phylogenetic-position/7480/

This is not yet implemented. Sorry.

Implementing this is non-trivial and I don't have much time to work on complicated brms features these days. Eventually, it will have to be implemented in `get_new_rdraws` which you can...

I think we need more tests though, possibly with much more chains than four to get more reliable estimates. There is significant between chains difference in speed and so the...

I have recently introduced the requirement for `data` in **brms**, since, as you say, not doing so comes with a whole lot of problems. In particular, users may not realize...

brms has the dpar argument so one can do posterior_epred(..., dpar = "mu") to get p. in brms, all families return their mean in posterior_epred by default regardless of their...

posterior_linpred(transform = TRUE) currently equals posterior_epred in brms which I agreee with is not ideal. I was planning to remove this argument at some point but now that you say...

They are so called distributional models or models of "location, scale, and shape". You may fit them with brms which also uses Stan at the backend (see https://cran.r-project.org/web/packages/brms/vignettes/brms_distreg.html for examples).