Frank Weber

Sorry, I currently don't have time for this. But I hope my comment above helps in fixing this.

> The Brier (Skill) Score and Precision Recall AUC could be useful evaluation statistics, and were suggested by Aki Vehtari in issue #25 but, it seems, not implemented. As stated...

As far as I understand the source code in lines https://github.com/stan-dev/projpred/blob/33047d6c9ea98d0644cb0c8503424f61ab0127af/R/misc.R#L29-L54 and lines https://github.com/stan-dev/projpred/blob/33047d6c9ea98d0644cb0c8503424f61ab0127af/R/summary_funs.R#L232-L248 as well as your explanations, I think it is the former of the two definitions, i.e....

> Implementing additional evaluation metrics is likely currently not one of your priorities in further developing projpred but for cases with highly imbalanced data it would be wonderful to, in...

> tss = tpr + tnr - 1 Interesting, I know this as Youden's Index (also seems to be called [Youden's J statistic](https://en.wikipedia.org/wiki/Youden%27s_J_statistic)). > I have struggled to wrap my...

> As Frank pointed out in https://github.com/stan-dev/projpred/issues/352, the bernoulli() distribution can be implemented by using brms:::get_refmodel.brmsfit() instead of init_refmodel(). Yes, thank you for updating the issue here, too. And for...

Ah, now that I read [this preprint](https://arxiv.org/abs/2109.04702v1), could it be that you were talking of what is mentioned in section 3.2 of that preprint, i.e., the fact that the PIRLS...

Related comments/discussions may be found in issue #173 (enumeration point 4) and PR #188. Now that I understand the purpose of `w` better, I think constant weights for LOO-CV are...

I have now read Magnusson et al. (2019) and Magnusson et al. (2020), but I'm still confused how **projpred** wants to perform the subsampled LOO CV. More importantly, I have...

An alternative to not supporting subsampled LOO CV at all anymore might be to throw a warning concerning a potentially incorrect implementation.