Jonah Gabry

@avehtari I just added the `model` column. Can you try it when you have time? I will make a new list of the reverse dependencies that need to be fixed,...

Here's an example of how rstanarm output looks using this branch. There are two methods. The first just uses regular `loo_compare()` and so it's no different than just using the...

> You can see how the comparison tables look like with tinytable at https://users.aalto.fi/~ave/casestudies/LOO_uncertainty/loo_uncertainty_diag.html It looks good, although none of the examples in this have anything in the `diag_elpd` column,...

> > You can see how the comparison tables look like with tinytable at https://users.aalto.fi/~ave/casestudies/LOO_uncertainty/loo_uncertainty_diag.html > > It looks good, although none of the examples in this have anything in...

> Maybe change `khat > 0.5` to `k_diff > 0.5` and `khat_psis > 0.7` to `k_psis > 0.7` to make them more different and also shorter. Also the current print.loo...

I made some small doc edits, fixed the tests, and moved the computation of `diag_diff` and `diag_elpd` to separate internal functions (defined at the end of the same R file)....

> It's p, because the equations use p(). I've not seen before anyone proposing we should give up using p() in equations. Instead of prob, more logical would be Pr...