Noa Kallioinen

Related to this, we've been discussing with @avehtari that it would be useful if messages/warnings related to e.g. convergence diagnostics would be available when the diagnostic is printed, not just...

rOpenSci recently published a guide/recommendations about this: https://ropensci.org/blog/2024/02/06/verbosity-control-packages/ See also [here](https://stats-devguide.ropensci.org/standards.html#output-verbosity), specifically discussing Bayesian software. Perhaps this would be a reasonable guide to follow for `posterior`?

I made a first [attempt](https://github.com/n-kall/posterior/commit/d865e57aecb8c03953f98311b80698af6b6013ae) at implementing the note-style messages in summarise draws. It seems that it could be done with attributes (see example below). I'm not sure if this...

Regarding possible levels or categories: Perhaps there could be different categories for different types of messages. For example: - diagnostic messages (e.g. high pareto_khat, high rhat, low ess) - computation...

I changed this to a draft as I am now thinking about whether a single `relative_eff` function is the way to go. Perhaps there should be one for different quantities...

As a single `r_eff` function doesn't really make sense (it should be dependent on the quantity of interest), I will close this PR.

I've made some updates regarding the extraction of the log_likelihood and log_prior. Below is a working example using CmdStanPy. I think it might make sense to include log_prior as a...

It can be useful to have the lprior as an array with different contributions to the joint log prior, so that it is possible to check the sensitivity to changing...

@OriolAbril Thanks for all the help with this PR! Greatly appreciated!

Thanks for the replies. > I would prefer to not add another hard dependency of loo Completely agree, I think it would fit under `suggests` fine. One potential hurdle is...