Plot or summarize priors about bambi HOT 5 CLOSED

I like the idea of plot_priors(), which should make it pretty much impossible for users not to understand what the priors are. That said, most of the plots will just be a bunch of identical-looking panels of Normal and HalfCauchy distributions, where the difference is just in the scale of the axes -- this redundancy is not necessarily problem, but it just means the plots may induce an eye-glazing reaction at first.

from bambi.

Agreed, but I think clarity is the most important thing here; I'm happy to bore users to tears as long as they understand what's going on. 🙂

from bambi.

In general I do think that the avoidance of redundancy is an overrated criterion for plots and tables.

from bambi.

I have some somewhat janky code that plots priors in the ESCS multiple regression notebook, but it doesn't yet work in the general case.

from bambi.

0b02417 implements this for all fixed effect parameters and the residual SD (in normal models), but it doesn't yet plot the priors for the random effect variances.

from bambi.