[Feature request] Bayesian Rsquare for GLM

[junpenglao]

Gelman, Goodrich, Gabry, & Ali 2017

The concept of R^2 and “explained variance” makes most sense for GLM with equal variance, but it seems easy to compute so why not ¯\_(ツ)_/¯

[aloctavodia]

I like the idea! I generally compute the R^2 when doing Bayesian linear regression. Also thanks for the paper.