R star MCMC convergence metric

[ben18785]

This is a slightly self-involved one but I do actually think it'd be useful. I would like to add my R* convergence metric that I describe in this paper, as I've found it works pretty well across all problems I've thrown at it; it also offers additional information about convergence relative to the standard Rhat stats...anyway, enough sell!

One thing it requires is the ability to train a gradient-boosted regression tree ML model. It looks like Scikit-learn has this capability. How would people feel about having this as an optional dependency?

[MichaelClerx]

What's a gradient-boosted regression tree ML model?