Resolve logistic regression parameters on simulated data

I see the subject of NUTS has been brought up before, just not getting the priority right now.