My gists
I don’t always have time to write up a full blog on a modeling topic (or haven’t gotten to it yet). When that happens, I’ll often put up a pair of files on Gist, one in Stan, one in R. The Stan file contains the model implmentation, while the R file will simulate some fake data and run it.
Heckman Correction in Stan
B-spline in Stan
Gaussian process heterogeneous treament effects instrumental variables in Stan
Biprobit
A very simple, very fast Naive Bayes-type classifier for very high dimensional data