Initial exploration of stan, including user defined functions
Playing around using sklearn wisconsin breast cancer dataset. Basic logistic regression using logit link function.
Incremental attempts to incoporate more interesting operations and arguments. To date, the best I have is a vector,vector --> double function: a user defined dot product.
I'm currently using Dan Foreman-Mackey's hacked pystan which allows includes as an argument to pystan.StanModel to permit custom user headers. Of course, this would be better pushed into a forked version of stan::math when I have a more useful function.
Now to code up a Bspline basis, function and derivative...