Hi everyone,
How would you recommend testing UDFs for PyStan model given that the functions can’t be exposed directly into Python?
Compile separate test programs to evaluate them? This seems like overkill (and still requires some kind of statistical model)
Thanks!