my numbers don't match up with Stata while trying out different models
def _ll_nbin(self, params, alpha, Q=0):
endog = self.endog
mu = np.exp(np.dot(self.exog, params))
I'm running several different models in the interpreter and still need to get a nice script to replicate and test case.
The test case that I'm working on is to verify cluster robust standard errors, PR will follow soon.
duplicate, I opened a new issues #1611
fixed in #1612