You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Unfortunately statsmodels doesn't.
statsmodels is pretty weak in traditional econometrics with endogenous
regressors.
The linearmodels package has more support for linear IV, statsmodels has
only GMM that can be used for endogenous regressors.
We don't have a good bivariate normal distribution. scipy has now also multivariate normal, but I'm not sure how fast and accurate the cdf is in the bivariate case.
We have an old version of bivariate normal distribution class in sandbox.distributions, but that's not ready for use.
The cdf uses double quad, instead we could use a single quad and the univariate cdf for the conditional distribution, which I guess would be more accurate.
Do you support the "heckprobit"(probit with sample selection) model?
Stata supports this model. Does statsmodels supporst this model?
Looking forward to your reply, thank you very much!
The text was updated successfully, but these errors were encountered: