inconvenient: logit when endog is (1,2) instead of (0,1) #1053

Closed
josef-pkt opened this Issue Aug 19, 2013 · 1 comment

Projects

None yet

1 participant

@josef-pkt
Member

This breaks with a Singular Matrix problem in the Hessian calculation.
It took me two hours of figuring out the singular matrix issue, before I thought of checking endog

I think we should add a basic check for this.

I haven't looked at the details yet.

related: I was looking at using Logit for continuous variables in [0,1] interval, and it seemed to work.

@josef-pkt
Member

http://stackoverflow.com/questions/20703733/logit-regression-and-singular-matrix-error-in-python

We should raise an error if there are no zero values (endog==0).sum() == 0.

@jseabold jseabold closed this in c4e71c0 Sep 20, 2014
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment