[Gretl-users] Data Error for Probit/Logit

[Allin Cottrell]

On Wed, 6 Aug 2008, Mike Fontaine wrote:

> Recently I have been receiving the same error message, "Data 
> Error", when I try to run a regression via Gretl. OLS, 
> Heteroskadasticity corrected and all other regressions seem to 
> work, however Logit and Probit do not work with the data set I 
> have. There is 571 points with 4 independent variables, however 
> I have successfully run a Logit regression before using the same 
> data set but with more points. Perhaps there is a minimum amount 
> of points, or that somehow my data got out of wack?

Please send me a copy of your data file, and the specs for the 
logit/probit regression.  I can't really guess what's happening 
without seeing the data.

Thanks,

Allin Cottrell