I am using lasso for model selection and predictions. The code looks as follows:


Code:
lassologit  y x1-x15 if x16 == 100

lassologit, lic(ebic) postresults
The first line runs just fine giving me the list of selected variables, pseudo R^2 etc.

For the second line I get the following error:

Code:
Use lambda=.7707203529877009 (selected by EBIC).
    st_matrixcolstripe():  3300  argument out of range
          ereturn_beta():     -  function returned error
           PostResults():     -  function returned error
                 <istmt>:     -  function returned error
r(3300);
I don't understand this error message. Is there anything I can do? I run this code several times with different datasets (defined by x16) and the error occurs multiple times.

I would assume that if the first line works fine, the second should as well, as the final model has already been estimated within the first line of code.

I would like to generate predicitons from the final model with
Code:
predict double phat, pr
Happy for any help, thanks!