st: nl with fixed effects

["Heiwai Tang" <heiwai.tang@tufts.edu>]

Hi,

 

I am running a bunch of non-linear least square regressions involving over
100 fixed effects (countries and industries). Interactive command of "nl"
gives me an error message saying ". too long." This is a well-known problem.

 

I am wondering if you are so kind to share with me some sample code that
does this task. In particular, how can I call the nl evaluator without
explicitly typing up hundreds of b1, b2, b3 . in the calling command.

 

Can I do something like the following

 

nl test @ y x _I*, nparameters(200) initial({b} 1)

 

where {b} captures all unspecified coefficients?

 

Thanks in advance.

 

Heiwai Tang

Assistant Professor of Economics

Tufts University
*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/