question on dummy variables

#1
Hi,

Im trying to run a regression with 400 dummy variables. I created 400 dummy variables but when I run a regression of Stata on those 400 dummy variables (with the 'noconstant' option), Stata automatically drops 50 or so dummies (there is no collinearity problem). I know xtreg or areg would be a better way to do this but was wondering if anyone knows how to fix this problem and get the coefficients for all 400 dummies.

Thanks
 

CB

Super Moderator
#2
I don't know much about either Stata or physics, but 400 dummy variables is a pretty stupendous number - the risk of overfitting due to the complexity of the model would be very high. What's your rationale for inclusion of these variables?