Dimitriy V. Masterov
Im glad you find my program useful and educational. In any case, there are two reasons why my -regxfe- is faster and matches to the baseline.
a) While you use double precision for the outer loop, I use float precision as the default. One can, of course, adjust that.
b) the correction of the degrees of freedom, since due to absorbed Fixed effects.
Nick Cox
Thanks for an earlier post. I really thought that commands coming from Stata Journal were automatically added to SSC database. I know now that isn't the case, and I have submitted the program to the right place for "easier" access.
Fernando
Im glad you find my program useful and educational. In any case, there are two reasons why my -regxfe- is faster and matches to the baseline.
a) While you use double precision for the outer loop, I use float precision as the default. One can, of course, adjust that.
b) the correction of the degrees of freedom, since due to absorbed Fixed effects.
Nick Cox
Thanks for an earlier post. I really thought that commands coming from Stata Journal were automatically added to SSC database. I know now that isn't the case, and I have submitted the program to the right place for "easier" access.
Fernando
Comment