[Limdep Nlogit List] spot the error?

Mikołaj Czajkowski miq at wne.uw.edu.pl
Wed Aug 20 02:17:40 EST 2008


Philip A. Viton wrote:
> 
> Can anyone detect the problem with the .lim command file shown below? 
> (nlogit 3). It's driving me nuts. What happens is this: I load in my 
> data, and then the command file. I highlight the commands and run them. 
> Dstats on the result shows that t1962 and t1994 have mean zero, ie that 
> these two lines have not "taken". (The data for the time variable does 
> in fact contain entries for these two years). This is repeatable, both 
> by re-highlighting and re-running; and by repeating the entire sequence 
> after re-starting nlogit.
> 
> But if I now open a new command window and type "create;" then 
> cut-and-paste the 1964 line ; and finally replace the ending ";" with a 
> "$" and run THAT line, t1962 now seems to be OK. I'm a little concerned 
> that I can't see what is going wrong, because it might also affect other 
> transformations where the failure isn't as obvious.
> 
> (I can supply a test dataset if anyone wants to experiment).

Philip - I can't spot the problem.
But if you're using Limdep 9.0 you might want to try the:
create; expand (name)$
which expands categorically coded 'name' into dummy variables.

Regards,

-- 
     Mikołaj Czajkowski

     Warsaw Ecological Economics Center
     Warsaw University
     http://www.woee.pl/



More information about the Limdep mailing list