[Limdep Nlogit List] Help with panel data and Error Comp. Logit

William Greene wgreene at stern.nyu.edu
Mon May 3 20:10:54 EST 2010


Gulsah: For the problem as you described it, you uave ;PDS=5.
The diagnostic implies that you do not have the right number of
observations consistent with your specification of ;PDS and
the number of choices, for the entire sample. The program loops
through your current sample using PDS and the number of choices
you have specified. If it gets to the end of the sample and the
last choice in the last period is not the last row of your 
sample, you have made a mistake in there somewhere, and the
diagnostic is given.
/B. Greene

----- Original Message -----
From: "Gulsah Akar" <akar.3 at osu.edu>
To: "Limdep and Nlogit Mailing List" <limdep at limdep.itls.usyd.edu.au>
Sent: Sunday, May 2, 2010 10:39:23 PM GMT -05:00 Colombia
Subject: [Limdep Nlogit List] Help with panel data and Error Comp. Logit


Hi,

I am having a problem related to the panel specification for discrete choice models (Pds=...). My data is in the discrete choice format. I have 30 choices for each individual and repetitive observations. The number of observations vary among the individuals.

So, let's say I have 5 observations from the first individual. Then, I have 5 x 30 =150 rows associated with this individual. So should the _groupti variable be 5 or 150 for this individual? Actually I tried both and I am getting an error in both cases, "the number of observations do not match the individuals" I can not figure out what exactly I am missing.

Thanks!

Gulsah


Gulsah Akar, PhD
Visiting Assistant Professor
City and Regional Planning
Knowlton School of Architecture
The Ohio State University
275 West Woodruff Av.
Columbus, OH 43210
614-292-6426
akar.3 at osu.edu  


_______________________________________________
Limdep site list
Limdep at limdep.itls.usyd.edu.au
http://limdep.itls.usyd.edu.au


More information about the Limdep mailing list