[Limdep Nlogit List] Fixed effects model: Limdep-code for p-valueoutput per group

William Greene wgreene at stern.nyu.edu
Tue Jul 15 07:41:16 EST 2008


Mr. Shreay: You did not specify any covariates in your model.
XC is the variable that tells how many lines of data (how many
intervals) you observed. Each line should give the ending time of 
that spell. Suppose you have 3 intervals that ended at weeks 5, 8 
and 14 (which is the ending time).  Then the 3 lines of data would be
lhs = 5, rhs = the covariates in spell 1, XC=3
lhs = 8, rhs = the covariates in spell 2, XC=3
lhs = 14, rhs = the covariates in spell 3, XC=3
If you observed three spells for each person, you would use
;PDS=3. Otherwise it might vary.  
If the first 102 rows of your data all have duration equal to
66, then you are not interpreting the model correctly. Please 
refer to the description on pages E36-24 to E36-26 in your
manual.
Sincerely,
/B. Greene

----- Original Message -----
From: sshreay at mail.wsu.edu
To: "Limdep and Nlogit Mailing List" <limdep at limdep.itls.usyd.edu.au>
Cc: "Limdep and Nlogit Mailing List" <limdep at limdep.itls.usyd.edu.au>
Sent: Monday, July 14, 2008 4:08:51 PM (GMT-0500) America/Bogota
Subject: Re: [Limdep Nlogit List] Fixed effects model: Limdep-code for  p-valueoutput per group

Dear Dr Green and other LIMDEP users,

I am trying to run Survival analysis with time varying covariates. So I
have a duration column (I took log of this) and then I used the following
command.

Survival;lhs=logd;rhs=one;pds=xc;model=weibull $

So logd is log of my duration variable and pds=xc is time varying
covariate. For each observation I have 102 values for time varying
covariate so I arranged my data in a panel. So when I run my model using
the above command I get the following error message.

Error   559: Data error. Individual     0 T0=  66.0 T1=  66.0

To my understanding the first 102 rows in my data corresponds to duration
value of 66 since I created a panel for time varying covariate. What
should I do so that LIMDEP understands that first 102 values correspond to
first observation? Anyways I took log so it should have shown log of 66
instead of 66 in my error?


Sincerely,
Sanatan Shreay
PhD student,
Washington State University
_______________________________________________
Limdep site list
Limdep at limdep.itls.usyd.edu.au
http://limdep.itls.usyd.edu.au


-- 
Professor William Greene
Department of Economics
Stern School of Business
New York University
44 West 4th St., Rm. 7-78
New York, NY   10012
http://www.stern.nyu.edu/~wgreene
212.998.0876




More information about the Limdep mailing list