[Limdep Nlogit List] NLogit Error 1099

[Thomas C. Eagle]

Your choice variables has to be coded as a 0,1 indicator and every choice
set MUST be of size three.  I bet your reject statement is building choice
sets of less than three and then there is overlap of the indicators such
that they sum to more than 1.0.

Tom Eagle

-----Original Message-----
From: limdep-bounces at limdep.itls.usyd.edu.au
[mailto:limdep-bounces at limdep.itls.usyd.edu.au] On Behalf Of Jose Sanchez
Sent: Friday, August 07, 2009 4:18 PM
To: limdep at limdep.itls.usyd.edu.au
Subject: [Limdep Nlogit List] NLogit Error 1099

Dear All,

I'm trying to run the following program (nested logistic):

sample;1-9729$ 
reject; choice=-999|cost=-999|chset1=9|chset2=9|chset3=9$
nlogit;LHS=choice;  choices=choicea,choiceb,choicec; 
                  RHS=cost,chance,damage,public,private,hi_pub,hi_pr$

and I get an error message: Error:  1099: Obs.  1987 responses should sum 
to 1.0. Sum is  2.0000

What exactly this error means? I looked in the manual and there is not 
much information.

Also, a colleague of mine ran the same program with the same data and did 
not get the error message. I'm running an older version 8.0.6 while my 
colleague is running version 9.  Is there a problem with the older 
version? If so, is there an update to fix the problem?

~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~
Jose Sanchez, Mathematical Statistician 
USDA Forest Service, Pacific Southwest Research Station
4955 Canyon Crest Dr.
Riverside, CA 92507-6099      Phone: (951)680-1589
jsanchez at fs.fed.us
~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~
_______________________________________________
Limdep site list
Limdep at limdep.itls.usyd.edu.au
http://limdep.itls.usyd.edu.au