[Limdep Nlogit List] working nlogit with >11 variables
Prof David Hensher
hgroup at optusnet.com.au
Fri Nov 14 14:10:49 EST 2008
Bill
I had separately advised her exactly this so she got it twice which
is fine. Also I think she has some vars that do not vary across all 3 alts.
Working on Ordered choice doc. Up to page 30.
David
At 02:07 PM 14/11/2008, you wrote:
>Dear Marit.
>(1) You need to press the Enter key in those utility functions to
>break up the lines. The internal limit on number of attributes
>is at least 100.
>(2) You are using Vista. There is an incompatibility between Vista
>and NLOGIT's help engine. We are working on the solution. Meantime,
>the support page at nlogit.com has some more information.
>(3) NLOGIT contains some built in routines for computing WTP. These
>are described in the manual.
>/Bill Greene
>
>----- Original Message -----
>From: "Marit Kragt" <marit.kragt at anu.edu.au>
>To: limdep at limdep.itls.usyd.edu.au
>Sent: Thursday, November 13, 2008 9:47:44 PM GMT -05:00 Columbia
>Subject: [Limdep Nlogit List] working nlogit with >11 variables
>
>
>Dear Limdep users,
>
>I have recently started using NLOGIT4.0 to analyse discrete choice
>data. There are a couple of issues with the program that I nor my
>colleagues here can solve. I am hoping to get some help from you.
>
>1) Every time I try to run a simple conditional logit model with
>more than 11 explanatory variables the program gets stuck. There is
>no problem whatsoever with the dataset (no missing data, no
>collinearity). I don't understand why the model doesn't accept more
>variables. I am using the command NLOGIT
>;lhs=choice, cset, altij
>;choices=SQ,change1,change2
>;Model:
>U(SQ) = ASCSQ + cost*cost + env1*attribute1 + env2*attribute2 +
>env3*attribute3 + env4*attribute4 + visit*ascvis + gender*ascgen +
>income*ascincom + education*asceduc + hhsize*aschhsize + farm*ascfarm/
>U(change1) = cost*cost + env1*attribute1 + env2*attribute2 +
>env3*attribute3 + env4*attribute4 + visit*ascvis + gender*ascgen +
>income*ascincom + education*asceduc + hhsize*aschhsize + farm*ascfarm/
>U(change2) = cost*cost + env1*attribute1 + env2*attribute2 +
>env3*attribute3 + env4*attribute4 + visit*ascvis + gender*ascgen +
>income*ascincom + education*asceduc + hhsize*aschhsize + farm*ascfarm$
>
>2) Every time I open the HELP function it give me an error "There
>was a problem running the macro (1037)". Clicking "Help Topics" in
>the appearing screen does not do anything. Why does the help
>function not work correctly in NLOGIT4.0?
>
>3) And lastly, I was wondering if anyone has an example of how to
>calculate WTP using bootstrapping? I have done parametric
>bootstrapping for logit models in other software programs, but am
>not positive on how to use this approach in NLOGIT. Since the help
>function doesn't work, it is hard to be sure.
>I suspect I need to use the PROCEDURE command to estimate a model
>and calculate the WTP based on this model, and then EXECUTE ;N =
>#repl ;Bootstrap = ...$
>
>
>Thank you very much for your time, your feedback is much appreciated!
>
>Regards,
>Marit
>
>-----
>Marit Kragt
>PhD scholar
>
>Crawford School of Economics and Government & Integrated Catchment
>Assessment and Management Centre (iCAM)
>Building 13, Ellery Crescent
>The Australian National University
>Canberra ACT 0200 Australia
>
>T: +61 2 6125 6557
>W: http://icam.anu.edu.au,
>http://www.crawford.anu.edu.au/students/showphd.php?surname=Kragt
>
>_______________________________________________
>Limdep site list
>Limdep at limdep.itls.usyd.edu.au
>http://limdep.itls.usyd.edu.au
>_______________________________________________
>Limdep site list
>Limdep at limdep.itls.usyd.edu.au
>http://limdep.itls.usyd.edu.au
More information about the Limdep
mailing list