Issue with Mixed Logit Model Combining Panel effect

[Amneh Nassar]

Dear Prof. Berlier,

I need your assistance with an issue I am encountering while estimating a Mixed Logit Model that combines latent variables and panel data.

I receive the error shown below, 

The norm of the gradient at ASC_S=1, BUS_AGE_1=0, BUS_ASC=0, BUS_COST=0, BUS_CROWDED_1=0, BUS_CROWDED_3=0, BUS_FREQUENCY=0, BUS_LANE=0, BUS_OLD_AGE=0, BUS_TRANSFERS=0, BUS_TRAVEL_TIME=0, BUS_WALK_TIME=0, CARPOOL_ASC=0, CARPOOL_COST=0, CARPOOL_PARK_TIME=0 is nan: g=nan, nan, nan, nan, nan, nan, nan, nan, nan, nan, nan, nan, nan, nan, nan

I tried to estimate the model based on initial Beta values from the MNL model results, but I received the same error.
Attached my script

 Thanks a lot for your help!

Amneh
 

[Michel Bierlaire]

Try to increase the initial value of the standard deviation parameters. A value of 1 is sometimes insufficient. Try with 10...
Also, estimate first the version without panel effect.

> --
> You received this message because you are subscribed to the Google Groups "Biogeme" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to biogeme+u...@googlegroups.com.
> To view this discussion visit https://groups.google.com/d/msgid/biogeme/805af8cc-8b19-4ff1-91e9-9de50ae4039an%40googlegroups.com.
> <SMNL ML.py>

Michel Bierlaire
Transport and Mobility Laboratory
School of Architecture, Civil and Environmental Engineering
EPFL - Ecole Polytechnique Fédérale de Lausanne
http://transp-or.epfl.ch
http://people.epfl.ch/michel.bierlaire