Panel Integrated choice latent variable model

8 views
Skip to first unread message

kiran.t...@gmail.com

unread,
Aug 5, 2021, 3:56:32 AMAug 5
to Biogeme
Dear Prof. Bierlaire,

I have  a panel data and I am developing the integrated choice latent variable model on it. During the iterations, the following erreor comes up:

 RuntimeWarning: overflow encountered in double_scalars rho = num / denom

What does this error mean? 
Also when I look at the results page, it says that the cause of termination of iterations is : the  trust region is too small. It means it didnt converge, right?
Thanks for your help

Kind Regards,
Kiran  

Bierlaire Michel

unread,
Aug 7, 2021, 11:40:00 AMAug 7
to kiran.t...@gmail.com, Bierlaire Michel, Biogeme
On 5 Aug 2021, at 02:21, kiran.t...@gmail.com <kiran.t...@gmail.com> wrote:

Dear Prof. Bierlaire,

I have  a panel data and I am developing the integrated choice latent variable model on it. During the iterations, the following erreor comes up:

 RuntimeWarning: overflow encountered in double_scalars rho = num / denom

What does this error mean? 

It is an overflow warning. These models are quite complicated, and there may be numerical issues during estimation.

Also when I look at the results page, it says that the cause of termination of iterations is : the  trust region is too small. It means it didnt converge, right?

No, indeed. You can decide if the results are acceptable based on the value of the norm of the projected gradient that is reported in the html file. 

Thanks for your help

Kind Regards,
Kiran  

--
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 on the web visit https://groups.google.com/d/msgid/biogeme/9292f872-e478-4d1e-bf69-e886687d0e92n%40googlegroups.com.

Reply all
Reply to author
Forward
0 new messages