How to shrink the number of iterations or stop iterating

[linlin zhang]

Dear Michel,

I am using Pandas Biogeme to estimate the NL and CNL modle at the same time.

How to shrink the number of iterations or stop iterating? I can see the Loglikelihood did not reduced too much after 15-20 iterations.

Many thanks!

Best, Lin

[Bierlaire Michel]

It is usually a sign that there is a specification issue with the model. 

You can control the stopping criteria  by providing a dict of to the optimization algorithm.

The parameters are documented here: http://biogeme.epfl.ch/sphinx/optimization.html#biogeme.optimization.simpleBoundsNewtonAlgorithmForBiogeme

This dict is transmitted to the algorithm through the algoParameters arguments of the function “estimate”. 

http://biogeme.epfl.ch/sphinx/biogeme.html#biogeme.biogeme.BIOGEME.estimate

Many thanks!

Best, Lin

--
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/d52c4515-140a-400e-b6ab-bab8cad9f3c7n%40googlegroups.com.