Recent problems with optimiziation algorithm

[Philipp Kleffel]

Dear Professor Bierlaire,

since a couple of days, the optimization algoritm for my continous latent class model with panel data always stops at local minima. Even if I try to run exactly the same estimations as I did some days before, I don't get the results of these prior estimation runs. For normal models with random parameters, the itinerations are fine and deliver the same results as before.

Do you have any idea what might be the reason?

Thank you very much for your time and help!

Kind regards

Philipp Kleffel

University of Bamberg, Germany

[Bierlaire Michel]

By default, Biogeme saves each iteration in a file with a .iter extension. When it is relaunched, it uses the values from this file as a starting point for the estimation. This is designed to save time when a run must be interrupted. 

You can erase the .iter file. 

You can also turn off this feature by setting the parameter saveIterations to False

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

--
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/3c5492be-e73c-4f7b-85f4-5da26feaea65n%40googlegroups.com.

[Philipp Kleffel]

Dear Professor Bierlaire,

Thank you very much for your quick help. Indeed, it helped!

Kind regards

Philipp Kleffel