Hi folks,
I have a problem with termination conditions using LBFGS. I want an
optimization to run just a long time to check convergence behaviour. The
optimization itself works just fine but there is a problem with termination.
I disabled all criteria by setting them to the values specified in the
reference. After that I set maxtime to some value, e.g. 864000 seconds
and start the simulation.
Optimization is running and gets closer and closer to the expected
result, but after some time which's much less than the specified maxtime
the optimization stops returning '1' (general success). I can
reproduce this behaviour using different initial values, different
values for maxtime, etc.
Is there any hard-coded termination in the implementation of LBFGS in
NLOpt? Does anybody has experienced the same problem? Any suggestions
how to solve this?
Thanks a lot in advance.
Best regards,
Timo
_______________________________________________
NLopt-discuss mailing list
[email protected]
http://ab-initio.mit.edu/cgi-bin/mailman/listinfo/nlopt-discuss
