Hi everyone! I tried nlopt, some methods are fast enough, it's free and good alternative, but is there a possibility to limit distance for steps for derivatives methods? Something like [cid:[email protected]] In particular I'm interested on NLOPT_LD_SLSQP it uses too big steps, it jumps from one side of bounds to another and since I have several local minimums it converges to unpredictable minimum, despite on good start points. No question for derivatives free methods cause as I understood it is possible tune them by initial_step.
[cid:[email protected]] Best regards, Anatoliy.
_______________________________________________ NLopt-discuss mailing list [email protected] http://ab-initio.mit.edu/cgi-bin/mailman/listinfo/nlopt-discuss
