In case it helps, I think you cannot use lower bound = upper bound with some algorithms. The domain has to be non-empty (for example, I think it is the case for MLSL).
A. On Fri, Feb 26, 2016 at 4:31 PM, Kaushik Matia <[email protected]> wrote: > Eventually I did just that but I am curious to understand why it should > fail if I pass a parameter as fixed. > On Feb 26, 2016 4:27 PM, "federico vaggi" <[email protected]> > wrote: > >> I understand. But why don't you just pass 35 parameters to the library >> instead of passing 35 + a 'fixed' parameter? >> >> On Fri, 26 Feb 2016 at 21:55 Kaushik Matia <[email protected]> >> wrote: >> >>> Hi Federico, >>> >>> your understanding is correct. The problem is for 35 parameters. >>> The 36th parameter is just a constant given the other 35. >>> >>> I pass 36 parameter to the library and I am asking it to hold the 36th >>> parameter fixed (by passing "upper bound" = "lower bound" = "guess 36th >>> parameter guess value"). >>> >>> This should make the problem a 35 parameter problem to the library and >>> it should not fail. >>> So that is my confusion. Why is the library failing when I try to call >>> the library this way. >>> >>> When I pass "n+m" parameters and fix "m" of the parameter should the >>> library not try to optimize my function on the "n" parameter that it is >>> allowed to move around? >>> >>> Best >>> >>> . >>> >>> >>> >>> >>> >>> On Fri, Feb 26, 2016 at 2:00 PM, federico vaggi < >>> [email protected]> wrote: >>> >>>> Unless I wildly misunderstood, doesn't your problem reduce to 35 >>>> parameters? The 36th parameter is just a constant, given the other 35. >>>> >>>> On Fri, 26 Feb 2016 at 18:00 <[email protected]> >>>> wrote: >>>> >>>>> Send NLopt-discuss mailing list submissions to >>>>> [email protected] >>>>> >>>>> To subscribe or unsubscribe via the World Wide Web, visit >>>>> >>>>> http://ab-initio.mit.edu/cgi-bin/mailman/listinfo/nlopt-discuss >>>>> <https://urldefense.proofpoint.com/v2/url?u=http-3A__ab-2Dinitio.mit.edu_cgi-2Dbin_mailman_listinfo_nlopt-2Ddiscuss&d=AwMFaQ&c=-dg2m7zWuuDZ0MUcV7Sdqw&r=8jNZOJYj_uffjo9ZgZlRE2gf1cRFzde3vf5YpxT6qnU&m=HpRVwF989fBWeLpx5Ip7BGCtVd3bfl_Cj93eGDWZ60U&s=qD1fcn5M5vz5bPuMBO7Ri__n83TMUs5xm-hRsHrFVao&e=> >>>>> or, via email, send a message with subject or body 'help' to >>>>> [email protected] >>>>> >>>>> You can reach the person managing the list at >>>>> [email protected] >>>>> >>>>> When replying, please edit your Subject line so it is more specific >>>>> than "Re: Contents of NLopt-discuss digest..." >>>>> >>>>> >>>>> Today's Topics: >>>>> >>>>> 1. Fixing a parameter causing failure. (Kaushik Matia) >>>>> >>>>> >>>>> ---------------------------------------------------------------------- >>>>> >>>>> Message: 1 >>>>> Date: Thu, 25 Feb 2016 15:28:17 -0500 >>>>> From: Kaushik Matia <[email protected]> >>>>> To: [email protected] >>>>> Subject: [NLopt-discuss] Fixing a parameter causing failure. >>>>> Message-ID: >>>>> < >>>>> caoui8dxs+orltchiytxmtzyjmpnh_c2gp7c0ctpjgqbhqn0...@mail.gmail.com> >>>>> Content-Type: text/plain; charset="utf-8" >>>>> >>>>> Hi, >>>>> >>>>> I have optimization problem with 36 parameters. I have a equality >>>>> constraint where the last parameter is equal to the sum of the others >>>>> i.e. >>>>> x_36 =(x_1+...+x_35). >>>>> >>>>> I try to optimize as follows: >>>>> >>>>> 1) I set up problem with 36 parameters. >>>>> 2) put lowerbound = upperbound =guess param for the 36th parameter >>>>> with the >>>>> intention to fix it from changing. >>>>> 3) in my objective function valuation i just compute x_36 as given >>>>> above >>>>> and use it for function evaluation. >>>>> >>>>> When I try to use algorithm COBYLA or Isres I get a failure. >>>>> >>>>> The error happens inside static double elimdim_func() at the line >>>>> >>>>> val = d->f(); >>>>> >>>>> Essential the function pointer is null and the code breaks. >>>>> >>>>> May I please request someone to point out what I may be doing wrong >>>>> >>>>> Best >>>>> -------------- next part -------------- >>>>> An HTML attachment was scrubbed... >>>>> URL: < >>>>> http://ab-initio.mit.edu/pipermail/nlopt-discuss/attachments/20160225/8f4b7130/attachment-0001.html >>>>> <https://urldefense.proofpoint.com/v2/url?u=http-3A__ab-2Dinitio.mit.edu_pipermail_nlopt-2Ddiscuss_attachments_20160225_8f4b7130_attachment-2D0001.html&d=AwMFaQ&c=-dg2m7zWuuDZ0MUcV7Sdqw&r=8jNZOJYj_uffjo9ZgZlRE2gf1cRFzde3vf5YpxT6qnU&m=HpRVwF989fBWeLpx5Ip7BGCtVd3bfl_Cj93eGDWZ60U&s=vNBowW8e0ec3X0kle9GVeZ8w9G07VfXaPZLIuI6OCjk&e=> >>>>> > >>>>> >>>>> ------------------------------ >>>>> >>>>> Subject: Digest Footer >>>>> >>>>> _______________________________________________ >>>>> NLopt-discuss mailing list >>>>> [email protected] >>>>> http://ab-initio.mit.edu/cgi-bin/mailman/listinfo/nlopt-discuss >>>>> <https://urldefense.proofpoint.com/v2/url?u=http-3A__ab-2Dinitio.mit.edu_cgi-2Dbin_mailman_listinfo_nlopt-2Ddiscuss&d=AwMFaQ&c=-dg2m7zWuuDZ0MUcV7Sdqw&r=8jNZOJYj_uffjo9ZgZlRE2gf1cRFzde3vf5YpxT6qnU&m=HpRVwF989fBWeLpx5Ip7BGCtVd3bfl_Cj93eGDWZ60U&s=qD1fcn5M5vz5bPuMBO7Ri__n83TMUs5xm-hRsHrFVao&e=> >>>>> >>>>> >>>>> ------------------------------ >>>>> >>>>> End of NLopt-discuss Digest, Vol 75, Issue 4 >>>>> ******************************************** >>>>> >>>> >>>> _______________________________________________ >>>> NLopt-discuss mailing list >>>> [email protected] >>>> http://ab-initio.mit.edu/cgi-bin/mailman/listinfo/nlopt-discuss >>>> <https://urldefense.proofpoint.com/v2/url?u=http-3A__ab-2Dinitio.mit.edu_cgi-2Dbin_mailman_listinfo_nlopt-2Ddiscuss&d=AwMFaQ&c=-dg2m7zWuuDZ0MUcV7Sdqw&r=8jNZOJYj_uffjo9ZgZlRE2gf1cRFzde3vf5YpxT6qnU&m=HpRVwF989fBWeLpx5Ip7BGCtVd3bfl_Cj93eGDWZ60U&s=qD1fcn5M5vz5bPuMBO7Ri__n83TMUs5xm-hRsHrFVao&e=> >>>> >>>> >>> > _______________________________________________ > NLopt-discuss mailing list > [email protected] > > https://urldefense.proofpoint.com/v2/url?u=http-3A__ab-2Dinitio.mit.edu_cgi-2Dbin_mailman_listinfo_nlopt-2Ddiscuss&d=AwIGaQ&c=-dg2m7zWuuDZ0MUcV7Sdqw&r=8jNZOJYj_uffjo9ZgZlRE2gf1cRFzde3vf5YpxT6qnU&m=HpRVwF989fBWeLpx5Ip7BGCtVd3bfl_Cj93eGDWZ60U&s=qD1fcn5M5vz5bPuMBO7Ri__n83TMUs5xm-hRsHrFVao&e= > >
_______________________________________________ NLopt-discuss mailing list [email protected] http://ab-initio.mit.edu/cgi-bin/mailman/listinfo/nlopt-discuss
