On May 1, 2009, at 7:58 PM, Thomas Savitsky wrote:
> I looked into global optimizers, and scipy's anneal is not the > answer. It gives poor results quite often. A better choice would > be ASA, located at http://alumnus.caltech.edu/~ingber/#ASA > Unfortunately, there are functions which give this (and other) > optimizer problems, such as the Rosenbrock function: (1-x)^2 + 100 > (y - x^2)^2. Nevertheless, a numerical approximation for a global > minimum would benefit sage imo. But if it were added, how should > the end user be informed that the answer given may very wrong, and > significant effort spent tuning options to the optimizer may be > required for a good result? > I quite like the ASA optimizer and I used a MATLAB interface to it for my Master's thesis. However, I'm fairly certain the license isn't GPL compatible. It would be a nice optional package, though. Cheers, Tim. --- Tim Lahey PhD Candidate, Systems Design Engineering University of Waterloo http://www.linkedin.com/in/timlahey --~--~---------~--~----~------------~-------~--~----~ To post to this group, send email to sage-devel@googlegroups.com To unsubscribe from this group, send email to sage-devel-unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/sage-devel URLs: http://www.sagemath.org -~----------~----~----~----~------~----~------~--~---