Recent content by aTn

I have no experience with Maxima, therefore I can't comment. However, does Maxima have the same functionnalities as Mathematica (in other terms, on what aspects do you want to compare the two) ? Hurkyl, I totally agree with you.

There are numerous algorithms available for black box optimization (or derivative free optimization as mathematicians and operations research experts call them). I strongly suggest looking at the following presentation. http://egon.cheme.cmu.edu/ewocp/docs/SahinidisEWO_DFO2010.pdf It...

For numerical optimization of medium to large scale problems, I strongly do not recommend Mathematica (use AMPL or GAMS instead; open-source solvers are also available through COIN-OR and other online sources; see NEOS server for optimization also). For symbolic computations, I highly...

This usually occurs because you problem is badly scaled. In broad terms: it is ill-conditioned because the eigenvalues of your matrices have differend orders of magnitude. Try to define scaled variables by dividing your original variables by (non-zero) constants (or via linear tranformations) so...

I obviously depends on what you want to do with it. Background: I've used Mathematica for 12 years now. I mostly use it to do pure math symbolic calculations (i.e. group theory). 1. Numerical optimization in Mathematica. I strongly DO NOT recommend Mathematica for that task if your...

What if every function evaluation takes 10 seconds or more ? This "multi-start" approach is good only as a first rough test in the case of inexpensive function evaluations (in terms of calculation times), but it is rather slow and will fail on many problems.

Here are my 2 cents. Short answer NMinimize and FindMinimum don't work with "black-box" functions. The only Mathematica package I know that supports this is Global Optimization for Mathematica. Long answer 1. Using finite differences to approximate derivatives can lead to disastrous results...