[Top][All Lists]

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Matlab central and octave

From: Dmitrey
Subject: Re: Matlab central and octave
Date: Tue, 06 Feb 2007 20:36:04 +0200
User-agent: Thunderbird (Windows/20061207)

Hi David,

David Bateman пишет:
However, the problem is there are lots of people who
are willing to suggest what needs to be done and few to actually do the
work. If there was someone committed enough to support such a forum,
then I'd support them wholehearted, but till someone steps up and says
something like, "I've just created a test website implementing MFA for
Octave, and will support it in the future" then I won't believe in it..
Maybe, after this issue someone will apper?
As for me, I would gladly commit my OpenOpt solvers to the area - even now some of them are good enough, for example ralg() on non-smooth & noisy tasks is better than MATLAB fminsearch (/fminsearchbnd, fminsearchcon, fminsearchOS), adsmax <>, mdsmax <>, other nelder-mead <> implementations, even commercial snopt(), at least at 95% cases that I have tried (but I can't say I tried too much*). *Now it can handle all type of constraints - lb, ub, Ax<=b, Aeq*x=beq, c(x)<0, h(x)=0. New feature - nonSmoothSolve - fsolve analog for non-smooth funcs. I intend to publish new relize 5-10 days later, currently you can taste latest v 0.35 at

All of the people currently developing code for octave and octave-forge
are all to busy on other projects, so it would take someone new to step
forward to do this work..

I don't know anything of the situation, so I can only assume that you are right. However, I have no info what are you doing now (since Octave (like it seems to me) doesn't inform their users what will next reliase contain). As for me, I guess, 2 key features that Octave miss vs MATLAB or even free Python(/scipy, numpy) is 1) at least one convinient IDE (containing easy settlement of breakpoints, better - conditional BP), for example eclipse wrapper - some years ago, as I noticed, there was a discussion about eclipse but no further steps were made - and I guess it was great error.
2) JIT

Best regards, D.

reply via email to

[Prev in Thread] Current Thread [Next in Thread]