On 3/17/07, Nicolas Pettiaux <nicolas.pettiaux at ael.be> wrote:
> I also wanted to ask a similar question, as I my department, I was due
> to show why to use octave instead of matlab for the numerical analysis
> courses,
The advantage of Matlab over Octave (as I have heard from
a friend who is a professor in a college of engineering) is that
you can buy add-on packages which are very powerful.
That's not useful for students in numerical analysis classes,
who are expected (indeed, required) to re-invent the wheel.
Having used Octave for various numerical problems, I think
it would be fine for teaching students numerical analysis.
If you used Matlab for the same purpose, the programs would
not differ in any substantial way.
> I am buiding my arguments, that we should use a free system, and if
> looking beyond simply numerical computations, then go for SAGE for
> example, that offers a free environnement (it is a free software)
> python as programming language, with the power and features of a cas
> (thanks to Maxima and others), numerical tool (thanks to python
> matlplotlib and scipy, but also octave if needed) and a nice web
> interface (much like Mathematica).
Well, at this point I like the R statistical system for strictly
numerical computations (it subsumes the capabilities of Octave
in this respect and then has all the statistical stuff, and better
graphics). For numerics + symbolics, I hope eventually Maxima
will be the answer. Recently (last year or two) Maxima's
numerical capabilities have expanded a lot, but coverage is
patchy. But we are making good progress.
FWIW
Robert