why python for the academe?

[Paul Magwene]

On Sat, 16 Mar 2002 21:00:57 -0500, huck wrote:

> hello all,
> 
> i've been lurking for a while and am amused at how many people in the
> list here use python for scientific computing. why not matlab (or octave
> or scilab or whatever)?
> 
> 

Speaking only for myself of course - flexibility, extensibility,
readability, and "shareability"

Flexibility:  As a general programming language I find Python to be a lot
more flexible. I've done the programming thing in specialized environment
like Mathematica/Matlab/Octave/R.  While particular operations are
sometimes simpler in these environements, the general structure of
programs seems to be much more constrained.  When I'm implementing some
new bit of code for my research, often as not, I'm (hopefully) not doing
the same things that people have done previously, and so I appreciate the
flexibility of a general programming language.

Extensibility:  I can use precanned extensions like Numeric when I need an
extra speed boost, or write a quickie interface to some piece of C code I
have lying about (though I HATE to do this - it makes me appreciate python
that much more).

Readability:  I can hand my code to one of my colleagues and chances are
they'll understand what's going on, regardless of whether or not they know
python.

"Shareability": Python is free and widely available (so are Octave Scilab,
R, etc., but not Matlab) and there is a large amount of freely shared code
available, much of it useful.  While many researchers in the US/Europe can
probably afford the $500+ licensing fees for something like Matlab, many
of our colleagues around the world can not (And I'm generally tired of
shelling out more moolah every 12-18 months when the developers of the
commercial packages have gotten around to fixing the bugs in the previous
release)


My-own-two-bits-ly-y'rs,
Paul