[SciPy-user] Matlab, Scipy and teaching science

[Peter Bowyer]

Hi Robert,

At 10:33 20/10/2005, Robert Kern wrote:
>Matlab cripples scientists who learn it as their first introduction to
>"real" programming. They continue to do *everything* in Matlab even when
>Matlab is the most hideously wrong tool to use because they've learned
>Matlab and using any other language would mean spending the effort to
>learn the new language. When they are convinced to learn another
>language, it's usually a low-level language so they can write faster
>code. But then they only write that low-level code as interfaces to
>Matlab so they can continue to do high-level things relatively easily.

Playing the devils advocate, doesn't that apply to learning any new 
programming language - because there's always an inertia to 
overcome?  If you substitute Python for Matlab in the above, do you 
say it doesn't hold true?

One point that people always make when they've heard of Python is 
that "Python is slow, so to write simulations the students will have 
to learn <insert low level language> anyway".  I believe enough 
(informal) studies have been done to show that students with no 
programming experience learn programming faster using Python.

However, then comes the point that "Python does things differently, 
so it's harder to use another langauge".

>As you say, they might have to learn Matlab anyways to deal with other
>researcher's code, but if you get them started right with Python, they
>won't be contributing to the problem.

 From the purists point of view I agree, however one criticism of the 
current course (using the department's own dialect of Basic) is that 
the language doesn't look good on the CV and isn't useful - something 
that Matlab may be able to overcome.

Of course a bigger problem is that the students don't want to learn 
to program in the first place ("We'd have done computer science if we 
liked computers").

>We're working on it.
>   http://peak.telecommunity.com/DevCenter/EasyInstall

Fantastic!

>We have fairly good docstring coverage in scipy at least. We need to
>build some searching capability, though.

Search would be good.

Peter

-- 
Maple Design - quality web design and programming
http://www.mapledesign.co.uk