Why use Python for scientific computing?

For example, recently a very similar thread was started:

http://mail.scipy.org/pipermail/scipy-user/2009-July/021706.html

you can also read it here formatted by google if you prefer:

http://groups.google.it/group/scipy-user/browse_thread/thread/0eb1517095963c9a#

There are a number of points made there that I won’t repeat, which you could find useful.

I was never a ‘faithful’ matlab user, but I did use IDL a lot while in graduate school, and I have many colleagues who had thousands of lines of production matlab code and still found it worthwhile to switch to python (even years ago, when the various tools were far less mature and powerful than they are today). But for each person, the answer to this question can be different, and what works for me may not work for you.

I’ve made a number of ‘pro python’ points in that article and elsewhere, but python is certainly not perfect (no tool is). I find it to be the best compromise tool from what exists today, for high-level scientific computing development, and it continually improves.