If you have numerical data to process, if you want to code some complex mathematical operations, or if you want to output some figures on nice graphs, scipy could be for you.
This package adds a module to the python language that allows it to do scientific data processing.
Using it, and a few friends (ipython, python-matplotlib) you have a numeric capable high-level language that allows both to replace matlab for interactive data processing, to run extensive computations, and even to build powerful GUIs for experiment control.
In the late 70s optimized fortran routines where wrapped in a high level language, MATLAB. The resulting specialized language, and its competitors, had an enormous success among engineers, as it allowed them to focus on their mathematical problem, without worrying about computing problems, such as variable types.
However these languages are very rich in math operations, but very poor in other fields. The scipy python modules adds rich numerical types and mathematical operations to an already very rich and conveniant language: python. This profits both the engineer using scipy, as he can benefit from python’s extensive librairy, and the python programmer, who can pick optimized numerical functions in scipy’s toolbox for his general purpose program.
The scipy community is very active and scipy is gaining momentum. It is a great tool to teach computing to physics and engineering studing. The goal of the project is to make coding math as simple as possible.
python-scipy is available in Debian (0.3.2 in sarge, 0.5.1 in testing/unstable) and Ubuntu (0.3.2 in dapper, 0.5.1 in edgy).