Remarks on Python for scientific computing

Introduction - what is Python and why ?

Python is an interpreted, object-oriented programming language. In addition to scripts/programs python can be used interactively.
Together with the NumPy (ie Numeric or Numerical) extension (see below) and SciPy it is an interesting free alternative to MATLAB or similar packages; in addition one has the advantage of a complete programming language.

There are many introductions to python both for beginners in programming and those familiar with other languages, see the documentation section on the Python Homepage. Personally, I found the following quite helpful

Python Tutorial
How to Think Like a Computer Scientist
Dive Into Python (aimed at more experienced programmers).

In addition you may want to have a look at some material I collected for our course on computational physics with python

Webpage der Vorlesung (German)
Some python links

What packages do you need ?

The following is an incomplete list of some relevant packages

Numeric
SciPy
For interactive work I highly recommend to use IPython
Gnuplot.py - python interface to gnuplot

Some other pages of interest

PyScience Home contains several links related to Python and Scientific Computing (though the listed SIGs (Special Interest Groups) Matrix-SIG and Plot-SIG are retired and thus more of historical interest).
Plotting is (as always) a bit more tricky, for lists of possible plotting packages which can be used from python see Graphical Representations of Data and Plot-Resources.
For visualization see the MayaVi Data Visualizer.