Programs included in this archive are written in Python 2.5, and requires the following standard scientific libraries: 
* Numpy (numerical arrays)
* Scipy (scientific tools)

In addition, you may want to install matplotlib (defaut plotting system, syntax similar to Matlab) and ipython (a convenient Python shell). 

All these packages are available in the repositories of major Linux distributions.
For instance, in Ubuntu or Debian, try: 
sudo aptitude install python-numpy python-scipy python-matplotlib ipython

Then, for instance, launch ipython, and type
run -i probit.py

Windows users may download Python Enthought distribution, at
http://www.enthought.com/products/epd.php
which includes all necessary scientific libraries in one convenient package. 


Specific note for decentred Gaussian example: 
In this program, I wanted to experiment with weave, a scipy module that allows to include a C program `inline', that is inside your Python program, and compiles it for you behind the scenes. This works very well and makes the program very fast, but, in case you don't manage to make it work, you can remove the C part and replace it with
the Python lines 138-141, which are currently commented. Note the C bit also requires
the GSL (Gnu Scientific Library).

Specific note for stochastic volatility example:
this needs a few routines found in the particle_tools package, so you need to place
particle_tools in a directory where Python looks for packages to be imported. (You may as well copy and paste these routines in the main program, of course.)