Prerequisites

Refer to Installing for the list of required packages. If the required packages have already been installed then skip the corresponding commands below.

(Step 0) Essential Compilers and Libraries

The instructions below assume that the following Ubuntu packages have been installed:

sudo apt-get install build-essential gfortran libatlas-sse2-dev 
sudo apt-get install python-all-dev ipython
sudo apt-get install subversion

When using other Fortran compilers (e.g. f77) or BLAS/LAPACK libraries, ensure that the combination is compatible.

(Step 1) Download Source Code

More detailed instructions for NumPy and SciPy are available here.

(Recommended) Obtain official versions of nose, NumPy, SciPy, and PyAMG

wget http://python-nose.googlecode.com/files/nose-0.10.1.tar.gz
wget http://superb-east.dl.sourceforge.net/sourceforge/numpy/numpy-1.2.1.tar.gz
wget http://voxel.dl.sourceforge.net/sourceforge/scipy/scipy-0.7.0.tar.gz
wget http://pyamg.googlecode.com/files/pyamg-1.0.0.tar.gz
tar xvfz nose-0.10.1.tar.gz
tar xvfz numpy-1.2.1.tar.gz
tar xvfz scipy-0.7.0.tar.gz
tar xvfz pyamg-1.0.0.tar.gz

Alternative Links: Nose 0.10.1 Numpy 1.2.1 SciPy 0.7 PyAMG 1.0

(Alternative) Obtain development versions of nose, NumPy, SciPy, and PyAMG

svn checkout http://python-nose.googlecode.com/svn/trunk/ nose
svn checkout http://svn.scipy.org/svn/numpy/trunk/ numpy
svn checkout http://svn.scipy.org/svn/scipy/trunk/ scipy
svn checkout http://pyamg.googlecode.com/svn/trunk/ pyamg

(Step 2) Install Packages

Install nose

cd nose
sudo python setup.py install
cd ..

Install NumPy

cd numpy
python setup.py build
sudo python setup.py install
cd ..

Install SciPy

If you have umfpack installed from apt-get install suitesparse suitesparse-dev, put the following site.cfg file in your scipy directory:

[amd]
library_dirs = /usr/lib
include_dirs = /usr/include/suitesparse
amd_libs = amd

[umfpack]
library_dirs = /usr/lib
include_dirs = /usr/include/suitesparse
umfpack_libs = umfpack

Then follow with

cd scipy
python setup.py build
sudo python setup.py install
cd ..

Install PyAMG

cd pyamg
python setup.py build
sudo python setup.py install

(Step 3) Test Installation

Using Python or IPython (recommended) to run the following commands

import pyamg
pyamg.test()

should have output similar to

Running unit tests for pyamg
NumPy version 1.2.1
NumPy is installed in /usr/lib/python2.5/site-packages/numpy
SciPy version 0.7.0
SciPy is installed in /usr/lib/python2.5/site-packages/scipy
Python version 2.5.2 (r252:60911, Jul 31 2008, 17:31:22) [GCC 4.2.3 (Ubuntu 4.2.3-2ubuntu7)]
nose version 0.10.4
PyAMG version 1.0.0
PyAMG is installed in /usr/lib/python2.5/site-packages/pyamg
..........................................................................................................................
----------------------------------------------------------------------
Ran 122 tests in 27.593s

OK

Testing NumPy and SciPy is done in similar fashion:

import numpy
numpy.test()

import scipy
scipy.test()