INSTALL

COMPILING AND INSTALLING

For the compilation of CAMFR, a C++ compiler is needed that supports
many features from the latest C++ standard. Use e.g. a recent version
(3.3 or later) of the gcc compiler from http://gcc.gnu.org.

CAMFR also requires the following external programs or libraries, 
which must be installed first and built with the same version of gcc used
to compile CAMFR:.

- Python 2.4 from http://www.python.org
- NumPy and SciPy from http://www.scipy.org/
- matplotlib from http://matplotlib.sourceforge.net/
- The Scons building tool from http://www.scons.org. (version 0.9 or later)
- numpy 1.0 or later from http://numpy.scipy.org
- The Boost Python Library from http://www.boost.org. At least version 1.33
  is required. For building it, use a comand like e.g. 
    bjam -sBUILD=release -sPYTHON_VERSION=2.4 -sPYTHON_ROOT=/usr
- Blitz++ from http://oonumerics.org/blitz. Use version 0.9 or later.
- the Fortran linear algebra libraries BLAS and LAPACK, which you can get from 
  http://www.netlib.org. If possible however, use a vendor-supplied version 
  (like Intel MKL) for maximum performance.
- for visualisation, the Python Image Library from 
  www.pythonware.com/products/pil/.

Now copy one of the machine_cfg.py.{various_platforms} files to 
machine_cfg.py, and edit it to reflect the location of these libraries, as 
well as other machine-dependent parameters and compiler flags.

Install the CAMFR library by typing 'python setup.py install' as root.