[MMTK] Scientific python installation problems

Konrad Hinsen hinsen at cnrs-orleans.fr
Mon Oct 6 20:28:59 CEST 2003

On Monday 06 October 2003 10:27, Jonathan Turner wrote:

> Now, it would appear that the whole installation is doing funny things as
> it doesn't even get to the first line of the script.

That usually indicates an installation problem, right.

> I re-installed everything. I went in the order python2.2, numeric, netCDF,
> scientific, then MMTK. Running the site.py script to identify site
> packages, it doesn't recognise the netCDF - is this OK? (I assume that as
> it is not a python package per se that it will be OK)

Right. netCDF has nothing do to with Python. The Python interface for netCDF 
is part of Scientific Python (module Scientific.IO.NetCDF)

As for ideas... if you used binary RPMs, they could be incompatible. In 
particular, C modules that use Numeric (e.g. the netCDF interface, but also 
many MMTK modules) must be compiled with the header files for exactly the 
same version of Numeric that is installed.

In case of doubt, build your own binary RPMs for ScientificPython and MMTK. 
The procedure is simple:
1) Get the source distribution (*.tar.gz)
2) Unpack and enter the newly created directory
3) python setup.py bdist_rpm
4) The binary RPM is in the directory "dist".

Make sure to build and install ScientificPython before building MMTK.

Konrad Hinsen                            | E-Mail: hinsen at cnrs-orleans.fr
Centre de Biophysique Moleculaire (CNRS) | Tel.: +33-
Rue Charles Sadron                       | Fax:  +33-
45071 Orleans Cedex 2                    | Deutsch/Esperanto/English/
France                                   | Nederlands/Francais

More information about the mmtk mailing list