[MMTK] netCDF module is missing.

Konrad Hinsen hinsen at cnrs-orleans.fr
Fri Apr 11 06:24:45 UTC 2008

On 10.04.2008, at 17:33, Ulderico wrote:

> But the error with xNmoldin is still there.
> We have check the installation of the various libnetcdf*  in  /usr/ 
> lib (also usr/local/lib) and they seem ok.
> Is there a way to verify that MMTK and netCDF do actually interact?

Yes, but it's a two-step procedure:

1) Check that ScientificPython knows about netCDF by running the  
following Python script:

	from Scientific.IO.NetCDF import NetCDF

Yes, just that one line. If nothing happens, ScientificPython is  
fine. If it prints any error message (about not being able to import  
something, or about missing symbols in the shared library), then  
something went wrong during installation and the only solution is to  
redo the installation of ScientificPython from scratch.

2) Check that MMTK interacts correctly with netCDF by running the  
following script:

	from MMTK.Trajectory import Trajectory

If nothing happens, everything is fine.  If it says "Trajectories are  
not available because the netCDF module is missing.", and assuming  
that the first test showed no problems, MMTK was compiled before a  
complete installation of ScientificPython was done. The solution is  
to reinstall MMTK from scratch.

Konrad Hinsen
Centre de Biophysique Moléculaire, CNRS Orléans
Synchrotron Soleil - Division Expériences
Saint Aubin - BP 48
91192 Gif sur Yvette Cedex, France
Tel. +33-1 69 35 97 15
E-Mail: hinsen at cnrs-orleans.fr

More information about the mmtk mailing list