problem in

Konrad Hinsen
Mon, 15 Feb 1999 11:58:25 +0100

> I've installed MMTK in a RS/6000 and I'm currently testing it.  I ran
> the example but at the end it crashed with the following
> message:
> Traceback (innermost last):
>   File "<stdin>", line 81, in ?
>   File "/usr/local/mmtk/", line 48, in __call__
>     _dynamics.integrateVV(self.universe, f, p,
> NetCDFError: Unknown error

First of all, this works for me, and I am using an RS/6000 as well.
Can you run the example ""? It uses exactly the same
trajectory generation procedure, but runs much quicker ;-)

As for the "unknown error", this just means "something went wrong";
the netCDF library is not very specific about error reporting, or at
least it doesn't follow its own documentation, at least not in all
versions. In other words, error checking with netCDF is a pain, and I
have never felt motivated enough to explore in detail how it really

Two typical causes for "unknown errors" are insufficient disk space
and insufficient file access permissions.

You should also make sure to have the latest versions of netCDF and
the Python netcdf module.

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