[SciPy-User] [Fwd: Re: ANN: SfePy 2009.4]

Robert Cimrman cimrman3@ntc.zcu...
Fri Nov 27 02:08:44 CST 2009


Hi Chris,

thanks for trying sfepy!

Let's discuss this at sfepy-devel, or, if you do not want to register there, 
write me personally, this is IMHO rather off-topic for the scipy-user list.

S. Chris Colbert wrote:
> I'm getting all sorts of errors trying to run sfepy tests and examples:

Can you copy & paste the output and send it to me? I wonder what errors you get.

> It builds fine. 
> 
> But I fail one of the solvers test because of a bug with OpenMPI (whichever 
> solver is using Petsc4py which btw, is not listed as a dependency).

Yes, PETSc (with Petsc4py) are optional packages, used only by 
test_linear_solvers.py test file. It should not fail, however, only skip the test.

> The schroedinger example runs, but produces erroneous output (~300% error). 
> The poisson and valec examples produce error results.  
> 
> System specs:
> Kubuntu 9.10 x64
> 
> Self built/easy_insall:
> Numpy 1.3.0
> Scipy 0.7.1 
> Newest umfpack scikit
> Newest Petsc4Py
> pytables 
> 
> from the repos:
> hdf5-serial 
> openmpi 1.6.6
> pysparse
> 
> Any help would be awesome!

I will gladly help you, but I need more information - could you send me 
(off-list) the full outputs of the simulations that do not work, or produce 
error results? Also attach the solution files if you find it necessary.

Thanks,
r.


More information about the SciPy-User mailing list