[petsc-users] Partitioning a mesh with petsc4py

[Matthew Knepley]

On Wed, Jul 8, 2015 at 5:16 AM, Javier Quinteros <javier at gfz-potsdam.de>
wrote:

> -----BEGIN PGP SIGNED MESSAGE-----
> Hash: SHA1
>
> Hi all.
> We use petsc4py to develop. We are trying to partition a mesh (DMPlex)
> but we receive different errors when calling to the "distribute"
> method of the "DMPlex" class.
>
> We use versions 3.6 from PETSc and petsc4py.
> We run it in two different installations. One has been installed with
> - --download-metis and the other with --download-chaco, but the error is
> always:
>
>   File "PETSc/DMPlex.pyx", line 514, in
> petsc4py.PETSc.DMPlex.distribute (src/petsc4py.PETSc.c:220066)
>     self.dm.distribute()
>   File "PETSc/DMPlex.pyx", line 514, in
> petsc4py.PETSc.DMPlex.distribute (src/petsc4py.PETSc.c:220066)
> petsc4py.PETSc.Errorpetsc4py.PETSc.Error: error code 56
> [0] DMPlexDistribute() line 1505 in
> /home/javier/soft/petsc-3.6.0/src/dm/impls/plex/plexdistribute.c
> [0] PetscPartitionerPartition() line 653 in
> /home/javier/soft/petsc-3.6.0/src/dm/impls/plex/plexpartition.c
> [0] PetscPartitionerPartition_Chaco() line 1048 in
> /home/javier/soft/petsc-3.6.0/src/dm/impls/plex/plexpartition.c
> [0] No support for this operation for this object type
> [0] Mesh partitioning needs external package support.
> Please reconfigure with --download-chaco.
>

Can you send the configure.log from the installation using --download-chaco?

  Thanks,

     Matt


> We build our mesh using "createFromCellList" as in firedrake (plex is
> empty on ranks different than 0), and there is the folowwing comment:
>
> # Provide empty plex on other ranks
> # A subsequent call to plex.distribute() takes care of parallel
> # partitioning
>
> But for us is not working.
>
> Does anyone have a working example of how to partition a mesh in petsc4py?
>
> Thanks in advance!
>
> - --
> Javier Quinteros
> - -------------------------------------------
> 2.4/Seismologie
> Tel.: +49 (0)331/288-1931
> Fax:  +49 (0)331/288-1277
> Email: javier at gfz-potsdam.de
> ___________________________________
>
> Helmholtz-Zentrum Potsdam
> Deutsches GeoForschungsZentrum GFZ
> Stiftung des öff. Rechts Land Brandenburg
> Telegrafenberg, 14473 Potsdam
> -----BEGIN PGP SIGNATURE-----
> Version: GnuPG v1
>
> iQEcBAEBAgAGBQJVnPhnAAoJEB7cuSNr8cJ5oxUIALmvXzbCD+vbiPjlDLcTOzV8
> QLeYPW7kLHqL6jZuPkIu+lubsgsdYfsvE8vtfm3FwvlYlQyuez8vMI5V74aLVGsb
> yMQGqjWvFMbJ8ymvJ8K+COEM7G8oahB5Y9I5Wl30+psRhE41ZCWHeVSruqidSCo5
> T3buMWOjUq1Il7dtpfgnaiRReYWvqjcuW9LW6sU6HPux7LgVzhbtNhiurYfJ8FZX
> PKT+mW9PFPJe4U+HQcvyT/augmr3+Zt8QQQcSFMe0GVMHtiZb8eXg9CcouC3qA91
> UaMVkCOz2XXmsgY6aKih9xhOh6QP3i8HM7n46YC1B7BdKqWtwHPh8PqgNH2WWbE=
> =KSm9
> -----END PGP SIGNATURE-----
>



-- 
What most experimenters take for granted before they begin their
experiments is infinitely more interesting than any results to which their
experiments lead.
-- Norbert Wiener
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.mcs.anl.gov/pipermail/petsc-users/attachments/20150708/615abc7a/attachment.html>