I forward to Francois Pellegrini.
We take a look on that problem tomorrow, but I think you can
automatically remove the metis files from scotch in PETSc without
problems. Scotch uses them only to provide to users the Metis and
ParMetis interface with the libraries libscotchmetis.a and
libscotc
On Jan 5, 2010, at 4:56 PM, Mathieu Faverge wrote:
> I forward to Francois Pellegrini.
> We take a look on that problem tomorrow, but I think you can
> automatically remove the metis files from scotch in PETSc without
> problems.
Thanks for the suggestion, when PETSc installs Scotch it
On Tue, 5 Jan 2010 14:24:36 -0600, Dmitry Karpeev
wrote:
> So an assembled MatFwk would store these block index arrays internally
> (and, possibly, R_ks matrix entries, if we need more than just
> ADD_VALUES, as is the case for BDDC
How is something other than ADD_VALUES necessary for BDDC? Onc
Scotch has a file called metis.h this is a big problem when
automatic installers are installing both ParMetis and Scotch since
automatic installers are not smart enough to read header file notes
and figure out exactly what file to delete. I would have reported this
as a bug to Scotch b
any results to which their
>> experiments lead.
>> -- Norbert Wiener
>>
>
>
--
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-dev/attachments/20100105/0653eb9f/attachment.html>
On Tue, Jan 5, 2010 at 11:42 AM, Jed Brown wrote:
> On Tue, 5 Jan 2010 10:52:36 -0600, Dmitry Karpeev
> wrote:
>> Automating decomposition/assembly of operators relative to a
>> decomposition of identity (doi) would, in my opinion, go a long way
>> towards decoupling Mat from DM. ?DM would then
Sorry for jumping in late.
PDE discretizations and solvers many geometric decompositions of the
linear space
they ultimately operate on. However, many of these decompositions can
be expressed
rather naturally in linear algebraic terms (e.g., "factorization" of
the stiffness matrix into
a "sum" of
On Tue, 5 Jan 2010 10:52:36 -0600, Dmitry Karpeev
wrote:
> Automating decomposition/assembly of operators relative to a
> decomposition of identity (doi) would, in my opinion, go a long way
> towards decoupling Mat from DM. DM would then convert their geometric
> data into a corresponding doi an