[Dune] Problem when instantiating a CG solver object with a Gauss Seidel Preconditioner

[Oswald Benedikt]

Hi Markus, thanks. The idea was to test the sequential Gauss-Seidel Preconditioner before
parallelizing it.

To wrap the seq GS, what would I do precisely ?

Just declare a new preconditioner class and inherit the  respective functions
from the sequential Gauss Seidel ?

Greetings, Benedikt


------------------------------------------------------------------------------------------------------------------
Benedikt Oswald, Dr. sc. techn., dipl. El. Ing. ETH, www.psi.ch, Computational Accelerator Scientist
Paul Scherrer  Institute (PSI), CH-5232 Villigen, Suisse, benedikt.oswald at psi.ch, +41(0)56 310 32 12
"Passion is required for any great work, and for the Revolution passion and audacity are required in big doses.", 
Ernesto 'Che' Guevara, Letter to his parents.
http://amas.web.psi.ch
------------------------------------------------------------------------------------------------------------------




-----Original Message-----
From: dune-bounces+benedikt.oswald=psi.ch at dune-project.org on behalf of Markus Blatt
Sent: Sat 5/1/2010 20:31
To: dune at dune-project.org
Subject: Re: [Dune] Problem when instantiating a CG solver object with a Gauss Seidel Preconditioner
 
Hi Benedikt,

On Fri, Apr 30, 2010 at 11:31:59PM +0200, Oswald Benedikt wrote:
> Hi Christian, we need a non-overlapping scalar product since the solver itself runs parallel
> and thus needs a parallel implementation of the dot product.
> 

Sill, the operator, scalarproduct, and preconditioner need to be of
the same category. Apparently, you are trying to use a sequential
preconditioner with a non-overlapping (i.e. parallel) scalar
product. This is an error.

BTW I doubt that it worked with SeqJacobi. Maybe used a parallel
wrapper around the preconditioner before?

Cheers,

Markus


_______________________________________________
Dune mailing list
Dune at dune-project.org
http://lists.dune-project.org/mailman/listinfo/dune