dolfin team mailing list archive

[Anders Logg <logg@xxxxxxxxx>]

Thanks. I will look at it next week.

I'll be adding unit tests.

This is mostly for use in the tensorproduct/kronecker/timedependent
branches but will have some side effect features like this one.

--
Anders


On Fri, Aug 17, 2012 at 07:40:24PM +0100, Garth N. Wells wrote:
> On 17 August 2012 19:15, Anders Logg <logg@xxxxxxxxx> wrote:
> > ok, can you give any details or code that I can look at?
> >
>
> You probably want to implement the Epetra_Operator class:
>
>     http://trilinos.sandia.gov/packages/docs/r10.12/packages/epetra/doc/html/classEpetra__Operator.html
>
> We could really do with some matrix-free unit tests.
>
> Garth
>
>
> >
> >
> > On Fri, Aug 17, 2012 at 08:06:42PM +0200, Kent-Andre Mardal wrote:
> >>    Yes, I think so. I've used ML in this fashion.
> >>    Kent
> >>
> >>    On 17 August 2012 16:28, Anders Logg <[1]logg@xxxxxxxxx> wrote:
> >>
> >>      I want to use PETSc MatShell to create a matrix-free representation
> >>      of
> >>      a linear system using PETScKrylovMatrix. Is there similar
> >>      functionality in Trilinos, in which case I could look into adding a
> >>      common interface matrix-free Krylov solvers?
> >>
> >> Referenser
> >>
> >>    1. mailto:logg@xxxxxxxxx
> >>    2. https://launchpad.net/%7Edolfin
> >>    3. mailto:dolfin@xxxxxxxxxxxxxxxxxxx
> >>    4. https://launchpad.net/%7Edolfin
> >>    5. https://help.launchpad.net/ListHelp
> >
> > _______________________________________________
> > Mailing list: https://launchpad.net/~dolfin
> > Post to     : dolfin@xxxxxxxxxxxxxxxxxxx
> > Unsubscribe : https://launchpad.net/~dolfin
> > More help   : https://help.launchpad.net/ListHelp
>
>
>