Dear Roy, I don't need to do the refinements. Currently, my code is correct in sequential implementation. However, it is wrong in parallel implementation. I want to check what happen by comparison between sequential and parallel implementation. If I do it, I need to get the same Mass matrix with the same index. Is it possible? how to do it? thanks so much.
Regards, Yujie On Wed, May 26, 2010 at 10:26 AM, Roy Stogner <[email protected]> wrote: > > On Wed, 26 May 2010, Yujie wrote: > >> Whether it is possible to keep the same indices of nodes of the mesh >> and Dofs in sequential and parallel implemenation? > > Definitely not easily, and probably not a good idea. You can't even > keep the same DoF indices in the same run after a refinement step, so > if you're writing code that depends on those indices being constant > you may want to find another way to do it. > --- > Roy > ------------------------------------------------------------------------------ _______________________________________________ Libmesh-users mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/libmesh-users
