On Thu, Mar 17, 2016 at 4:15 PM, Roy Stogner <royst...@ices.utexas.edu> wrote:
> > On Thu, 17 Mar 2016, David Knezevic wrote: > > This would add all constrained dofs to the list of condensed dofs. >> That should do what we want, I agree. >> > > I'll put in a PR now. > > Looks like a few other bits to fix while I'm at it; e.g. unsigned int > instead of dof_id_type. > > Would you want to get rid of the argument "std::set<unsigned int> & >> global_dirichlet_dofs_set"? I can imagine using that to set "extra" >> condensed dofs, but I don't think that's a very common use-case, and >> anyway users could just add DofConstraintRows to handle that case if >> needed. >> > > I'll make it const and give it an empty set default value. No need > to break the API, and who knows, someone might want to condense away > unconstrained dofs too someday. > OK, perfect, thanks. We should update the eigenproblem_ex3 to use DirichletBoundary too. I can do that later if you don't get around to it in this PR. Torquil: Once this change is in, it'd be good if you can modify your example to use DirichletBoundary and re-run it to check if it fixes your problem. David ------------------------------------------------------------------------------ Transform Data into Opportunity. Accelerate data analysis in your applications with Intel Data Analytics Acceleration Library. Click to learn more. http://pubads.g.doubleclick.net/gampad/clk?id=278785231&iu=/4140 _______________________________________________ Libmesh-users mailing list Libmesh-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/libmesh-users
