adam-sim-dev

> > Also a suggestion, rename "Superlu" to "superlu_dist" in mfem. > > > > Do you mean just in the source code? Like `SuperLUSolver` to `SuperLUDistSolver` or something like...

Any progress on this PR? I am interested in the Dorfler marking scheme. Maybe @hughcars could have a review. I see https://github.com/awslabs/palace/pull/110

> > Any progress on this PR? I am interested in the Dorfler marking scheme. Maybe @hughcars could have a review. I see [awslabs/palace#110](https://github.com/awslabs/palace/pull/110) > > We took a different...

> > > > Any progress on this PR? I am interested in the Dorfler marking scheme. Maybe @hughcars could have a review. I see [awslabs/palace#110](https://github.com/awslabs/palace/pull/110) > > > >...

> A next step might be to use different MatrixConstantCoefficient values in different regions of your domain tagged using the element attributes of the mesh. Hi Mark, @mlstowell , is...

Dear @acfisher and other developers, my application need all the elements containing one vertex for the `ParMesh`. For the mesh, I could call `Mesh::GetVertexToElementTable()`, which was introduced in this PR....

> Hi @adam-sim-dev, > > Are you referring to [this discussion](https://hypre.readthedocs.io/en/latest/solvers-ams.html#:~:text=For%20(singular)%20problems,HYPRE_ParVector%20x) in the AMS documentation? > > We don't have direct support for `HYPRE_AMSSetInteriorNodes` but that something that can be...

I now changed PCG to MINRES. FGMRES/MINRES can solve the linear system, however, it converges too slowly for some models （the low frequency, and high order models）. So, I am...

I have modified `void HypreAMS::MakeGradientAndInterpolation(ParFiniteElementSpace *edge_fespace, int cycle_type, Array &zero)` to ``` // generate and set the vertex coordinates or Nedelec interpolation matrices x = y = z = NULL;...

> > (3) I do not know how to get vertex for the high-order meshes. > > Is your Nedelec space high-order or is it first order Nedelec space on...