William Dawn

Thank you so much for taking a look! I'm using the C API right now. My main code uses the [MFEM](https://mfem.org) framework ([github](https://github.com/mfem/mfem)) so I have routines to generate `AMGX_matrix_handle`...

Hi! Just wanted to ping to say that this is a feature I need for my application as well. Thanks!

@v-dobrev I've been experimenting with this and have found a problem for which it does not work. Unfortunately, I'm having trouble understand why it is breaking. I have a very...

@v-dobrev I was able to run `ex1p-test` using my mesh. I also commented out the block where the `Mesh` object is created since it is unnecessary. I'm starting to suspect...

@v-dobrev I think I found my original problem. There seems to be an error when using `attribute` on the mesh. To replicate this error with `ex1p-test`, simply use a `PWConstCoefficient`...

That's a good catch, but unfortunately that doesn't fix the problem. Here is my solution to `ex1p-test` using ```cpp Mesh mesh(mesh_file, 1, 1); ParMesh pmesh(MPI_COMM_WORLD, mesh); ``` And using ```cpp...

@rcarson3 confusingly, the materials seem to be mapped properly when visualizing with glvis. For testing purposes, the mesh that I'm using is actually pretty small (

@rcarson3 I agree. It looks like something is not being communicated while the matrix/operator is being constructed based on the sparsity patterns. I think this is handled by the `GroupTopology`...

@v-dobrev This fixed my issue with `ex1p-test` and I have also tested it and it fixes my issue in my personal code as well. It would have been very hard...

Yes. See #66. The API will function through the end of the year for those with an API key but no more keys are being issued.