craffael

Results 18 comments of craffael
trafficstars

Okay I've visualized the square_quads.msh using ParaView and second order elements: ![grafik](https://user-images.githubusercontent.com/15138672/57638182-89035600-75ad-11e9-8845-9fd8bc82d47a.png) Then if we refine it once: ![grafik](https://user-images.githubusercontent.com/15138672/57638135-68d39700-75ad-11e9-8302-6ec25aee83b1.png) As you can see there is not really a difference between...

@hiptmair @anianruoss What do you think?

Yes, as a skeleton examples are very useful. But we should strive to have not too many of those examples because, most of them will probably look very similarly. I...

Aha I see, that wasn't clear to me when I looked at the header file and also @TobiasRohner started using one of the methods without knowing that these are not...

@victorapm I've setup a small test program which shows the issue. Along with the test program I provide you matrices for four different mesh sizes (the mesh size is halfed...

@victorapm It would be great if you could have a look. I'm still a bit puzzled by the fact that a few additional basis functions can trouble AMG that much....

Hi all a few comments: - AMS/ADS are not applicable, it is a scalar valued Laplace problem. - I've already tried various (in)complete Schur complement approaches, but they are quite...

@liruipeng I try to summarize a bit what I've tried. For the discussion lets assume the system matrix has the following block structure: ![grafik](https://github.com/hypre-space/hypre/assets/15138672/41448835-2d2f-4591-ab64-eff17969d352) Where C is small `n x...