[deal.II] How can I visit a face exactly once when running with MPI?

2023-04-27 Thread Abbas
I want to postporcess some data across an interface with feInterface. I have a fully::distributed::triangulation. I am doing is something that "roughly" looks like this: template void TSL_prllel::interface_gradients(Tensor<1, dim> & interface_gradients) { interface_gradients = 0; QGauss

[deal.II] Re: How can I visit a face exactly once when running with MPI?

2023-04-27 Thread Bruno Turcksin
Hello, On Thursday, April 27, 2023 at 7:56:50 AM UTC-4 abbas.b...@gmail.com wrote: This works with a single MPI but not for more. Is it because the cell->face( face)->clear_user_flag() doesn't clear the user flag in the other distributed meshes? That's correct. The flag is only cleared

[deal.II] Re: How can I visit a face exactly once when running with MPI?

2023-04-27 Thread Abbas Ballout
This works!! Thanks. I am guessing something like: if(cell_neighbour->is_ghost()==true) cell->face(neighbour_face_number)->clear_user_flag(); won't clear the user flag everywhere either, because that didn't work. Right? On Thursday, April 27, 2023 at 2:48:23 PM UTC+2 bruno.t...@gmail.com

[deal.II] deal.II Newsletter #251

2023-04-27 Thread 'Rene Gassmoeller' via deal.II User Group
Hello everyone! This is deal.II newsletter #251. It automatically reports recently merged features and discussions about the deal.II finite element library. ## Below you find a list of recently proposed or merged features: #15151: Allow MappingInfo::reinit_cells() to work on containers

Re: [deal.II] Re: How can I visit a face exactly once when running with MPI?

2023-04-27 Thread Bruno Turcksin
Right, it won't clear it everywhere. You would need to use MPI to clear the flag on the other processor. If you would do that you would create a huge amount of messages and your code would be extremely slow. Best, Bruno Le jeu. 27 avr. 2023 à 10:09, Abbas Ballout a écrit : > This works!!

[deal.II] Internal deal.II error when instantiating TrilinosWrappers::block_operator

2023-04-27 Thread Lucas Myers
Hi folks, I'm trying to use the block_operator technique on Trilinos vectors and matrices, but I can't quite get it to compile. For simplicity, the relevant lines of code are: ``` #include #include using vec = dealii::LinearAlgebraTrilinos::MPI::Vector; using block_vec =

Re: [deal.II] Re: How can I visit a face exactly once when running with MPI?

2023-04-27 Thread Timo Heister
I would suggest that you use MeshWorker::mesh_loop() to do this (or look at the complicated implementation of it). On Thu, Apr 27, 2023 at 11:18 AM Bruno Turcksin wrote: > > Right, it won't clear it everywhere. You would need to use MPI to clear the > flag on the other processor. If you would

Re: [deal.II] Negative determinant error with tetrahedrals

2023-04-27 Thread Wolfgang Bangerth
Kumar, can you come up with a small example that illustrates the problem? We can't tell where the error might come from without seeing a (complete) piece of code and an input mesh. Best W. On 4/27/23 16:09, Kumar Saurabh wrote: *** Caution: EXTERNAL Sender *** When I am trying to use

Re: [deal.II] Internal deal.II error when instantiating TrilinosWrappers::block_operator

2023-04-27 Thread Wolfgang Bangerth
Lucas: I'm trying to use the block_operator technique on Trilinos vectors and matrices, but I can't quite get it to compile. For simplicity, the relevant lines of code are: ``` #include #include using vec = dealii::LinearAlgebraTrilinos::MPI::Vector; using block_vec =