|
schwz
Generated automatically from develop
|
The solver metadata struct. More...
#include <settings.hpp>
Classes | |
| struct | post_process_data |
| The struct used for storing data for post-processing. More... | |
Public Attributes | |
| MPI_Comm | mpi_communicator |
| The MPI communicator. | |
| gko::size_type | global_size = 0 |
| The size of the global matrix. | |
| gko::size_type | oned_laplacian_size = 0 |
| The size of the 1 dimensional laplacian grid. | |
| gko::size_type | local_size = 0 |
| The size of the local subdomain matrix. | |
| gko::size_type | local_size_x = 0 |
| The size of the local subdomain matrix + the overlap. | |
| gko::size_type | local_size_o = 0 |
| The size of the local subdomain matrix + the overlap. | |
| gko::size_type | overlap_size = 0 |
| The size of the overlap between the subdomains. | |
| gko::size_type | num_subdomains = 1 |
| The number of subdomains used within the solver. | |
| int | my_rank |
| The rank of the subdomain. | |
| int | my_local_rank |
| The local rank of the subdomain. | |
| int | local_num_procs |
| The local number of procs in the subdomain. | |
| int | comm_size |
| The number of subdomains used within the solver, size of the communicator. | |
| int | num_threads |
| The number of threads used within the solver for each subdomain. | |
| IndexType | iter_count |
| The iteration count of the solver. | |
| ValueType | tolerance |
| The tolerance of the complete solver. More... | |
| ValueType | local_solver_tolerance |
| The tolerance of the local solver in case of an iterative solve. More... | |
| IndexType | max_iters |
| The maximum iteration count of the Schwarz solver. | |
| IndexType | local_max_iters |
| The maximum iteration count of the local iterative solver. | |
| IndexType | updated_max_iters |
| The updated maximum iteration count of the local iterative solver. | |
| std::string | local_precond |
| Local preconditioner. | |
| unsigned int | precond_max_block_size |
| The maximum block size for the preconditioner. | |
| ValueType | current_residual_norm = -1.0 |
| The current residual norm of the subdomain. | |
| ValueType | min_residual_norm = -1.0 |
| The minimum residual norm of the subdomain. | |
| std::vector< std::tuple< int, int, int, std::string, std::vector< ValueType > > > | time_struct |
| The struct used to measure the timings of each function within the solver loop. | |
| std::vector< std::tuple< int, std::vector< std::tuple< int, int > >, std::vector< std::tuple< int, int > >, int, int > > | comm_data_struct |
| The struct used to measure the timings of each function within the solver loop. | |
| post_process_data | post_process_data |
| double | init_mpi_wtime = 0.0 |
| std::shared_ptr< gko::Array< IndexType > > | global_to_local |
| The mapping containing the global to local indices. | |
| std::shared_ptr< gko::Array< IndexType > > | local_to_global |
| The mapping containing the local to global indices. | |
| gko::Array< IndexType > | overlap_row |
| The overlap row indices. | |
| std::shared_ptr< gko::Array< IndexType > > | first_row |
| The starting row of each subdomain in the matrix. | |
| std::shared_ptr< gko::Array< IndexType > > | permutation |
| The permutation used for the re-ordering. | |
| std::shared_ptr< gko::Array< IndexType > > | i_permutation |
| The inverse permutation used for the re-ordering. | |
The solver metadata struct.
| ValueType | The type of the floating point values. |
| IndexType | The type of the index type values. |
| ValueType schwz::Metadata< ValueType, IndexType >::local_solver_tolerance |
The tolerance of the local solver in case of an iterative solve.
The residual norm reduction required.
| ValueType schwz::Metadata< ValueType, IndexType >::tolerance |
The tolerance of the complete solver.
The residual norm reduction required.
1.8.13