version 0.6.0
Loading...
Searching...
No Matches
type_solver_hypre::t_hypre_precond_parcsr_boomeramg Type Reference

ParCSR BoomerAMG precond.

+ Inheritance diagram for type_solver_hypre::t_hypre_precond_parcsr_boomeramg:

Public Attributes

real(c_double) strong_threshold = 0.25_C_DOUBLE
 AMG strength threshold.
 
integer(hypre_int) coarsen_type = 6
 Parallel coarsening algorithm.
 
integer(hypre_int) interp_type = 0
 Parallel interpolation operator.
 
integer(hypre_int) post_interp_type = 0
 
integer(hypre_int) relax_type = 1
 Smoother algorithm.
 
integer(hypre_int) agg_num_levels = 0
 Number of levels of aggressive coarsening.
 
integer(hypre_int) agg_num_paths = 1
 Number of paths for aggressive coarsening.
 
real(c_double) interp_trunc = 0.0_C_DOUBLE
 Interpolation Truncation.
 
integer(hypre_int) p_max = 5
 Maximum number of elements per row for the interpolation.
 
integer(hypre_int) inter_type_agg = 4
 Interolation used on levels of aggressive coarsening.
 
- Public Attributes inherited from type_solver_hypre::t_hypre_solver_interface
logical single_setup = .true.
 Do Hypre set-up only at first iteration.
 
logical single_setup_flag = .false.
 Associated flag (set to true after 1 iteration)
 
logical single_setup_cst_matrix_coef = .false.
 If matrix coefficients do not change with time, single set up can be improved.
 
logical single_setup_cst_matrix_coef_flag = .false.
 Associated flag (set to true after 1 iteration)
 
integer(hypre_int) logging = 1
 Amount of logging to do.
 
integer(hypre_int) print_level = 0
 Amount of printing to the screen to do.
 
- Public Attributes inherited from type_solver::t_iterative_solver
integer max_iter = 1000
 Maximum allowed number of iterations.
 
double precision tolerance = 1d-13
 Residual relative tolerence.
 
integer final_iter = -1
 Number of iteration of the last exectution.
 
- Public Attributes inherited from type_solver::t_linear_solver
logical use_jacobi_initial_preconditioner = .true.
 Apply Jacobi preconditioning before calling the linear solver.
 
double precision residual = -1d0
 Residual of the last execution.
 

Member Data Documentation

◆ strong_threshold

real(c_double) type_solver_hypre::t_hypre_precond_parcsr_boomeramg::strong_threshold = 0.25_C_DOUBLE

The default is 0.25. For 2d Laplace operators, 0.25 is a good value, for 3d Laplace operators, 0.5 or 0.6 is a better value. For elasticity problems, a large strength threshold, such as 0.9, is often better.(eh??)


The documentation for this type was generated from the following file: