An LinearSolver class using the suite of Hypre solvers to allow. More...

#include <hypre_solver.h>

Inheritance diagram for oomph::HypreSolver:

Public Member Functions
	HypreSolver ()
	Constructor.

	~HypreSolver ()
	Empty destructor.

	HypreSolver (const HypreSolver &)=delete
	Broken copy constructor.

void	operator= (const HypreSolver &)=delete
	Broken assignment operator.

void	disable_resolve ()
	Disable resolve function (overloads the LinearSolver disable_resolve function).

unsigned &	max_iter ()
	Access function to Max_iter.

double &	tolerance ()
	Access function to Tolerance.

unsigned &	hypre_method ()
	Access function to Hypre_method flag – specified via enumeration.

unsigned &	internal_preconditioner ()
	Access function to Internal_preconditioner flag – specified via enumeration.

void	amg_using_simple_smoothing ()
	Function to select use of 'simple' AMG smoothers as controlled by AMG_simple_smoother flag.

unsigned &	amg_simple_smoother ()
	Access function to AMG_simple_smoother flag.

void	amg_using_complex_smoothing ()
	Function to select use of 'complex' AMG smoothers as controlled by AMG_complex_smoother flag.

unsigned &	amg_complex_smoother ()
	Access function to AMG_complex_smoother flag.

unsigned &	amg_print_level ()
	Access function to AMG_print_level.

unsigned &	amg_smoother_iterations ()
	Access function to AMG_smoother_iterations.

unsigned &	amg_coarsening ()
	Access function to AMG_coarsening flag.

unsigned &	amg_max_levels ()
	Access function to AMG_max_levels.

double &	amg_damping ()
	Access function to AMG_damping parameter.

double &	amg_strength ()
	Access function to AMG_strength.

double &	amg_max_row_sum ()
	Access function to AMG_max_row_sum.

double &	amg_truncation ()
	Access function to AMG_truncation.

int &	parasails_symmetry ()
	Access function to ParaSails symmetry flag.

int &	parasails_nlevel ()
	Access function to ParaSails nlevel parameter.

double &	parasails_thresh ()
	Access function to ParaSails thresh parameter.

double &	parasails_filter ()
	Access function to ParaSails filter parameter.

double &	euclid_droptol ()
	Access function to Euclid drop tolerance parameter.

int &	euclid_level ()
	Access function to Euclid level parameter for ILU(k) factorization.

void	enable_euclid_rowScale ()
	Enable euclid rowScaling.

void	disable_euclid_rowScale ()
	Disable euclid row scaling.

void	enable_euclid_using_BJ ()
	Enable use of Block Jacobi as opposed to PILU.

void	disable_euclid_using_BJ ()
	Disable use of Block Jacobi, so PILU will be used.

void	euclid_using_ILUK ()
	Function to switch on ILU(k) factorization for Euclid (default is ILU(k) factorization)

void	euclid_using_ILUT ()
	Function to switch on ILUT factorization for Euclid (default is ILU(k) factorization)

unsigned &	euclid_print_level ()
	Function to set the level of printing from Euclid when the Euclid destructor is called 0: no printing (default) 1: prints summary of runtime settings and timings 2: as 1 plus prints memory usage.

unsigned &	krylov_print_level ()
	Access function to Krylov_print_level.

void	enable_delete_matrix ()
	Hypre copies matrix data from oomph-lib's CRDoubleMatrix or DistributedCRDoubleMatrix into its own data structures, doubling the memory requirements for the matrix. As far as the Hypre solvers are concerned the oomph-lib matrix is no longer required and could be deleted to save memory. However, this will not be the expected behaviour and therefore needs to be requested explicitly by the user by calling this function which changes the flag from false (its default) to true.

void	disable_delete_matrix ()
	Hypre copies matrix data from oomph-lib's CRDoubleMatrix or DistributedCRDoubleMatrix into its own data structures, doubling the memory requirements for the matrix. Calling this function ensures that the matrix is not deleted.

void	solve (Problem *const &problem_pt, DoubleVector &solution)
	Function which uses problem_pt's get_jacobian(...) function to generate a linear system which is then solved. This function deletes any existing internal data and then generates a new Hypre solver.

void	solve (DoubleMatrixBase *const &matrix_pt, const DoubleVector &rhs, DoubleVector &solution)
	Function to solve the linear system defined by matrix_pt and rhs. This function will delete any existing internal data and generate a new Hypre solver. Note: The matrix has to be of type CRDoubleMatrix or Distributed CRDoubleMatrix. Note: Hypre copies matrix data from oomph-lib's CRDoubleMatrix or DistributedCRDoubleMatrix into its own data structures, doubling the memory requirements for the matrix. As far as the Hypre solvers are concerned, the oomph-lib matrix is no longer required and could be deleted to save memory. However, this will not be the expected behaviour and therefore needs to be requested explicitly by the user by calling the enable_delete_matrix() function.

void	resolve (const DoubleVector &rhs, DoubleVector &solution)
	Function to resolve a linear system using the existing solver data, allowing a solve with a new right hand side vector. This function must be used after a call to solve(...) with enable_resolve set to true.

void	clean_up_memory ()
	Function deletes all solver data.

Public Member Functions inherited from oomph::LinearSolver
	LinearSolver ()
	Empty constructor, initialise the member data.

	LinearSolver (const LinearSolver &dummy)=delete
	Broken copy constructor.

void	operator= (const LinearSolver &)=delete
	Broken assignment operator.

virtual	~LinearSolver ()
	Empty virtual destructor.

void	enable_doc_time ()
	Enable documentation of solve times.

void	disable_doc_time ()
	Disable documentation of solve times.

bool	is_doc_time_enabled () const
	Is documentation of solve times enabled?

bool	is_resolve_enabled () const
	Boolean flag indicating if resolves are enabled.

virtual void	enable_resolve ()
	Enable resolve (i.e. store matrix and/or LU decomposition, say) Virtual so it can be overloaded to perform additional tasks.

virtual void	solve (DoubleMatrixBase *const &matrix_pt, const Vector< double > &rhs, Vector< double > &result)
	Linear-algebra-type solver: Takes pointer to a matrix and rhs vector and returns the solution of the linear system.

virtual void	solve_transpose (Problem *const &problem_pt, DoubleVector &result)
	Solver: Takes pointer to problem and returns the results vector which contains the solution of the linear system defined by the problem's fully assembled Jacobian and residual vector (broken virtual).

virtual void	solve_transpose (DoubleMatrixBase *const &matrix_pt, const DoubleVector &rhs, DoubleVector &result)
	Linear-algebra-type solver: Takes pointer to a matrix and rhs vector and returns the solution of the linear system.

virtual void	solve_transpose (DoubleMatrixBase *const &matrix_pt, const Vector< double > &rhs, Vector< double > &result)
	Linear-algebra-type solver: Takes pointer to a matrix and rhs vector and returns the solution of the linear system.

virtual void	resolve_transpose (const DoubleVector &rhs, DoubleVector &result)
	Solver: Resolve the system defined by the last assembled jacobian and the rhs vector. Solution is returned in the vector result. (broken virtual)

virtual double	jacobian_setup_time () const
	returns the time taken to assemble the Jacobian matrix and residual vector (needs to be overloaded for each solver)

virtual double	linear_solver_solution_time () const
	return the time taken to solve the linear system (needs to be overloaded for each linear solver)

virtual void	enable_computation_of_gradient ()
	function to enable the computation of the gradient required for the globally convergent Newton method

void	disable_computation_of_gradient ()
	function to disable the computation of the gradient required for the globally convergent Newton method

void	reset_gradient ()
	function to reset the size of the gradient before each Newton solve

void	get_gradient (DoubleVector &gradient)
	function to access the gradient, provided it has been computed

Public Member Functions inherited from oomph::DistributableLinearAlgebraObject
	DistributableLinearAlgebraObject ()
	Default constructor - create a distribution.

	DistributableLinearAlgebraObject (const DistributableLinearAlgebraObject &matrix)=delete
	Broken copy constructor.

void	operator= (const DistributableLinearAlgebraObject &)=delete
	Broken assignment operator.

virtual	~DistributableLinearAlgebraObject ()
	Destructor.

LinearAlgebraDistribution *	distribution_pt () const
	access to the LinearAlgebraDistribution

unsigned	nrow () const
	access function to the number of global rows.

unsigned	nrow_local () const
	access function for the num of local rows on this processor.

unsigned	nrow_local (const unsigned &p) const
	access function for the num of local rows on this processor.

unsigned	first_row () const
	access function for the first row on this processor

unsigned	first_row (const unsigned &p) const
	access function for the first row on this processor

bool	distributed () const
	distribution is serial or distributed

bool	distribution_built () const
	if the communicator_pt is null then the distribution is not setup then false is returned, otherwise return true

void	build_distribution (const LinearAlgebraDistribution *const dist_pt)
	setup the distribution of this distributable linear algebra object

void	build_distribution (const LinearAlgebraDistribution &dist)
	setup the distribution of this distributable linear algebra object

Public Member Functions inherited from oomph::HypreInterface
	HypreInterface ()
	Constructor.

	~HypreInterface ()
	Destructor.

	HypreInterface (const HypreInterface &)=delete
	Broken copy constructor.

void	operator= (const HypreInterface &)=delete
	Broken assignment operator.

void	enable_hypre_error_messages ()
	Turn on the hypre error messages.

void	disable_hypre_error_messages ()
	Turn off hypre error messages.

unsigned	existing_solver ()
	Function to return value of which solver (if any) is currently stored.

unsigned	existing_preconditioner ()
	Function return value of which preconditioner (if any) is stored.

Private Attributes
bool	Delete_matrix
	Hypre copies matrix data from oomph-lib's CRDoubleMatrix or DistributedCRDoubleMatrix into its own data structures, doubling the memory requirements for the matrix. As far as the Hypre solvers are concerned the oomph-lib matrix is no longer required and could be deleted to save memory. However, this will not be the expected behaviour and therefore needs to be requested explicitly by the user by changing this flag from false (its default) to true.

Additional Inherited Members
Public Types inherited from oomph::HypreInterface
enum	Hypre_method_types { CG , GMRES , BiCGStab , BoomerAMG , Euclid , ParaSails , None }
	Enumerated flag to define which Hypre methods are used CAREFUL: DON'T CHANGE THE ORDER OF THESE! More...

Protected Member Functions inherited from oomph::DistributableLinearAlgebraObject
void	clear_distribution ()
	clear the distribution of this distributable linear algebra object

Protected Member Functions inherited from oomph::HypreInterface
void	hypre_clean_up_memory ()
	Function deletes all solver data.

void	hypre_matrix_setup (CRDoubleMatrix *matrix_pt)
	Function which sets values of First_global_row, Last_global_row and other partitioning data and creates the distributed Hypre matrix (stored in Matrix_ij/Matrix_par) from the CRDoubleMatrix.

void	hypre_solver_setup ()
	Sets up the data required for to use as an oomph-lib LinearSolver or Preconditioner. This must be called after the Hypre matrix has been generated using hypre_matrix_setup(...).

void	hypre_solve (const DoubleVector &rhs, DoubleVector &solution)
	Helper function performs a solve if any solver exists.

void	doc_hypre_parameters ()
	Doc parameters used in hypre solver.

Protected Attributes inherited from oomph::LinearSolver
bool	Enable_resolve
	Boolean that indicates whether the matrix (or its factors, in the case of direct solver) should be stored so that the resolve function can be used.

bool	Doc_time
	Boolean flag that indicates whether the time taken.

bool	Compute_gradient
	flag that indicates whether the gradient required for the globally convergent Newton method should be computed or not

bool	Gradient_has_been_computed
	flag that indicates whether the gradient was computed or not

DoubleVector	Gradient_for_glob_conv_newton_solve
	DoubleVector storing the gradient for the globally convergent Newton method.

Protected Attributes inherited from oomph::HypreInterface
bool	Output_info
	Flag is true to output info and results of timings.

unsigned	Max_iter
	Maximum number of iterations used in solver.

double	Tolerance
	Tolerance used to terminate solver.

unsigned	Hypre_method
	Hypre method flag. Valid values are specified in enumeration.

unsigned	Internal_preconditioner
	Preconditioner method flag used with Hypre's PCG, GMRES and BiCGStab in solve(...) or resolve(...). Valid values are BoomerAMG, Euclid, ParaSails or None (all enumerated above), for any other value no preconditioner is set.

unsigned	AMG_print_level
	Used to set the Hypre printing level for AMG 0: no printout 1: print setup information 2: print solve information 3: print setup and solve information.

unsigned	AMG_max_levels
	Maximum number of levels used in AMG.

double	AMG_max_row_sum
	Parameter to identify diagonally dominant parts of the matrix in AMG.

bool	AMG_using_simple_smoothing
	Flag to determine whether simple smoothers (determined by the AMG_simple_smoother flag) or complex smoothers (determined by the AMG_complex_smoother flag are used in AMG.

unsigned	AMG_simple_smoother
	Simple smoothing methods used in BoomerAMG. To use these methods set AMG_using_simple_smoothing to true. Here are the current options (from hypre's HYPRE_parcsr_ls.h):

unsigned	AMG_complex_smoother
	Complex smoothing methods used in BoomerAMG. To use these methods set AMG_using_simple_smoothing to false Here are the current options (from hypre's HYPRE_parcsr_ls.h):

unsigned	AMG_smoother_iterations
	The number of smoother iterations to apply.

double	AMG_damping
	Damping factor for BoomerAMG smoothed Jacobi or hybrid SOR.

double	AMG_strength
	Connection strength threshold parameter for BoomerAMG.

double	AMG_truncation
	Interpolation truncation factor for BoomerAMG.

unsigned	AMG_coarsening
	AMG coarsening strategy. Coarsening types include: 0 = CLJP (parallel coarsening using independent sets) 1 = classical RS with no boundary treatment (not recommended in parallel) 3 = modified RS with 3rd pass to add C points on the boundaries 6 = Falgout (uses 1 then CLJP using interior coarse points as first independent set) 8 = PMIS (parallel coarsening using independent sets - lower complexities than 0, maybe also slower convergence) 10= HMIS (one pass RS on each processor then PMIS on interior coarse points as first independent set) 11= One pass RS on each processor (not recommended)

int	ParaSails_symmetry
	ParaSails symmetry flag, used to inform ParaSails of Symmetry of definitenss of problem and type of ParaSails preconditioner: 0 = nonsymmetric and/or indefinite problem, nonsymmetric preconditioner 1 = SPD problem, and SPD (factored preconditioner) 2 = nonsymmetric, definite problem and SDP (factored preconditoner)

int	ParaSails_nlevel
	ParaSails nlevel parameter.

double	ParaSails_thresh
	ParaSails thresh parameter.

double	ParaSails_filter
	ParaSails filter parameter.

double	Euclid_droptol
	Euclid drop tolerance for ILU(k) and ILUT factorization.

bool	Euclid_rowScale
	Flag to switch on Euclid row scaling.

bool	Euclid_using_ILUT
	Flag to determine if ILUT (if true) or ILU(k) is used in Euclid.

bool	Euclid_using_BJ
	Flag to determine if Block Jacobi is used instead of PILU.

int	Euclid_level
	Euclid level parameter for ILU(k) factorization.

unsigned	Euclid_print_level
	Flag to set the level of printing from Euclid when the Euclid destructor is called 0: no printing (default) 1: prints summary of runtime settings and timings 2: as 1 plus prints memory usage.

bool	AMGEuclidSmoother_use_block_jacobi

bool	AMGEuclidSmoother_use_row_scaling

bool	AMGEuclidSmoother_use_ilut

unsigned	AMGEuclidSmoother_level

double	AMGEuclidSmoother_drop_tol

unsigned	AMGEuclidSmoother_print_level

unsigned	Krylov_print_level
	Used to set the Hypre printing level for the Krylov subspace solvers.

bool	Hypre_error_messages
	Flag to determine if non-zero values of the Hypre error flag plus Hypre error messages are output to screen at various points in the solve function, i.e. after:

bool	Delete_input_data
	Internal flag which is true when hypre_setup or hypre_solve can delete input matrix.

bool	Using_distributed_rhs
	Internal flag which tell the solver if the rhs Vector is distributed or not.

bool	Returning_distributed_solution
	Internal flag which tell the solver if the solution Vector to be returned is distributed or not.

Detailed Description

An LinearSolver class using the suite of Hypre solvers to allow.

BoomerAMG (AMG), CG, GMRES or BiCGStab

to be used for LinearSolver::solve(...) or LinearSolver::resolve(...).

The Krylov subspace solvers (CG, GMRES and BiCGStab) may be preconditioned using:

    BoomerAMG, Euclid or ParaSails

By default GMRES without preconditioning is used.

Definition at line 702 of file hypre_solver.h.

Constructor & Destructor Documentation

◆ HypreSolver() [1/2]

oomph::HypreSolver::HypreSolver ( )

inline

Constructor.

Definition at line 706 of file hypre_solver.h.

References Delete_matrix, and oomph::LinearSolver::Doc_time.

◆ ~HypreSolver()

oomph::HypreSolver::~HypreSolver ( )

inline

Empty destructor.

Definition at line 723 of file hypre_solver.h.

◆ HypreSolver() [2/2]

oomph::HypreSolver::HypreSolver ( const HypreSolver & )

delete

Broken copy constructor.

Member Function Documentation

◆ amg_coarsening()

unsigned & oomph::HypreSolver::amg_coarsening ( )

inline

Access function to AMG_coarsening flag.

Definition at line 803 of file hypre_solver.h.

References oomph::HypreInterface::AMG_coarsening.

◆ amg_complex_smoother()

unsigned & oomph::HypreSolver::amg_complex_smoother ( )

inline

Access function to AMG_complex_smoother flag.

Definition at line 785 of file hypre_solver.h.

References oomph::HypreInterface::AMG_complex_smoother.

◆ amg_damping()

double & oomph::HypreSolver::amg_damping ( )

inline

Access function to AMG_damping parameter.

Definition at line 815 of file hypre_solver.h.

References oomph::HypreInterface::AMG_damping.

◆ amg_max_levels()

unsigned & oomph::HypreSolver::amg_max_levels ( )

inline

Access function to AMG_max_levels.

Definition at line 809 of file hypre_solver.h.

References oomph::HypreInterface::AMG_max_levels.

◆ amg_max_row_sum()

double & oomph::HypreSolver::amg_max_row_sum ( )

inline

Access function to AMG_max_row_sum.

Definition at line 827 of file hypre_solver.h.

References oomph::HypreInterface::AMG_max_row_sum.

◆ amg_print_level()

unsigned & oomph::HypreSolver::amg_print_level ( )

inline

Access function to AMG_print_level.

Definition at line 791 of file hypre_solver.h.

References oomph::HypreInterface::AMG_print_level.

◆ amg_simple_smoother()

unsigned & oomph::HypreSolver::amg_simple_smoother ( )

inline

Access function to AMG_simple_smoother flag.

Definition at line 772 of file hypre_solver.h.

References oomph::HypreInterface::AMG_simple_smoother.

◆ amg_smoother_iterations()

unsigned & oomph::HypreSolver::amg_smoother_iterations ( )

inline

Access function to AMG_smoother_iterations.

Definition at line 797 of file hypre_solver.h.

References oomph::HypreInterface::AMG_smoother_iterations.

◆ amg_strength()

double & oomph::HypreSolver::amg_strength ( )

inline

Access function to AMG_strength.

Definition at line 821 of file hypre_solver.h.

References oomph::HypreInterface::AMG_strength.

◆ amg_truncation()

double & oomph::HypreSolver::amg_truncation ( )

inline

Access function to AMG_truncation.

Definition at line 833 of file hypre_solver.h.

References oomph::HypreInterface::AMG_truncation.

◆ amg_using_complex_smoothing()

void oomph::HypreSolver::amg_using_complex_smoothing ( )

inline

Function to select use of 'complex' AMG smoothers as controlled by AMG_complex_smoother flag.

Definition at line 779 of file hypre_solver.h.

References oomph::HypreInterface::AMG_using_simple_smoothing.

◆ amg_using_simple_smoothing()

void oomph::HypreSolver::amg_using_simple_smoothing ( )

inline

Function to select use of 'simple' AMG smoothers as controlled by AMG_simple_smoother flag.

Definition at line 766 of file hypre_solver.h.

References oomph::HypreInterface::AMG_using_simple_smoothing.

◆ clean_up_memory()

void oomph::HypreSolver::clean_up_memory ( )

virtual

Function deletes all solver data.

clean_up_memory() deletes any existing solver data

Reimplemented from oomph::LinearSolver.

Definition at line 1649 of file hypre_solver.cc.

References oomph::HypreInterface::hypre_clean_up_memory().

Referenced by disable_resolve(), solve(), and solve().

◆ disable_delete_matrix()

void oomph::HypreSolver::disable_delete_matrix ( )

inline

Hypre copies matrix data from oomph-lib's CRDoubleMatrix or DistributedCRDoubleMatrix into its own data structures, doubling the memory requirements for the matrix. Calling this function ensures that the matrix is not deleted.

Definition at line 949 of file hypre_solver.h.

References Delete_matrix.

◆ disable_euclid_rowScale()

void oomph::HypreSolver::disable_euclid_rowScale ( )

inline

Disable euclid row scaling.

Definition at line 882 of file hypre_solver.h.

References oomph::HypreInterface::Euclid_rowScale.

◆ disable_euclid_using_BJ()

void oomph::HypreSolver::disable_euclid_using_BJ ( )

inline

Disable use of Block Jacobi, so PILU will be used.

Definition at line 896 of file hypre_solver.h.

References oomph::HypreInterface::Euclid_using_BJ.

◆ disable_resolve()

void oomph::HypreSolver::disable_resolve ( )

inlinevirtual

Disable resolve function (overloads the LinearSolver disable_resolve function).

Reimplemented from oomph::LinearSolver.

Definition at line 733 of file hypre_solver.h.

References clean_up_memory(), and oomph::LinearSolver::Enable_resolve.

◆ enable_delete_matrix()

void oomph::HypreSolver::enable_delete_matrix ( )

inline

Hypre copies matrix data from oomph-lib's CRDoubleMatrix or DistributedCRDoubleMatrix into its own data structures, doubling the memory requirements for the matrix. As far as the Hypre solvers are concerned the oomph-lib matrix is no longer required and could be deleted to save memory. However, this will not be the expected behaviour and therefore needs to be requested explicitly by the user by calling this function which changes the flag from false (its default) to true.

Definition at line 940 of file hypre_solver.h.

References Delete_matrix.

◆ enable_euclid_rowScale()

void oomph::HypreSolver::enable_euclid_rowScale ( )

inline

Enable euclid rowScaling.

Definition at line 876 of file hypre_solver.h.

References oomph::HypreInterface::Euclid_rowScale.

◆ enable_euclid_using_BJ()

void oomph::HypreSolver::enable_euclid_using_BJ ( )

inline

Enable use of Block Jacobi as opposed to PILU.

Definition at line 889 of file hypre_solver.h.

References oomph::HypreInterface::Euclid_using_BJ.

◆ euclid_droptol()

double & oomph::HypreSolver::euclid_droptol ( )

inline

Access function to Euclid drop tolerance parameter.

Definition at line 863 of file hypre_solver.h.

References oomph::HypreInterface::Euclid_droptol.

◆ euclid_level()

int & oomph::HypreSolver::euclid_level ( )

inline

Access function to Euclid level parameter for ILU(k) factorization.

Definition at line 870 of file hypre_solver.h.

References oomph::HypreInterface::Euclid_level.

◆ euclid_print_level()

unsigned & oomph::HypreSolver::euclid_print_level ( )

inline

Function to set the level of printing from Euclid when the Euclid destructor is called 0: no printing (default) 1: prints summary of runtime settings and timings 2: as 1 plus prints memory usage.

Definition at line 920 of file hypre_solver.h.

References oomph::HypreInterface::Euclid_print_level.

◆ euclid_using_ILUK()

void oomph::HypreSolver::euclid_using_ILUK ( )

inline

Function to switch on ILU(k) factorization for Euclid (default is ILU(k) factorization)

Definition at line 903 of file hypre_solver.h.

References oomph::HypreInterface::Euclid_using_ILUT.

◆ euclid_using_ILUT()

void oomph::HypreSolver::euclid_using_ILUT ( )

inline

Function to switch on ILUT factorization for Euclid (default is ILU(k) factorization)

Definition at line 910 of file hypre_solver.h.

References oomph::HypreInterface::Euclid_using_ILUT.

◆ hypre_method()

unsigned & oomph::HypreSolver::hypre_method ( )

inline

Access function to Hypre_method flag – specified via enumeration.

Definition at line 752 of file hypre_solver.h.

References oomph::HypreInterface::Hypre_method.

◆ internal_preconditioner()

unsigned & oomph::HypreSolver::internal_preconditioner ( )

inline

Access function to Internal_preconditioner flag – specified via enumeration.

Definition at line 759 of file hypre_solver.h.

References oomph::HypreInterface::Internal_preconditioner.

◆ krylov_print_level()

unsigned & oomph::HypreSolver::krylov_print_level ( )

inline

Access function to Krylov_print_level.

Definition at line 926 of file hypre_solver.h.

References oomph::HypreInterface::Krylov_print_level.

◆ max_iter()

unsigned & oomph::HypreSolver::max_iter ( )

inline

Access function to Max_iter.

Definition at line 740 of file hypre_solver.h.

References oomph::HypreInterface::Max_iter.

◆ operator=()

void oomph::HypreSolver::operator= ( const HypreSolver & )

delete

Broken assignment operator.

◆ parasails_filter()

double & oomph::HypreSolver::parasails_filter ( )

inline

Access function to ParaSails filter parameter.

Definition at line 857 of file hypre_solver.h.

References oomph::HypreInterface::ParaSails_filter.

◆ parasails_nlevel()

int & oomph::HypreSolver::parasails_nlevel ( )

inline

Access function to ParaSails nlevel parameter.

Definition at line 845 of file hypre_solver.h.

References oomph::HypreInterface::ParaSails_nlevel.

◆ parasails_symmetry()

int & oomph::HypreSolver::parasails_symmetry ( )

inline

Access function to ParaSails symmetry flag.

Definition at line 839 of file hypre_solver.h.

References oomph::HypreInterface::ParaSails_symmetry.

◆ parasails_thresh()

double & oomph::HypreSolver::parasails_thresh ( )

inline

Access function to ParaSails thresh parameter.

Definition at line 851 of file hypre_solver.h.

References oomph::HypreInterface::ParaSails_thresh.

◆ resolve()

void oomph::HypreSolver::resolve	(	const DoubleVector &	rhs,
		DoubleVector &	solution
	)

virtual

Function to resolve a linear system using the existing solver data, allowing a solve with a new right hand side vector. This function must be used after a call to solve(...) with enable_resolve set to true.

Resolve performs a linear solve using current solver data (if such data exists).

Reimplemented from oomph::LinearSolver.

Definition at line 1595 of file hypre_solver.cc.

References oomph::DistributableLinearAlgebraObject::distribution_pt(), oomph::LinearSolver::Doc_time, oomph::HypreInterface::existing_solver(), oomph::HypreInterface::hypre_solve(), oomph::HypreInterface::None, and oomph::HypreInterface::Output_info.

◆ solve() [1/2]

void oomph::HypreSolver::solve	(	DoubleMatrixBase *const &	matrix_pt,
		const DoubleVector &	rhs,
		DoubleVector &	solution
	)

virtual

Function to solve the linear system defined by matrix_pt and rhs. This function will delete any existing internal data and generate a new Hypre solver. Note: The matrix has to be of type CRDoubleMatrix or Distributed CRDoubleMatrix. Note: Hypre copies matrix data from oomph-lib's CRDoubleMatrix or DistributedCRDoubleMatrix into its own data structures, doubling the memory requirements for the matrix. As far as the Hypre solvers are concerned, the oomph-lib matrix is no longer required and could be deleted to save memory. However, this will not be the expected behaviour and therefore needs to be requested explicitly by the user by calling the enable_delete_matrix() function.

Uses HypreInterface::hypre_solve(...) to solve the linear system for a CRDoubleMatrix or a DistributedCRDoubleMatrix. In the latter case, the rhs needs to be distributed too, i.e. the length of the rhs vector must be equal to the number of rows stored locally. Note: the returned solution vector is never distributed, i.e. all processors hold all values.

Reimplemented from oomph::LinearSolver.

Definition at line 1499 of file hypre_solver.cc.

References oomph::DistributableLinearAlgebraObject::build_distribution(), clean_up_memory(), oomph::HypreInterface::Delete_input_data, Delete_matrix, oomph::DistributableLinearAlgebraObject::distribution_pt(), oomph::LinearSolver::Doc_time, oomph::LinearSolver::Enable_resolve, oomph::HypreInterface::hypre_matrix_setup(), oomph::HypreInterface::hypre_solve(), oomph::HypreInterface::hypre_solver_setup(), oomph::DoubleMatrixBase::ncol(), oomph::DoubleMatrixBase::nrow(), and oomph::HypreInterface::Output_info.

◆ solve() [2/2]

void oomph::HypreSolver::solve	(	Problem *const &	problem_pt,
		DoubleVector &	solution
	)

virtual

Function which uses problem_pt's get_jacobian(...) function to generate a linear system which is then solved. This function deletes any existing internal data and then generates a new Hypre solver.

Problem-based solve function to generate the Jacobian matrix and residual vector and use HypreInterface::hypre_solver_setup(...) and HypreInterface::hypre_solve(...) to solve the linear system. This function will delete any existing data. Note: the returned solution vector is NOT distributed, i.e. all processors hold all values because this is what the Newton solver requires.

Implements oomph::LinearSolver.

Definition at line 1446 of file hypre_solver.cc.

References clean_up_memory(), oomph::HypreInterface::Delete_input_data, oomph::LinearSolver::Doc_time, oomph::LinearSolver::Enable_resolve, oomph::Problem::get_jacobian(), oomph::HypreInterface::hypre_matrix_setup(), oomph::HypreInterface::hypre_solve(), oomph::HypreInterface::hypre_solver_setup(), oomph::oomph_info, oomph::HypreInterface::Output_info, and oomph::TimingHelpers::timer().

◆ tolerance()

double & oomph::HypreSolver::tolerance ( )

inline

Access function to Tolerance.

Definition at line 746 of file hypre_solver.h.

References oomph::HypreInterface::Tolerance.

Member Data Documentation

◆ Delete_matrix

bool oomph::HypreSolver::Delete_matrix

private

Hypre copies matrix data from oomph-lib's CRDoubleMatrix or DistributedCRDoubleMatrix into its own data structures, doubling the memory requirements for the matrix. As far as the Hypre solvers are concerned the oomph-lib matrix is no longer required and could be deleted to save memory. However, this will not be the expected behaviour and therefore needs to be requested explicitly by the user by changing this flag from false (its default) to true.

Definition at line 995 of file hypre_solver.h.

Referenced by disable_delete_matrix(), enable_delete_matrix(), HypreSolver(), and solve().

The documentation for this class was generated from the following files:

Public Member Functions

Private Attributes

Additional Inherited Members

Detailed Description

Constructor & Destructor Documentation

◆ HypreSolver() [1/2]

◆ ~HypreSolver()

◆ HypreSolver() [2/2]

Member Function Documentation

◆ amg_coarsening()

◆ amg_complex_smoother()

◆ amg_damping()

◆ amg_max_levels()

◆ amg_max_row_sum()

◆ amg_print_level()

◆ amg_simple_smoother()

◆ amg_smoother_iterations()

◆ amg_strength()

◆ amg_truncation()

◆ amg_using_complex_smoothing()

◆ amg_using_simple_smoothing()

◆ clean_up_memory()

◆ disable_delete_matrix()

◆ disable_euclid_rowScale()

◆ disable_euclid_using_BJ()

◆ disable_resolve()

◆ enable_delete_matrix()

◆ enable_euclid_rowScale()

◆ enable_euclid_using_BJ()

◆ euclid_droptol()

◆ euclid_level()

◆ euclid_print_level()

◆ euclid_using_ILUK()

◆ euclid_using_ILUT()

◆ hypre_method()

◆ internal_preconditioner()

◆ krylov_print_level()

◆ max_iter()

◆ operator=()

◆ parasails_filter()

◆ parasails_nlevel()

◆ parasails_symmetry()

◆ parasails_thresh()

◆ resolve()

◆ solve() [1/2]

◆ solve() [2/2]

◆ tolerance()

Member Data Documentation

◆ Delete_matrix