#include <invert_quda.h>

Inheritance diagram for quda::Solver:

Public Member Functions
	Solver (const DiracMatrix &mat, const DiracMatrix &matSloppy, const DiracMatrix &matPrecon, const DiracMatrix &matEig, SolverParam &param, TimeProfile &profile)

virtual	~Solver ()

virtual void	operator() (ColorSpinorField &out, ColorSpinorField &in)=0

virtual void	blocksolve (ColorSpinorField &out, ColorSpinorField &in)

const DiracMatrix &	M ()

const DiracMatrix &	Msloppy ()

const DiracMatrix &	Mprecon ()

const DiracMatrix &	Meig ()

virtual bool	hermitian ()=0

bool	convergence (double r2, double hq2, double r2_tol, double hq_tol)

bool	convergenceHQ (double r2, double hq2, double r2_tol, double hq_tol)
	Test for HQ solver convergence – ignore L2 residual. More...

bool	convergenceL2 (double r2, double hq2, double r2_tol, double hq_tol)
	Test for L2 solver convergence – ignore HQ residual. More...

void	PrintStats (const char *name, int k, double r2, double b2, double hq2)
	Prints out the running statistics of the solver (requires a verbosity of QUDA_VERBOSE) More...

void	PrintSummary (const char *name, int k, double r2, double b2, double r2_tol, double hq_tol)
	Prints out the summary of the solver convergence (requires a verbosity of QUDA_SUMMARIZE). Assumes SolverParam.true_res and SolverParam.true_res_hq has been set. More...

double	precisionEpsilon (QudaPrecision prec=QUDA_INVALID_PRECISION) const
	Returns the epsilon tolerance for a given precision, by default returns the solver precision. More...

void	constructDeflationSpace (const ColorSpinorField &meta, const DiracMatrix &mat)
	Constructs the deflation space and eigensolver. More...

void	destroyDeflationSpace ()
	Destroy the allocated deflation space. More...

void	extendSVDDeflationSpace ()
	Extends the deflation space to twice its size for SVD deflation. More...

void	injectDeflationSpace (std::vector< ColorSpinorField * > &defl_space)
	Injects a deflation space into the solver from the vector argument. Note the input space is reduced to zero size as a result of calling this function, with responsibility for the space transferred to the solver. More...

void	extractDeflationSpace (std::vector< ColorSpinorField * > &defl_space)
	Extracts the deflation space from the solver to the vector argument. Note the solver deflation space is reduced to zero size as a result of calling this function, with responsibility for the space transferred to the argument. More...

int	deflationSpaceSize () const
	Returns the size of deflation space. More...

void	setDeflateCompute (bool flag)
	Sets the deflation compute boolean. More...

void	setRecomputeEvals (bool flag)
	Sets the recompute evals boolean. More...

virtual double	flops () const
	Return flops. More...

Static Public Member Functions
static Solver *	create (SolverParam &param, const DiracMatrix &mat, const DiracMatrix &matSloppy, const DiracMatrix &matPrecon, const DiracMatrix &matEig, TimeProfile &profile)
	Solver factory. More...

static double	stopping (double tol, double b2, QudaResidualType residual_type)
	Set the solver L2 stopping condition. More...

Protected Attributes
const DiracMatrix &	mat

const DiracMatrix &	matSloppy

const DiracMatrix &	matPrecon

const DiracMatrix &	matEig

SolverParam &	param

TimeProfile &	profile

int	node_parity

EigenSolver *	eig_solve

bool	deflate_init

bool	deflate_compute

bool	recompute_evals

std::vector< ColorSpinorField * >	evecs

std::vector< Complex >	evals

Detailed Description

Definition at line 462 of file invert_quda.h.

Constructor & Destructor Documentation

◆ Solver()

quda::Solver::Solver	(	const DiracMatrix &	mat,
		const DiracMatrix &	matSloppy,
		const DiracMatrix &	matPrecon,
		const DiracMatrix &	matEig,
		SolverParam &	param,
		TimeProfile &	profile
	)

Holds the eigenvalues.

Definition at line 14 of file solver.cpp.

◆ ~Solver()

quda::Solver::~Solver ( )

virtual

Definition at line 33 of file solver.cpp.

Member Function Documentation

◆ blocksolve()

void quda::Solver::blocksolve	(	ColorSpinorField &	out,
		ColorSpinorField &	in
	)

virtual

Reimplemented in quda::CG.

Definition at line 302 of file solver.cpp.

◆ constructDeflationSpace()

void quda::Solver::constructDeflationSpace	(	const ColorSpinorField &	meta,
		const DiracMatrix &	mat
	)

Constructs the deflation space and eigensolver.

Parameters

[in]	meta	A sample ColorSpinorField with which to instantiate the eigensolver
[in]	mat	The operator to eigensolve
[in]	Whether	to compute the SVD

Definition at line 168 of file solver.cpp.

◆ convergence()

bool quda::Solver::convergence	(	double	r2,
		double	hq2,
		double	r2_tol,
		double	hq_tol
	)

@briefTest for solver convergence

Parameters

[in]	r2	L2 norm squared of the residual
[in]	hq2	Heavy quark residual
[in]	r2_tol	Solver L2 tolerance
[in]	hq_tol	Solver heavy-quark tolerance

Returns: Whether converged

Definition at line 328 of file solver.cpp.

◆ convergenceHQ()

bool quda::Solver::convergenceHQ	(	double	r2,
		double	hq2,
		double	r2_tol,
		double	hq_tol
	)

Test for HQ solver convergence – ignore L2 residual.

Parameters

[in]	r2	L2 norm squared of the residual
[in]	hq2	Heavy quark residual
[in]	r2_tol	Solver L2 tolerance
	[in[	hq_tol Solver heavy-quark tolerance

Returns: Whether converged

Definition at line 348 of file solver.cpp.

◆ convergenceL2()

bool quda::Solver::convergenceL2	(	double	r2,
		double	hq2,
		double	r2_tol,
		double	hq_tol
	)

Test for L2 solver convergence – ignore HQ residual.

Parameters

[in]	r2	L2 norm squared of the residual
[in]	hq2	Heavy quark residual
[in]	r2_tol	Solver L2 tolerance
[in]	hq_tol	Solver heavy-quark tolerance

Definition at line 361 of file solver.cpp.

◆ create()

Solver * quda::Solver::create	(	SolverParam &	param,
		const DiracMatrix &	mat,
		const DiracMatrix &	matSloppy,
		const DiracMatrix &	matPrecon,
		const DiracMatrix &	matEig,
		TimeProfile &	profile
	)

static

Solver factory.

Definition at line 42 of file solver.cpp.

◆ deflationSpaceSize()

int quda::Solver::deflationSpaceSize ( ) const

inline

Returns the size of deflation space.

Definition at line 615 of file invert_quda.h.

◆ destroyDeflationSpace()

void quda::Solver::destroyDeflationSpace ( )

Destroy the allocated deflation space.

Definition at line 229 of file solver.cpp.

◆ extendSVDDeflationSpace()

void quda::Solver::extendSVDDeflationSpace ( )

Extends the deflation space to twice its size for SVD deflation.

Definition at line 287 of file solver.cpp.

◆ extractDeflationSpace()

void quda::Solver::extractDeflationSpace ( std::vector< ColorSpinorField * > & defl_space )

Extracts the deflation space from the solver to the vector argument. Note the solver deflation space is reduced to zero size as a result of calling this function, with responsibility for the space transferred to the argument.

Parameters

[in,out] defl_space the extracted deflation space. On input, this vector should have zero size.

Definition at line 276 of file solver.cpp.

◆ flops()

virtual double quda::Solver::flops ( ) const

inlinevirtual

Return flops.

Returns: flops expended by this operator

Reimplemented in quda::MG.

Definition at line 633 of file invert_quda.h.

◆ hermitian()

virtual bool quda::Solver::hermitian ( )

pure virtual

Returns: Whether the solver is only for Hermitian systems

Implemented in quda::MG, quda::GMResDR, quda::IncEigCG, quda::PreconditionedSolver, quda::XSD, quda::SD, quda::CAGCR, quda::CACGNR, quda::CACGNE, quda::CACG, quda::MR, quda::GCR, quda::BiCGstabL, quda::MPBiCGstab, quda::SimpleBiCGstab, quda::BiCGstab, quda::PreconCG, quda::MPCG, quda::CG3NR, quda::CG3NE, quda::CG3, quda::CGNR, quda::CGNE, and quda::CG.

◆ injectDeflationSpace()

void quda::Solver::injectDeflationSpace ( std::vector< ColorSpinorField * > & defl_space )

Injects a deflation space into the solver from the vector argument. Note the input space is reduced to zero size as a result of calling this function, with responsibility for the space transferred to the solver.

Parameters

[in,out] defl_space the deflation space we wish to transfer to the solver.

Definition at line 266 of file solver.cpp.

◆ M()

const DiracMatrix& quda::Solver::M ( )

inline

Definition at line 489 of file invert_quda.h.

◆ Meig()

const DiracMatrix& quda::Solver::Meig ( )

inline

Definition at line 492 of file invert_quda.h.

◆ Mprecon()

const DiracMatrix& quda::Solver::Mprecon ( )

inline

Definition at line 491 of file invert_quda.h.

◆ Msloppy()

const DiracMatrix& quda::Solver::Msloppy ( )

inline

Definition at line 490 of file invert_quda.h.

◆ operator()()

virtual void quda::Solver::operator()	(	ColorSpinorField &	out,
		ColorSpinorField &	in
	)

pure virtual

◆ precisionEpsilon()

double quda::Solver::precisionEpsilon ( QudaPrecision prec = QUDA_INVALID_PRECISION ) const

Returns the epsilon tolerance for a given precision, by default returns the solver precision.

Parameters

[in] prec Input precision, default value is solver precision

Definition at line 412 of file solver.cpp.

◆ PrintStats()

void quda::Solver::PrintStats	(	const char *	name,
		int	k,
		double	r2,
		double	b2,
		double	hq2
	)

Prints out the running statistics of the solver (requires a verbosity of QUDA_VERBOSE)

Parameters

[in]	name	Name of solver that called this
[in]	k	iteration count
[in]	r2	L2 norm squared of the residual
[in]	hq2	Heavy quark residual

Definition at line 373 of file solver.cpp.

◆ PrintSummary()

void quda::Solver::PrintSummary	(	const char *	name,
		int	k,
		double	r2,
		double	b2,
		double	r2_tol,
		double	hq_tol
	)

Prints out the summary of the solver convergence (requires a verbosity of QUDA_SUMMARIZE). Assumes SolverParam.true_res and SolverParam.true_res_hq has been set.

Parameters

[in]	name	Name of solver that called this
[in]	k	iteration count
[in]	r2	L2 norm squared of the residual
[in]	hq2	Heavy quark residual
[in]	r2_tol	Solver L2 tolerance
[in]	hq_tol	Solver heavy-quark tolerance

Definition at line 386 of file solver.cpp.

◆ setDeflateCompute()

void quda::Solver::setDeflateCompute ( bool flag )

inline

Sets the deflation compute boolean.

Parameters

[in] flag Set to this boolean value

Definition at line 621 of file invert_quda.h.

◆ setRecomputeEvals()

void quda::Solver::setRecomputeEvals ( bool flag )

inline

Sets the recompute evals boolean.

Parameters

[in] flag Set to this boolean value

Definition at line 627 of file invert_quda.h.

◆ stopping()

double quda::Solver::stopping	(	double	tol,
		double	b2,
		QudaResidualType	residual_type
	)

static

Set the solver L2 stopping condition.

Parameters

[in]	Desired	solver tolerance
[in]	b2	L2 norm squared of the source vector
[in]	residual_type	The type of residual we want to solve for

Returns: L2 stopping condition

Definition at line 311 of file solver.cpp.

Member Data Documentation

◆ deflate_compute

bool quda::Solver::deflate_compute

protected

If true, the deflation space has been computed.

Definition at line 475 of file invert_quda.h.

◆ deflate_init

bool quda::Solver::deflate_init

protected

Eigensolver object.

Definition at line 474 of file invert_quda.h.

◆ eig_solve

EigenSolver* quda::Solver::eig_solve

protected

Definition at line 473 of file invert_quda.h.

◆ evals

std::vector<Complex> quda::Solver::evals

protected

Holds the eigenvectors.

Definition at line 478 of file invert_quda.h.

◆ evecs

std::vector<ColorSpinorField *> quda::Solver::evecs

protected

If true, instruct the solver to recompute evals from an existing deflation space.

Definition at line 477 of file invert_quda.h.

◆ mat

const DiracMatrix& quda::Solver::mat

protected

Definition at line 465 of file invert_quda.h.

◆ matEig

const DiracMatrix& quda::Solver::matEig

protected

Definition at line 468 of file invert_quda.h.

◆ matPrecon

const DiracMatrix& quda::Solver::matPrecon

protected

Definition at line 467 of file invert_quda.h.

◆ matSloppy

const DiracMatrix& quda::Solver::matSloppy

protected

Definition at line 466 of file invert_quda.h.

◆ node_parity

int quda::Solver::node_parity

protected

Definition at line 472 of file invert_quda.h.

◆ param

SolverParam& quda::Solver::param

protected

Definition at line 470 of file invert_quda.h.

◆ profile

TimeProfile& quda::Solver::profile

protected

Definition at line 471 of file invert_quda.h.

◆ recompute_evals

bool quda::Solver::recompute_evals

protected

If true, instruct the solver to create a deflation space.

Definition at line 476 of file invert_quda.h.

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

quda/include/invert_quda.h
quda/lib/solver.cpp

Public Member Functions

Static Public Member Functions

Protected Attributes

Detailed Description

Constructor & Destructor Documentation

◆ Solver()

◆ ~Solver()

Member Function Documentation

◆ blocksolve()

◆ constructDeflationSpace()

◆ convergence()

◆ convergenceHQ()

◆ convergenceL2()

◆ create()

◆ deflationSpaceSize()

◆ destroyDeflationSpace()

◆ extendSVDDeflationSpace()

◆ extractDeflationSpace()

◆ flops()

◆ hermitian()

◆ injectDeflationSpace()

◆ M()

◆ Meig()

◆ Mprecon()

◆ Msloppy()

◆ operator()()

◆ precisionEpsilon()

◆ PrintStats()

◆ PrintSummary()

◆ setDeflateCompute()

◆ setRecomputeEvals()

◆ stopping()

Member Data Documentation

◆ deflate_compute

◆ deflate_init

◆ eig_solve

◆ evals

◆ evecs

◆ mat

◆ matEig

◆ matPrecon

◆ matSloppy

◆ node_parity

◆ param

◆ profile

◆ recompute_evals