Amesos2 - Direct Sparse Solver Interfaces  Version of the Day
Public Types | Public Member Functions | Public Attributes | Static Public Attributes | Protected Member Functions | Protected Attributes | Private Types | Private Member Functions | Private Attributes | Friends | List of all members
Amesos2::Basker< Matrix, Vector > Class Template Reference

Amesos2 interface to the Baker package. More...

#include <Amesos2_Basker_decl.hpp>

Inheritance diagram for Amesos2::Basker< Matrix, Vector >:
Inheritance graph
[legend]
Collaboration diagram for Amesos2::Basker< Matrix, Vector >:
Collaboration graph
[legend]

Public Types

typedef Amesos2::Basker
< Matrix, Vector > 
type
 
typedef SolverCore
< Amesos2::Basker, Matrix,
Vector > 
super_type
 
typedef super_type::scalar_type scalar_type
 
typedef
super_type::local_ordinal_type 
local_ordinal_type
 
typedef
super_type::global_ordinal_type 
global_ordinal_type
 
typedef
super_type::global_size_type 
global_size_type
 
typedef super_type::node_type node_type
 
typedef TypeMap
< Amesos2::Basker, scalar_type > 
type_map
 
typedef type_map::type basker_type
 
typedef FunctionMap
< Amesos2::Basker, basker_type > 
function_map
 
typedef Matrix matrix_type
 
typedef MatrixAdapter
< matrix_type > 
matrix_adapter_type
 
typedef ConcreteSolver< Matrix,
Vector > 
solver_type
 
typedef Vector vector_type
 

Public Member Functions

 Basker (Teuchos::RCP< const Matrix > A, Teuchos::RCP< Vector > X, Teuchos::RCP< const Vector > B)
 
bool matrixShapeOK ()
 Returns true if the solver can handle this matrix shape. More...
 
void setA (const Teuchos::RCP< const Matrix > a, EPhase keep_phase=CLEAN)
 Sets the matrix A of this solver. More...
 
void setA (const Matrix *a, EPhase keep_phase=CLEAN)
 Sets the matrix A of this solver. More...
 
void setX (const Teuchos::RCP< Vector > x)
 Sets the LHS vector X.
 
void setX (Vector *x)
 Sets the LHS vector X using a raw pointer.
 
const Teuchos::RCP< Vector > getX ()
 Returns the vector that is the LHS of the linear system.
 
Vector * getXRaw ()
 Returns a raw pointer to the LHS of the linear system.
 
void setB (const Teuchos::RCP< const Vector > b)
 Sets the RHS vector B.
 
void setB (const Vector *b)
 Sets the RHS vector B using a raw pointer.
 
const Teuchos::RCP< const Vector > getB ()
 Returns the vector that is the RHS of the linear system.
 
const Vector * getBRaw ()
 Returns a raw pointer to the RHS of the linear system.
 
std::string description () const
 Returns a short description of this Solver.
 
void describe (Teuchos::FancyOStream &out, const Teuchos::EVerbosityLevel verbLevel=Teuchos::Describable::verbLevel_default) const
 
void printTiming (Teuchos::FancyOStream &out, const Teuchos::EVerbosityLevel verbLevel) const
 Prints timing information about the current solver. More...
 
void getTiming (Teuchos::ParameterList &timingParameterList) const
 Extracts timing information from the current solver. More...
 
std::string name () const
 Return the name of this solver. More...
 
Mathematical functions
super_typepreOrdering ()
 Pre-orders the matrix A for minimal fill-in. More...
 
super_typesymbolicFactorization ()
 Performs symbolic factorization on the matrix A. More...
 
super_typenumericFactorization ()
 Performs numeric factorization on the matrix A. More...
 
void solve ()
 Solves $ A X = B$ (or $ A^T X = B$ ) More...
 
void solve (const Teuchos::Ptr< Vector > X, const Teuchos::Ptr< const Vector > B) const
 Solve $ A X = B$ using the given X and B vectors. More...
 
void solve (Vector *X, const Vector *B) const
 Solve $ A X = B$ using the given X and B vectors. More...
 
Parameter methods
super_typesetParameters (const Teuchos::RCP< Teuchos::ParameterList > &parameterList)
 Set/update internal variables and solver options. More...
 
Teuchos::RCP< const
Teuchos::ParameterList > 
getValidParameters () const
 Return a const parameter list of all of the valid parameters that this->setParameterList(...) will accept. More...
 
void setParameterList (const Teuchos::RCP< Teuchos::ParameterList > &parameterList)
 Set or update internal variables and solver options. More...
 
Teuchos::RCP
< Teuchos::ParameterList > 
getNonconstParameterList ()
 This is a empty stub. More...
 
Teuchos::RCP
< Teuchos::ParameterList > 
unsetParameterList ()
 This is an empty stub. More...
 
Accessor methods
Teuchos::RCP< const
Teuchos::Comm< int > > 
getComm () const
 Returns a pointer to the Teuchos::Comm communicator with this operator.
 
StatusgetStatus () const
 Returns a reference to this solver's internal status object.
 

Public Attributes

decltype(type_map::dtype) typedef basker_dtype
 

Static Public Attributes

static const char * name = "Basker"
 Name of this solver interface.
 

Protected Member Functions

void setNnzLU (size_t nnz)
 Set the number of non-zero values in the $L$ and $U$ factors. More...
 

Protected Attributes

Teuchos::RCP< const
MatrixAdapter< Matrix > > 
matrixA_
 The LHS operator.
 
bool matrix_loaded_
 
Teuchos::RCP< Vector > multiVecX_
 The LHS vector/multi-vector.
 
Teuchos::RCP< const Vector > multiVecB_
 The RHS vector/multi-vector. More...
 
global_size_type globalNumRows_
 Number of global rows in matrixA_.
 
global_size_type globalNumCols_
 Number of global columns in matrixA_.
 
global_size_type globalNumNonZeros_
 Number of global non-zero values in matrixA_.
 
global_size_type rowIndexBase_
 Index base of rowmap of matrixA_.
 
global_size_type columnIndexBase_
 Index base of column map of matrixA_.
 
Status status_
 Holds status information about a solver.
 
Control control_
 Parameters for solving.
 
Timers timers_
 Various timing statistics.
 
int rank_
 The MPI rank of this image.
 
bool root_
 If true, then this is the root processor.
 
int nprocs_
 Number of process images in the matrix communicator.
 

Private Types

typedef
Kokkos::DefaultHostExecutionSpace 
HostSpaceType
 
typedef Kokkos::View
< local_ordinal_type
*, HostSpaceType > 
host_ordinal_type_array
 
typedef Kokkos::View
< basker_type *, HostSpaceType > 
host_value_type_array
 
typedef Kokkos::View
< basker_type
**, Kokkos::LayoutLeft,
HostSpaceType > 
host_solve_array_t
 

Private Member Functions

bool single_proc_optimization () const
 can we optimize size_type and ordinal_type for straight pass through, More...
 
int preOrdering_impl ()
 Performs pre-ordering on the matrix to increase efficiency. More...
 
int symbolicFactorization_impl ()
 
int numericFactorization_impl ()
 Basker specific numeric factorization. More...
 
int solve_impl (const Teuchos::Ptr< MultiVecAdapter< Vector > > X, const Teuchos::Ptr< const MultiVecAdapter< Vector > > B) const
 Basker specific solve. More...
 
bool matrixShapeOK_impl () const
 Determines whether the shape of the matrix is OK for this solver.
 
void setParameters_impl (const Teuchos::RCP< Teuchos::ParameterList > &parameterList)
 
Teuchos::RCP< const
Teuchos::ParameterList > 
getValidParameters_impl () const
 
bool loadA_impl (EPhase current_phase)
 Reads matrix data into internal structures. More...
 

Private Attributes

int num_threads
 
host_value_type_array host_nzvals_view_
 Stores the values of the nonzero entries for CHOLMOD.
 
host_ordinal_type_array host_rows_view_
 Stores the location in Ai_ and Aval_ that starts row j.
 
host_ordinal_type_array host_col_ptr_view_
 Stores the row indices of the nonzero entries.
 
bool is_contiguous_
 
host_solve_array_t xValues_
 Persisting 1D store for X.
 
int ldx_
 
host_solve_array_t bValues_
 Persisting 1D store for B.
 
int ldb_
 
mutable::BaskerClassicNS::BaskerClassic
< local_ordinal_type,
basker_dtype > 
basker
 

Friends

class SolverCore< Amesos2::Basker, Matrix, Vector >
 

Detailed Description

template<class Matrix, class Vector>
class Amesos2::Basker< Matrix, Vector >

Amesos2 interface to the Baker package.

See the summary of Basker parameters supported by this Amesos2 interface.

Member Function Documentation

template<class Matrix , class Vector >
bool Amesos2::Basker< Matrix, Vector >::single_proc_optimization ( ) const
private

can we optimize size_type and ordinal_type for straight pass through,

  • *- * also check that is_contiguous_ flag set to true
template<class Matrix , class Vector >
int Amesos2::Basker< Matrix, Vector >::preOrdering_impl ( )
private

Performs pre-ordering on the matrix to increase efficiency.

Come back to add support to Amesos for preordering

template<class Matrix , class Vector >
int Amesos2::Basker< Matrix, Vector >::numericFactorization_impl ( )
private

Basker specific numeric factorization.

Exceptions
std::runtime_errorBasker is not able to factor the matrix
template<class Matrix , class Vector >
int Amesos2::Basker< Matrix, Vector >::solve_impl ( const Teuchos::Ptr< MultiVecAdapter< Vector > >  X,
const Teuchos::Ptr< const MultiVecAdapter< Vector > >  B 
) const
private

Basker specific solve.

Uses the symbolic and numeric factorizations, along with the RHS vector B to solve the sparse system of equations. The solution is placed in X.

Exceptions
std::runtime_errorBasker is not able to solve the system.
template<class Matrix , class Vector >
Teuchos::RCP< const Teuchos::ParameterList > Amesos2::Basker< Matrix, Vector >::getValidParameters_impl ( ) const
private

Hooked in by Amesos2::SolverCore parent class.

Returns
a const Teuchos::ParameterList of all valid parameters for this solver.
template<class Matrix , class Vector >
bool Amesos2::Basker< Matrix, Vector >::loadA_impl ( EPhase  current_phase)
private

Reads matrix data into internal structures.

Parameters
[in]current_phasean indication of which solution phase this load is being performed for.
Returns
true if the matrix was loaded, false if not
template<template< class, class > class ConcreteSolver, class Matrix , class Vector >
Solver< Matrix, Vector > & Amesos2::SolverCore< ConcreteSolver, Matrix, Vector >::preOrdering ( void  )
virtualinherited

Pre-orders the matrix A for minimal fill-in.

Rearranges the rows and columns of the matrix A to minimize the amount of fill-in of the non-zero entries of the matrix. Pre-ordering may or may not be supported by the underlying solver. If not supported, a call to this method simply does nothing.

Returns
a reference to this .
See Also
symbolicFactorization(), numericFactorization(), and solve()

Implements Amesos2::Solver< Matrix, Vector >.

template<template< class, class > class ConcreteSolver, class Matrix , class Vector >
Solver< Matrix, Vector > & Amesos2::SolverCore< ConcreteSolver, Matrix, Vector >::symbolicFactorization ( void  )
virtualinherited

Performs symbolic factorization on the matrix A.

In addition to performing symbolic factorization on the matrix A, the call to symbolicFactorization() implies that no change will be made to the non-zero structure of the underlying matrix without a subsequent call to symbolicFactorization().

Precondition
Postcondition
Returns
a reference to this .
See Also
preOrdering(), numericFactorization(), and solve()

Implements Amesos2::Solver< Matrix, Vector >.

template<template< class, class > class ConcreteSolver, class Matrix , class Vector >
Solver< Matrix, Vector > & Amesos2::SolverCore< ConcreteSolver, Matrix, Vector >::numericFactorization ( void  )
virtualinherited

Performs numeric factorization on the matrix A.

In addition to performing numeric factorization on the matrix A, the call to numericFactorization() implies that no change will be made to the underlying matrix values without a subsequent call to numericFactorization().

Precondition
  • The non-zero structure of the matrix should not have changed since the last call to symbolicFactorization(). Other changes can have arbitrary consequences.
  • The distribution of the matrix should not have changed since the last call to symbolicFactorization().
Postcondition
Numeric factorization will be performed (or marked to be performed) allowing solve() to be performed correctly despite a potential change in the matrix values (though not in the non-zero structure).
Returns
a reference to this
See Also
preOrdering(), symbolicFactorization(), and solve()

Implements Amesos2::Solver< Matrix, Vector >.

template<template< class, class > class ConcreteSolver, class Matrix , class Vector >
void Amesos2::SolverCore< ConcreteSolver, Matrix, Vector >::solve ( void  )
virtualinherited

Solves $ A X = B$ (or $ A^T X = B$ )

Precondition
Postcondition
X will be set such that $ A X = B$ (or $ A^T X = B$ ), within the limits of the accuracy of the underlying solver.
Returns
void
See Also
preOrdering(), symbolicFactorization(), and numericFactorization()

Implements Amesos2::Solver< Matrix, Vector >.

template<template< class, class > class ConcreteSolver, class Matrix , class Vector>
void Amesos2::SolverCore< ConcreteSolver, Matrix, Vector >::solve ( const Teuchos::Ptr< Vector >  X,
const Teuchos::Ptr< const Vector >  B 
) const
virtualinherited

Solve $ A X = B$ using the given X and B vectors.

This overload of solve uses the given X and B vectors when solving. This X and B are used in place of any X and B that were given upon construction of the Amesos2 solver instance and are used only for this solve.

If a permanent change of X and B are required, see the setX() and setB() methods.

Postcondition
  • The (multi)vector X contains the solution to the system
  • The X and B given at construction time (if any) are unchanged.

Implements Amesos2::Solver< Matrix, Vector >.

References Amesos2::Solver< Matrix, Vector >::numericFactorization().

template<template< class, class > class ConcreteSolver, class Matrix , class Vector>
void Amesos2::SolverCore< ConcreteSolver, Matrix, Vector >::solve ( Vector *  X,
const Vector *  B 
) const
virtualinherited

Solve $ A X = B$ using the given X and B vectors.

This overload of solve uses the given X and B vectors when solving. This X and B are used in place of any X and B that were given upon construction of the Amesos2 solver instance and are used only for this solve.

If a permanent change of X and B are required, see the setX() and setB() methods.

Postcondition
  • The (multi)vector X contains the solution to the system
  • The X and B given at construction time (if any) are unchanged.

Implements Amesos2::Solver< Matrix, Vector >.

template<template< class, class > class ConcreteSolver, class Matrix , class Vector >
bool Amesos2::SolverCore< ConcreteSolver, Matrix, Vector >::matrixShapeOK ( void  )
virtualinherited

Returns true if the solver can handle this matrix shape.

Returns true if the matrix shape is one that the underlying concrete sparse direct solver can handle. Classes that work only on square matrices should return false for rectangular matrices. Classes that work only on symmetric matrices would return false for non-symmetric matrices. etc.

Implements Amesos2::Solver< Matrix, Vector >.

Referenced by Amesos2::SolverCore< ConcreteSolver, Matrix, Vector >::SolverCore().

template<template< class, class > class ConcreteSolver, class Matrix, class Vector >
void Amesos2::SolverCore< ConcreteSolver, Matrix, Vector >::setA ( const Teuchos::RCP< const Matrix >  a,
EPhase  keep_phase = CLEAN 
)
virtualinherited

Sets the matrix A of this solver.

Parameters
[in]aAn RCP to a matrix will will be used for future computation steps
[in]keep_phaseThis parameter tells the solver what state it should keep. For example, you may want to replace the matrix but keep the symbolic factorization because you know the structure of the new matrix is the same as the structure of the old matrix. In this case you would pass Amesos2::SYMBFACT as this parameter.

The default value for the second parameter is Amesos2::CLEAN, which means that the internal state of the solver will be completely reset. It will be as if no previous computational steps were performed.

Implements Amesos2::Solver< Matrix, Vector >.

template<template< class, class > class ConcreteSolver, class Matrix, class Vector>
void Amesos2::SolverCore< ConcreteSolver, Matrix, Vector >::setA ( const Matrix *  a,
EPhase  keep_phase = CLEAN 
)
inlinevirtualinherited

Sets the matrix A of this solver.

Parameters
[in]aAn raw C pointer to a matrix will will be used for future computation steps.
[in]keep_phaseThis parameter tells the solver what state it should keep. For example, you may want to replace the matrix but keep the symbolic factorization because you know the structure of the new matrix is the same as the structure of the old matrix. In this case you would pass Amesos2::SYMBFACT as this parameter.

The default value for the second parameter is Amesos2::CLEAN, which means that the internal state of the solver will be completely reset. It will be as if no previous computational steps were performed.

Implements Amesos2::Solver< Matrix, Vector >.

Referenced by Amesos2::SolverCore< Amesos2::Umfpack, Matrix, Vector >::setA().

template<template< class, class > class ConcreteSolver, class Matrix , class Vector >
Solver< Matrix, Vector > & Amesos2::SolverCore< ConcreteSolver, Matrix, Vector >::setParameters ( const Teuchos::RCP< Teuchos::ParameterList > &  parameterList)
virtualinherited

Set/update internal variables and solver options.

The setParameters method is consistent over all concrete solvers. It accepts general status and control parameters, as well as parameters specific to a given solver. If the solver does not recognize the parameter, then it will simply be ignored

Note
The ParameterList must be named "Amesos2". A list with any other name will be ignored.
Postcondition
  • Internal variables controlling the factorization and solve will be updated and take effect on all subsequent calls to numericFactorization() and solve().
  • All parameters whose value is to differ from the default values must be included in parameterList. Parameters not specified in parameterList revert to their default values.
Returns
a reference to this

Implements Amesos2::Solver< Matrix, Vector >.

Referenced by Amesos2::Lapack< Matrix, Vector >::Lapack(), Amesos2::SolverCore< Amesos2::Umfpack, Matrix, Vector >::setParameterList(), Amesos2::Superludist< Matrix, Vector >::Superludist(), and Amesos2::Superlumt< Matrix, Vector >::Superlumt().

template<template< class, class > class ConcreteSolver, class Matrix , class Vector >
Teuchos::RCP< const Teuchos::ParameterList > Amesos2::SolverCore< ConcreteSolver, Matrix, Vector >::getValidParameters ( void  ) const
virtualinherited

Return a const parameter list of all of the valid parameters that this->setParameterList(...) will accept.

Note
Check the documentation for your concrete solver to see a complete list of the values that each parameter may take. A solver may also recognize multiple data types as arguments for a particular parameters (eg. recognizing "YES" and "NO" as well as true and false ).

Implements Amesos2::Solver< Matrix, Vector >.

Referenced by Amesos2::Lapack< Matrix, Vector >::Lapack(), Amesos2::Superludist< Matrix, Vector >::Superludist(), and Amesos2::Superlumt< Matrix, Vector >::Superlumt().

template<template< class, class > class ConcreteSolver, class Matrix, class Vector>
void Amesos2::SolverCore< ConcreteSolver, Matrix, Vector >::setParameterList ( const Teuchos::RCP< Teuchos::ParameterList > &  parameterList)
inlineinherited

Set or update internal variables and solver options.

Redefined from Teuchos::ParameterListAcceptor

Note
Alias for setParameters()
Parameters
[in]parameterList
template<template< class, class > class ConcreteSolver, class Matrix, class Vector>
Teuchos::RCP<Teuchos::ParameterList> Amesos2::SolverCore< ConcreteSolver, Matrix, Vector >::getNonconstParameterList ( )
inlineinherited

This is a empty stub.

Returns
template<template< class, class > class ConcreteSolver, class Matrix, class Vector>
Teuchos::RCP<Teuchos::ParameterList> Amesos2::SolverCore< ConcreteSolver, Matrix, Vector >::unsetParameterList ( )
inlineinherited

This is an empty stub.

Returns
template<template< class, class > class ConcreteSolver, class Matrix , class Vector >
void Amesos2::SolverCore< ConcreteSolver, Matrix, Vector >::describe ( Teuchos::FancyOStream &  out,
const Teuchos::EVerbosityLevel  verbLevel = Teuchos::Describable::verbLevel_default 
) const
virtualinherited

Prints the status information about the current solver with some level of verbosity

Implements Amesos2::Solver< Matrix, Vector >.

References Amesos2::Util::printLine().

template<template< class, class > class ConcreteSolver, class Matrix , class Vector >
void Amesos2::SolverCore< ConcreteSolver, Matrix, Vector >::printTiming ( Teuchos::FancyOStream &  out,
const Teuchos::EVerbosityLevel  verbLevel 
) const
virtualinherited

Prints timing information about the current solver.

The Amesos2::SolverCore base class takes care of tracking total time spent in the Amesos2 interface. Concrete solver interface class are responsible for reporting other timing statistics, which include time spent in:

  • Redistribution of matrix objects,
  • Conversion of matrix objects to solver-specific formats,
  • Redistribution of multi-vector objects,
  • Conversion of multi-vector objects to solver formats,
  • TPL symbolic factorizations,
  • TPL numeric factorizations, and
  • TPL solves

Implements Amesos2::Solver< Matrix, Vector >.

References Amesos2::Util::printLine().

template<template< class, class > class ConcreteSolver, class Matrix , class Vector >
void Amesos2::SolverCore< ConcreteSolver, Matrix, Vector >::getTiming ( Teuchos::ParameterList &  timingParameterList) const
virtualinherited

Extracts timing information from the current solver.

Results are placed into the parameter list timingParameterList.

Parameters
[out]timingParameterListAccepts timing information from the current solver

Implements Amesos2::Solver< Matrix, Vector >.

template<template< class, class > class ConcreteSolver, class Matrix , class Vector >
std::string Amesos2::SolverCore< ConcreteSolver, Matrix, Vector >::name ( void  ) const
virtualinherited

Return the name of this solver.

The name is given by the underlying concrete solver instance

Returns
A std::string which is the name of this solver

Implements Amesos2::Solver< Matrix, Vector >.

template<template< class, class > class ConcreteSolver, class Matrix, class Vector>
void Amesos2::SolverCore< ConcreteSolver, Matrix, Vector >::setNnzLU ( size_t  nnz)
inlineprotectedinherited

Set the number of non-zero values in the $L$ and $U$ factors.

Concrete solver classes may call this method if they wish to (or are able to) report the number of conbined non-zero count for the $L$ and $U$ factors.

Member Data Documentation

template<template< class, class > class ConcreteSolver, class Matrix, class Vector>
bool Amesos2::SolverCore< ConcreteSolver, Matrix, Vector >::matrix_loaded_
protectedinherited

If true indicates that the current matrix A has been loaded into internal solver structures.

template<template< class, class > class ConcreteSolver, class Matrix, class Vector>
Teuchos::RCP<const Vector> Amesos2::SolverCore< ConcreteSolver, Matrix, Vector >::multiVecB_
protectedinherited

The RHS vector/multi-vector.

We point to a const Vector because Amesos2 should never directly modify B.

Referenced by Amesos2::SolverCore< Amesos2::Umfpack, Matrix, Vector >::getB(), Amesos2::SolverCore< Amesos2::Umfpack, Matrix, Vector >::getBRaw(), and Amesos2::SolverCore< Amesos2::Umfpack, Matrix, Vector >::setB().


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