KASKADE 7 development version
Public Member Functions | Friends | List of all members
Kaskade::NestedDissection< StorageScalar > Class Template Reference

a nested dissection factorization for symmetric positive definite sparse matrices More...

#include <domainDecompositionPreconditioner.hh>

Detailed Description

template<class StorageScalar>
class Kaskade::NestedDissection< StorageScalar >

a nested dissection factorization for symmetric positive definite sparse matrices

Template Parameters
StorageScalarthe floating point type used for storage of Cholesky factors

Definition at line 209 of file domainDecompositionPreconditioner.hh.

Public Member Functions

 NestedDissection ()=default
 
template<class SparseMatrix >
 NestedDissection (SparseMatrix const &A, size_t skipEntries, size_t skipDimension)
 Constructor. More...
 
template<class FactorizationScalar >
NestedDissection< StorageScalar > & operator= (NestedDissection< FactorizationScalar > &&nd)
 move assignment from different storage scalar type More...
 
template<class T >
void solve (T &y) const
 
size_t storageSize () const
 
size_t flops () const
 
size_t size ()
 reports the total number of separators in this nested dissection More...
 

Friends

template<class FactorizationScalar >
class NestedDissection
 

Constructor & Destructor Documentation

◆ NestedDissection() [1/2]

template<class StorageScalar >
Kaskade::NestedDissection< StorageScalar >::NestedDissection ( )
default

◆ NestedDissection() [2/2]

template<class StorageScalar >
template<class SparseMatrix >
Kaskade::NestedDissection< StorageScalar >::NestedDissection ( SparseMatrix const &  A,
size_t  skipEntries,
size_t  skipDimension 
)

Constructor.

Member Function Documentation

◆ flops()

template<class StorageScalar >
size_t Kaskade::NestedDissection< StorageScalar >::flops ( ) const

◆ operator=()

template<class StorageScalar >
template<class FactorizationScalar >
NestedDissection< StorageScalar > & Kaskade::NestedDissection< StorageScalar >::operator= ( NestedDissection< FactorizationScalar > &&  nd)

move assignment from different storage scalar type

Use this for compression after construction. E.g., the construction (including factorization) can be done with double precision, and the Cholesky factor be converted to single precision afterwards using this move assignment.

◆ size()

template<class StorageScalar >
size_t Kaskade::NestedDissection< StorageScalar >::size ( )
inline

reports the total number of separators in this nested dissection

Definition at line 239 of file domainDecompositionPreconditioner.hh.

◆ solve()

template<class StorageScalar >
template<class T >
void Kaskade::NestedDissection< StorageScalar >::solve ( T &  y) const

◆ storageSize()

template<class StorageScalar >
size_t Kaskade::NestedDissection< StorageScalar >::storageSize ( ) const

Friends And Related Function Documentation

◆ NestedDissection

template<class StorageScalar >
template<class FactorizationScalar >
friend class NestedDissection
friend

Definition at line 246 of file domainDecompositionPreconditioner.hh.


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