gsMatrixOp< MatrixType > Class Template Reference

Detailed Description

template<class MatrixType>
class gismo::gsMatrixOp< MatrixType >

Simple adapter class to use a matrix (or matrix-like object) as a linear operator. Needed for the iterative method classes.

Inheritance diagram for gsMatrixOp< MatrixType >:

Collaboration diagram for gsMatrixOp< MatrixType >:

Public Types
typedef memory::shared_ptr< gsMatrixOp >	Ptr
	Shared pointer for gsMatrixOp.
typedef memory::unique_ptr< gsMatrixOp >	uPtr
	Unique pointer for gsMatrixOp.

Public Member Functions
void	apply (const gsMatrix< T > &input, gsMatrix< T > &x) const
	apply the operator on the input vector and store the result in x
index_t	cols () const
	Returns the number of columns of the operator.
	gsMatrixOp (const MatrixType &mat)
	Constructor taking a reference.
	gsMatrixOp (MatrixPtr mat)
	Constructor taking a shared pointer.
NestedMatrix	matrix () const
	Returns the matrix.
MatrixPtr	matrixPtr () const
	Returns a shared pinter to the matrix.
index_t	rows () const
	Returns the number of rows of the operator.
virtual void	setOptions (const gsOptionList &)
	Set options based on a gsOptionList object.

Static Public Member Functions
static gsOptionList	defaultOptions ()
	Get the default options as gsOptionList object.
static gsIdentityOp< MatrixType::Scalar >	Identity (const index_t dim)
	Identity operator.
static uPtr	make (const MatrixType &mat)
	Make function returning a smart pointer.
static uPtr	make (MatrixPtr mat)
	Make function returning a smart pointer.

Private Attributes
NestedMatrix	m_expr
	Nested Eigen expression.
const MatrixPtr	m_mat
	Shared pointer to matrix (if needed)

Related Symbols
(Note that these are not member symbols.)
template<class Derived >
gsMatrixOp< Derived >::uPtr	makeMatrixOp (const gsEigen::EigenBase< Derived > &mat)
	This essentially just calls the gsMatrixOp constructor, but the use of a template functions allows us to let the compiler do type inference, so we don't need to type out the matrix type explicitly.
template<class Derived >
gsMatrixOp< Derived >::uPtr	makeMatrixOp (memory::shared_ptr< Derived > mat)
	This essentially just calls the gsMatrixOp constructor, but the use of a template functions allows us to let the compiler do type inference, so we don't need to type out the matrix type explicitly.

Constructor & Destructor Documentation

gsMatrixOp()

template<class MatrixType >

gsMatrixOp ( const MatrixType &  mat )

inline

Constructor taking a reference.

Note

This does not copy the matrix. Make sure that the matrix is not deleted too early (alternatively use constructor by shared pointer)

Member Function Documentation

apply()

template<class MatrixType >

void apply ( const gsMatrix< T > &  input, gsMatrix< T > &  x  ) const

inlinevirtual

apply the operator on the input vector and store the result in x

Parameters

input	Input vector
x	result vector

Implements gsLinearOperator< MatrixType::Scalar >.

make()

template<class MatrixType >

static uPtr make ( const MatrixType &  mat )

inlinestatic

Make function returning a smart pointer.

Note

This does not copy the matrix. Make sure that the matrix is not deleted too early or provide a shared pointer.

Friends And Related Symbol Documentation

makeMatrixOp() [1/2]

template<class Derived >

gsMatrixOp< Derived >::uPtr makeMatrixOp ( const gsEigen::EigenBase< Derived > &  mat )

related

This essentially just calls the gsMatrixOp constructor, but the use of a template functions allows us to let the compiler do type inference, so we don't need to type out the matrix type explicitly.

Examples:

gsMatrix<> M;
M.setRandom(10,10);
gsLinearOperator<>::Ptr op  = makeMatrixOp(M);
gsLinearOperator<>::Ptr opT = makeMatrixOp(M.transpose());
gsLinearOperator<>::Ptr opB = makeMatrixOp(M.block(0,0,5,5) );

Note that

gsLinearOperator<>::Ptr opInv = makeMatrixOp(M.inverse());

will re-compute the inverse of the matrix every time the operator opInv is applied, so this is not advised.

Note

If a matrix is provided, only a reference is stored. Make sure that the matrix is not deleted too early or provide a shared pointer.

makeMatrixOp() [2/2]

template<class Derived >

gsMatrixOp< Derived >::uPtr makeMatrixOp ( memory::shared_ptr< Derived >  mat )

related

This essentially just calls the gsMatrixOp constructor, but the use of a template functions allows us to let the compiler do type inference, so we don't need to type out the matrix type explicitly.

Example:

gsMatrix<>::Ptr M(new gsMatrix<>);
M->setRandom(10,10);
gsLinearOperator<>::Ptr op  = makeMatrixOp(M);

Alternatively:

gsMatrix<> M;
M.setRandom(10,10);
gsLinearOperator<>::Ptr op = makeMatrixOp(M.moveToPtr());