SparseCore_Module

Detailed Description

Classes
class	SparseMatrix< _Scalar, _Options, _Index >
	A versatible sparse matrix representation. More...
class	SparseMatrixBase< Derived >
	Base class of any sparse matrices or sparse expressions. More...
class	SparseSelfAdjointView< MatrixType, UpLo >
	Pseudo expression to manipulate a triangular sparse matrix as a selfadjoint matrix. More...
class	SparseVector< _Scalar, _Options, _Index >
	a sparse vector class More...
class	SparseView< MatrixType >
	Expression of a dense or sparse matrix with zero or too small values removed. More...
class	Triplet< Scalar, Index >
	A small structure to hold a non zero as a triplet (i,j,value). More...

Functions
const SparseView< Derived >	sparseView (const Scalar &m_reference=Scalar(0), const typename NumTraits< Scalar >::Real &m_epsilon=NumTraits< Scalar >::dummy_precision()) const

Function Documentation

sparseView()

template<typename Derived>

const SparseView< Derived > sparseView	(	const Scalar &	m_reference = Scalar(0),
		const typename NumTraits< Scalar >::Real &	m_epsilon = NumTraits<Scalar>::dummy_precision() ) const

Returns

a sparse expression of the dense expression *this with values smaller than reference * epsilon removed.

This method is typically used when prototyping to convert a quickly assembled dense Matrix D to a SparseMatrix S:

MatrixXd D(n,m);
SparseMatrix<double> S;
S = D.sparseView();             // suppress numerical zeros (exact)
S = D.sparseView(reference);
S = D.sparseView(reference,epsilon);

where reference is a meaningful non zero reference value, and epsilon is a tolerance factor defaulting to NumTraits<Scalar>::dummy_precision().

See also

SparseMatrixBase::pruned(), class SparseView