Blaze  3.6
Classes | Public Types | Public Member Functions | Static Public Attributes | Private Types | Private Attributes | Static Private Attributes | List of all members
blaze::SVecDVecMultExpr< VT1, VT2, TF > Class Template Reference

Expression object for sparse vector-dense vector multiplications.The SVecDVecMultExpr class represents the compile time expression for componentwise multiplications between a sparse vector and a dense vector. More...

#include <SVecDVecMultExpr.h>

Inherits blaze::VecVecMultExpr< SparseVector< SVecDVecMultExpr< VT1, VT2, TF >, TF > >, and blaze::Computation.

Classes

class  ConstIterator
 Iterator over the elements of the sparse vector-dense vector multiplication expression. More...
 

Public Types

using This = SVecDVecMultExpr< VT1, VT2, TF >
 Type of this SVecDVecMultExpr instance.
 
using BaseType = SparseVector< This, TF >
 Base type of this SVecDVecMultExpr instance.
 
using ResultType = MultTrait_t< RT1, RT2 >
 Result type for expression template evaluations.
 
using TransposeType = TransposeType_t< ResultType >
 Transpose type for expression template evaluations.
 
using ElementType = ElementType_t< ResultType >
 Resulting element type.
 
using ReturnType = const If_t< returnExpr, ExprReturnType, ElementType >
 Return type for expression template evaluations.
 
using CompositeType = If_t< useAssign, const ResultType, const SVecDVecMultExpr & >
 Data type for composite expression templates.
 
using LeftOperand = If_t< IsExpression_v< VT1 >, const VT1, const VT1 & >
 Composite type of the left-hand side sparse vector expression.
 
using RightOperand = If_t< IsExpression_v< VT2 >, const VT2, const VT2 & >
 Composite type of the right-hand side dense vector expression.
 
using VectorType = SVecDVecMultExpr< VT1, VT2, TF >
 Type of the vector.
 

Public Member Functions

 SVecDVecMultExpr (const VT1 &lhs, const VT2 &rhs) noexcept
 Constructor for the SVecDVecMultExpr class. More...
 
ReturnType operator[] (size_t index) const
 Subscript operator for the direct access to the vector elements. More...
 
ReturnType at (size_t index) const
 Checked access to the vector elements. More...
 
ConstIterator begin () const
 Returns an iterator to the first non-zero element of the sparse vector. More...
 
ConstIterator end () const
 Returns an iterator just past the last non-zero element of the sparse vector. More...
 
size_t size () const noexcept
 Returns the current size/dimension of the vector. More...
 
size_t nonZeros () const
 Returns the number of non-zero elements in the sparse vector. More...
 
ConstIterator find (size_t index) const
 Searches for a specific vector element. More...
 
ConstIterator lowerBound (size_t index) const
 Returns an iterator to the first index not less then the given index. More...
 
ConstIterator upperBound (size_t index) const
 Returns an iterator to the first index greater then the given index. More...
 
LeftOperand leftOperand () const noexcept
 Returns the left-hand side sparse vector operand. More...
 
RightOperand rightOperand () const noexcept
 Returns the right-hand side dense vector operand. More...
 
template<typename T >
bool canAlias (const T *alias) const noexcept
 Returns whether the expression can alias with the given address alias. More...
 
template<typename T >
bool isAliased (const T *alias) const noexcept
 Returns whether the expression is aliased with the given address alias. More...
 
BLAZE_ALWAYS_INLINE constexpr VectorTypeoperator~ () noexcept
 Conversion operator for non-constant vectors. More...
 
BLAZE_ALWAYS_INLINE constexpr const VectorTypeoperator~ () const noexcept
 Conversion operator for constant vectors. More...
 

Static Public Attributes

static constexpr bool smpAssignable = false
 Compilation switch for the expression template assignment strategy.
 
static constexpr bool transposeFlag
 Transpose flag of the vector.
 

Private Types

using RT1 = ResultType_t< VT1 >
 Result type of the left-hand side sparse vector expression.
 
using RT2 = ResultType_t< VT2 >
 Result type of the right-hand side dense vector expression.
 
using RN1 = ReturnType_t< VT1 >
 Return type of the left-hand side sparse vector expression.
 
using RN2 = ReturnType_t< VT2 >
 Return type of the right-hand side dense vector expression.
 
using CT1 = CompositeType_t< VT1 >
 Composite type of the left-hand side sparse vector expression.
 
using CT2 = CompositeType_t< VT2 >
 Composite type of the right-hand side dense vector expression.
 
using ExprReturnType = decltype(std::declval< RN1 >() *std::declval< RN2 >())
 Expression return type for the subscript operator.
 

Private Attributes

LeftOperand lhs_
 Left-hand side sparse vector of the multiplication expression.
 
RightOperand rhs_
 Right-hand side dense vector of the multiplication expression.
 

Static Private Attributes

static constexpr bool returnExpr = ( !IsTemporary_v<RN1> && !IsTemporary_v<RN2> )
 Compilation switch for the selection of the subscript operator return type. More...
 
static constexpr bool useAssign = ( RequiresEvaluation_v<VT1> || RequiresEvaluation_v<VT2> )
 Compilation switch for the evaluation strategy of the multiplication expression. More...
 

Detailed Description

template<typename VT1, typename VT2, bool TF>
class blaze::SVecDVecMultExpr< VT1, VT2, TF >

Expression object for sparse vector-dense vector multiplications.

The SVecDVecMultExpr class represents the compile time expression for componentwise multiplications between a sparse vector and a dense vector.

Constructor & Destructor Documentation

◆ SVecDVecMultExpr()

template<typename VT1 , typename VT2 , bool TF>
blaze::SVecDVecMultExpr< VT1, VT2, TF >::SVecDVecMultExpr ( const VT1 &  lhs,
const VT2 &  rhs 
)
inlineexplicitnoexcept

Constructor for the SVecDVecMultExpr class.

Parameters
lhsThe left-hand side sparse vector operand of the multiplication expression.
rhsThe right-hand side dense vector operand of the multiplication expression.

Member Function Documentation

◆ at()

template<typename VT1 , typename VT2 , bool TF>
ReturnType blaze::SVecDVecMultExpr< VT1, VT2, TF >::at ( size_t  index) const
inline

Checked access to the vector elements.

Parameters
indexAccess index. The index has to be in the range $[0..N-1]$.
Returns
The resulting value.
Exceptions
std::out_of_rangeInvalid vector access index.

◆ begin()

template<typename VT1 , typename VT2 , bool TF>
ConstIterator blaze::SVecDVecMultExpr< VT1, VT2, TF >::begin ( ) const
inline

Returns an iterator to the first non-zero element of the sparse vector.

Returns
Iterator to the first non-zero element of the sparse vector.

◆ canAlias()

template<typename VT1 , typename VT2 , bool TF>
template<typename T >
bool blaze::SVecDVecMultExpr< VT1, VT2, TF >::canAlias ( const T *  alias) const
inlinenoexcept

Returns whether the expression can alias with the given address alias.

Parameters
aliasThe alias to be checked.
Returns
true in case the expression can alias, false otherwise.

◆ end()

template<typename VT1 , typename VT2 , bool TF>
ConstIterator blaze::SVecDVecMultExpr< VT1, VT2, TF >::end ( ) const
inline

Returns an iterator just past the last non-zero element of the sparse vector.

Returns
Iterator just past the last non-zero element of the sparse vector.

◆ find()

template<typename VT1 , typename VT2 , bool TF>
ConstIterator blaze::SVecDVecMultExpr< VT1, VT2, TF >::find ( size_t  index) const
inline

Searches for a specific vector element.

Parameters
indexThe index of the search element.
Returns
Iterator to the element in case the index is found, end() iterator otherwise.

◆ isAliased()

template<typename VT1 , typename VT2 , bool TF>
template<typename T >
bool blaze::SVecDVecMultExpr< VT1, VT2, TF >::isAliased ( const T *  alias) const
inlinenoexcept

Returns whether the expression is aliased with the given address alias.

Parameters
aliasThe alias to be checked.
Returns
true in case an alias effect is detected, false otherwise.

◆ leftOperand()

template<typename VT1 , typename VT2 , bool TF>
LeftOperand blaze::SVecDVecMultExpr< VT1, VT2, TF >::leftOperand ( ) const
inlinenoexcept

Returns the left-hand side sparse vector operand.

Returns
The left-hand side sparse vector operand.

◆ lowerBound()

template<typename VT1 , typename VT2 , bool TF>
ConstIterator blaze::SVecDVecMultExpr< VT1, VT2, TF >::lowerBound ( size_t  index) const
inline

Returns an iterator to the first index not less then the given index.

Parameters
indexThe index of the search element.
Returns
Iterator to the first index not less then the given index, end() iterator otherwise.

◆ nonZeros()

template<typename VT1 , typename VT2 , bool TF>
size_t blaze::SVecDVecMultExpr< VT1, VT2, TF >::nonZeros ( ) const
inline

Returns the number of non-zero elements in the sparse vector.

Returns
The number of non-zero elements in the sparse vector.

◆ operator[]()

template<typename VT1 , typename VT2 , bool TF>
ReturnType blaze::SVecDVecMultExpr< VT1, VT2, TF >::operator[] ( size_t  index) const
inline

Subscript operator for the direct access to the vector elements.

Parameters
indexAccess index. The index has to be in the range $[0..N-1]$.
Returns
The resulting value.

◆ operator~() [1/2]

BLAZE_ALWAYS_INLINE constexpr VectorType& blaze::Vector< SVecDVecMultExpr< VT1, VT2, TF > , TF >::operator~ ( )
inlinenoexceptinherited

Conversion operator for non-constant vectors.

Returns
Reference of the actual type of the vector.

◆ operator~() [2/2]

BLAZE_ALWAYS_INLINE constexpr const VectorType& blaze::Vector< SVecDVecMultExpr< VT1, VT2, TF > , TF >::operator~ ( ) const
inlinenoexceptinherited

Conversion operator for constant vectors.

Returns
Const reference of the actual type of the vector.

◆ rightOperand()

template<typename VT1 , typename VT2 , bool TF>
RightOperand blaze::SVecDVecMultExpr< VT1, VT2, TF >::rightOperand ( ) const
inlinenoexcept

Returns the right-hand side dense vector operand.

Returns
The right-hand side dense vector operand.

◆ size()

template<typename VT1 , typename VT2 , bool TF>
size_t blaze::SVecDVecMultExpr< VT1, VT2, TF >::size ( ) const
inlinenoexcept

Returns the current size/dimension of the vector.

Returns
The size of the vector.

◆ upperBound()

template<typename VT1 , typename VT2 , bool TF>
ConstIterator blaze::SVecDVecMultExpr< VT1, VT2, TF >::upperBound ( size_t  index) const
inline

Returns an iterator to the first index greater then the given index.

Parameters
indexThe index of the search element.
Returns
Iterator to the first index greater then the given index, end() iterator otherwise.

Member Data Documentation

◆ returnExpr

template<typename VT1 , typename VT2 , bool TF>
constexpr bool blaze::SVecDVecMultExpr< VT1, VT2, TF >::returnExpr = ( !IsTemporary_v<RN1> && !IsTemporary_v<RN2> )
staticprivate

Compilation switch for the selection of the subscript operator return type.

The returnExpr compile time constant expression is a compilation switch for the selection of the ReturnType. If either vector operand returns a temporary vector or matrix, returnExpr will be set to false and the subscript operator will return it's result by value. Otherwise returnExpr will be set to true and the subscript operator may return it's result as an expression.

◆ useAssign

template<typename VT1 , typename VT2 , bool TF>
constexpr bool blaze::SVecDVecMultExpr< VT1, VT2, TF >::useAssign = ( RequiresEvaluation_v<VT1> || RequiresEvaluation_v<VT2> )
staticprivate

Compilation switch for the evaluation strategy of the multiplication expression.

The useAssign compile time constant expression represents a compilation switch for the evaluation strategy of the multiplication expression. In case either the sparse or the dense vector operand requires an intermediate evaluation, useAssign will be set to true and the multiplication expression will be evaluated via the assign function family. Otherwise useAssign will be set to false and the expression will be evaluated via the subscript operator.


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