Classes | Public Types | Public Member Functions | Static Public Attributes | Private Types | Private Attributes | Static Private Attributes | List of all members
blaze::DMatDMatAddExpr< MT1, MT2, SO > Class Template Reference

Expression object for dense matrix-dense matrix additions.The DMatDMatAddExpr class represents the compile time expression for additions between dense matrices with identical storage order. More...

#include <DMatDMatAddExpr.h>

Inherits blaze::MatMatAddExpr< DenseMatrix< DMatDMatAddExpr< MT1, MT2, SO >, SO > >, and blaze::Computation.

Classes

class  ConstIterator
 Iterator over the elements of the dense matrix. More...
 

Public Types

using This = DMatDMatAddExpr< MT1, MT2, SO >
 Type of this DMatDMatAdd instance.
 
using BaseType = DenseMatrix< This, SO >
 Base type of this DMatDMatAdd instance.
 
using ResultType = AddTrait_t< RT1, RT2 >
 Result type for expression template evaluations.
 
using OppositeType = OppositeType_t< ResultType >
 Result type with opposite storage order 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 DMatDMatAddExpr &>
 Data type for composite expression templates.
 
using LeftOperand = If_t< IsExpression_v< MT1 >, const MT1, const MT1 &>
 Composite type of the left-hand side dense matrix expression.
 
using RightOperand = If_t< IsExpression_v< MT2 >, const MT2, const MT2 &>
 Composite type of the right-hand side dense matrix expression.
 
using MatrixType = MT
 Type of the matrix.
 

Public Member Functions

 DMatDMatAddExpr (const MT1 &lhs, const MT2 &rhs) noexcept
 Constructor for the DMatDMatAddExpr class. More...
 
ReturnType operator() (size_t i, size_t j) const
 2D-access to the matrix elements. More...
 
ReturnType at (size_t i, size_t j) const
 Checked access to the matrix elements. More...
 
BLAZE_ALWAYS_INLINE auto load (size_t i, size_t j) const noexcept
 Access to the SIMD elements of the matrix. More...
 
ConstIterator begin (size_t i) const
 Returns an iterator to the first non-zero element of row i. More...
 
ConstIterator end (size_t i) const
 Returns an iterator just past the last non-zero element of row i. More...
 
size_t rows () const noexcept
 Returns the current number of rows of the matrix. More...
 
size_t columns () const noexcept
 Returns the current number of columns of the matrix. More...
 
LeftOperand leftOperand () const noexcept
 Returns the left-hand side dense matrix operand. More...
 
RightOperand rightOperand () const noexcept
 Returns the right-hand side dense matrix 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...
 
bool isAligned () const noexcept
 Returns whether the operands of the expression are properly aligned in memory. More...
 
bool canSMPAssign () const noexcept
 Returns whether the expression can be used in SMP assignments. More...
 
BLAZE_ALWAYS_INLINE constexpr MatrixTypeoperator~ () noexcept
 Conversion operator for non-constant matrices. More...
 
BLAZE_ALWAYS_INLINE constexpr const MatrixTypeoperator~ () const noexcept
 Conversion operator for constant matrices. More...
 

Static Public Attributes

static constexpr bool simdEnabled
 Compilation switch for the expression template evaluation strategy. More...
 
static constexpr bool smpAssignable = ( MT1::smpAssignable && MT2::smpAssignable )
 Compilation switch for the expression template assignment strategy.
 
static constexpr size_t SIMDSIZE = SIMDTrait<ElementType>::size
 The number of elements packed within a single SIMD element.
 
static constexpr bool storageOrder = SO
 Storage order of the matrix.
 

Private Types

using RT1 = ResultType_t< MT1 >
 Result type of the left-hand side dense matrix expression.
 
using RT2 = ResultType_t< MT2 >
 Result type of the right-hand side dense matrix expression.
 
using RN1 = ReturnType_t< MT1 >
 Return type of the left-hand side dense matrix expression.
 
using RN2 = ReturnType_t< MT2 >
 Return type of the right-hand side dense matrix expression.
 
using CT1 = CompositeType_t< MT1 >
 Composite type of the left-hand side dense matrix expression.
 
using CT2 = CompositeType_t< MT2 >
 Composite type of the right-hand side dense matrix expression.
 
using ET1 = ElementType_t< MT1 >
 Element type of the left-hand side dense matrix expression.
 
using ET2 = ElementType_t< MT2 >
 Element type of the right-hand side dense matrix expression.
 
using ExprReturnType = decltype(std::declval< RN1 >()+std::declval< RN2 >())
 Expression return type for the subscript operator.
 

Private Attributes

LeftOperand lhs_
 Left-hand side dense matrix of the addition expression.
 
RightOperand rhs_
 Right-hand side dense matrix of the addition 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
 Compilation switch for the serial evaluation strategy of the addition expression. More...
 

Detailed Description

template<typename MT1, typename MT2, bool SO>
class blaze::DMatDMatAddExpr< MT1, MT2, SO >

Expression object for dense matrix-dense matrix additions.

The DMatDMatAddExpr class represents the compile time expression for additions between dense matrices with identical storage order.

Constructor & Destructor Documentation

◆ DMatDMatAddExpr()

template<typename MT1 , typename MT2 , bool SO>
blaze::DMatDMatAddExpr< MT1, MT2, SO >::DMatDMatAddExpr ( const MT1 &  lhs,
const MT2 &  rhs 
)
inlineexplicitnoexcept

Constructor for the DMatDMatAddExpr class.

Parameters
lhsThe left-hand side operand of the addition expression.
rhsThe right-hand side operand of the addition expression.

Member Function Documentation

◆ at()

template<typename MT1 , typename MT2 , bool SO>
ReturnType blaze::DMatDMatAddExpr< MT1, MT2, SO >::at ( size_t  i,
size_t  j 
) const
inline

Checked access to the matrix elements.

Parameters
iAccess index for the row. The index has to be in the range $[0..M-1]$.
jAccess index for the column. The index has to be in the range $[0..N-1]$.
Returns
The resulting value.
Exceptions
std::out_of_rangeInvalid matrix access index.

◆ begin()

template<typename MT1 , typename MT2 , bool SO>
ConstIterator blaze::DMatDMatAddExpr< MT1, MT2, SO >::begin ( size_t  i) const
inline

Returns an iterator to the first non-zero element of row i.

Parameters
iThe row index.
Returns
Iterator to the first non-zero element of row i.

◆ canAlias()

template<typename MT1 , typename MT2 , bool SO>
template<typename T >
bool blaze::DMatDMatAddExpr< MT1, MT2, SO >::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.

◆ canSMPAssign()

template<typename MT1 , typename MT2 , bool SO>
bool blaze::DMatDMatAddExpr< MT1, MT2, SO >::canSMPAssign ( ) const
inlinenoexcept

Returns whether the expression can be used in SMP assignments.

Returns
true in case the expression can be used in SMP assignments, false if not.

◆ columns()

template<typename MT1 , typename MT2 , bool SO>
size_t blaze::DMatDMatAddExpr< MT1, MT2, SO >::columns ( ) const
inlinenoexcept

Returns the current number of columns of the matrix.

Returns
The number of columns of the matrix.

◆ end()

template<typename MT1 , typename MT2 , bool SO>
ConstIterator blaze::DMatDMatAddExpr< MT1, MT2, SO >::end ( size_t  i) const
inline

Returns an iterator just past the last non-zero element of row i.

Parameters
iThe row index.
Returns
Iterator just past the last non-zero element of row i.

◆ isAliased()

template<typename MT1 , typename MT2 , bool SO>
template<typename T >
bool blaze::DMatDMatAddExpr< MT1, MT2, SO >::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.

◆ isAligned()

template<typename MT1 , typename MT2 , bool SO>
bool blaze::DMatDMatAddExpr< MT1, MT2, SO >::isAligned ( ) const
inlinenoexcept

Returns whether the operands of the expression are properly aligned in memory.

Returns
true in case the operands are aligned, false if not.

◆ leftOperand()

template<typename MT1 , typename MT2 , bool SO>
LeftOperand blaze::DMatDMatAddExpr< MT1, MT2, SO >::leftOperand ( ) const
inlinenoexcept

Returns the left-hand side dense matrix operand.

Returns
The left-hand side dense matrix operand.

◆ load()

template<typename MT1 , typename MT2 , bool SO>
BLAZE_ALWAYS_INLINE auto blaze::DMatDMatAddExpr< MT1, MT2, SO >::load ( size_t  i,
size_t  j 
) const
inlinenoexcept

Access to the SIMD elements of the matrix.

Parameters
iAccess index for the row. The index has to be in the range $[0..M-1]$.
jAccess index for the column. The index has to be in the range $[0..N-1]$.
Returns
Reference to the accessed values.

◆ operator()()

template<typename MT1 , typename MT2 , bool SO>
ReturnType blaze::DMatDMatAddExpr< MT1, MT2, SO >::operator() ( size_t  i,
size_t  j 
) const
inline

2D-access to the matrix elements.

Parameters
iAccess index for the row. The index has to be in the range $[0..M-1]$.
jAccess index for the column. The index has to be in the range $[0..N-1]$.
Returns
The resulting value.

◆ operator~() [1/2]

template<typename MT, bool SO>
BLAZE_ALWAYS_INLINE constexpr MatrixType& blaze::Matrix< MT, SO >::operator~ ( )
inlinenoexceptinherited

Conversion operator for non-constant matrices.

Returns
Reference of the actual type of the matrix.

◆ operator~() [2/2]

template<typename MT, bool SO>
BLAZE_ALWAYS_INLINE constexpr const MatrixType& blaze::Matrix< MT, SO >::operator~ ( ) const
inlinenoexceptinherited

Conversion operator for constant matrices.

Returns
Constant reference of the actual type of the matrix.

◆ rightOperand()

template<typename MT1 , typename MT2 , bool SO>
RightOperand blaze::DMatDMatAddExpr< MT1, MT2, SO >::rightOperand ( ) const
inlinenoexcept

Returns the right-hand side dense matrix operand.

Returns
The right-hand side dense matrix operand.

◆ rows()

template<typename MT1 , typename MT2 , bool SO>
size_t blaze::DMatDMatAddExpr< MT1, MT2, SO >::rows ( ) const
inlinenoexcept

Returns the current number of rows of the matrix.

Returns
The number of rows of the matrix.

Member Data Documentation

◆ returnExpr

template<typename MT1 , typename MT2 , bool SO>
constexpr bool blaze::DMatDMatAddExpr< MT1, MT2, SO >::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 matrix 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.

◆ simdEnabled

template<typename MT1 , typename MT2 , bool SO>
constexpr bool blaze::DMatDMatAddExpr< MT1, MT2, SO >::simdEnabled
static
Initial value:
=
( MT1::simdEnabled && MT2::simdEnabled && HasSIMDAdd_v<ET1,ET2> )

Compilation switch for the expression template evaluation strategy.

◆ useAssign

template<typename MT1 , typename MT2 , bool SO>
constexpr bool blaze::DMatDMatAddExpr< MT1, MT2, SO >::useAssign
staticprivate
Initial value:
=
( RequiresEvaluation_v<MT1> || RequiresEvaluation_v<MT2> || !returnExpr )

Compilation switch for the serial evaluation strategy of the addition expression.

The useAssign compile time constant expression represents a compilation switch for the serial evaluation strategy of the addition expression. In case either of the two dense matrix operands requires an intermediate evaluation or the subscript operator can only return by value, useAssign will be set to 1 and the addition expression will be evaluated via the assign function family. Otherwise useAssign will be set to 0 and the expression will be evaluated via the function call operator.


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