Classes | Public Types | Public Member Functions | Private Types | List of all members
blaze::HybridMatrix< Type, M, N, SO > Class Template Reference

Efficient implementation of a dynamically sized matrix with static memory.The HybridMatrix class template combines the flexibility of a dynamically sized matrix with the efficiency and performance of a fixed-size matrix. It is implemented as a crossing between the blaze::StaticMatrix and the blaze::DynamicMatrix class templates: Similar to the static matrix it uses static stack memory instead of dynamically allocated memory and similar to the dynamic matrix it can be resized (within the extend of the static memory). The type of the elements, the maximum number of rows and columns and the storage order of the matrix can be specified via the four template parameters: More...

#include <HybridMatrix.h>

Inherits blaze::DenseMatrix< HybridMatrix< Type, M, N, SO >, SO >.

Classes

struct  Rebind
 Rebind mechanism to obtain a HybridMatrix with different data/element type. More...
 
struct  Resize
 Resize mechanism to obtain a HybridMatrix with different fixed dimensions. More...
 

Public Types

enum  : bool { simdEnabled = IsVectorizable<Type>::value }
 Compilation flag for SIMD optimization. More...
 
enum  : bool { smpAssignable = false }
 Compilation flag for SMP assignments. More...
 
using This = HybridMatrix< Type, M, N, SO >
 Type of this HybridMatrix instance.
 
using BaseType = DenseMatrix< This, SO >
 Base type of this HybridMatrix instance.
 
using ResultType = This
 Result type for expression template evaluations.
 
using OppositeType = HybridMatrix< Type, M, N,!SO >
 Result type with opposite storage order for expression template evaluations.
 
using TransposeType = HybridMatrix< Type, N, M,!SO >
 Transpose type for expression template evaluations.
 
using ElementType = Type
 Type of the matrix elements.
 
using SIMDType = SIMDTrait_< ElementType >
 SIMD type of the matrix elements.
 
using ReturnType = const Type &
 Return type for expression template evaluations.
 
using CompositeType = const This &
 Data type for composite expression templates.
 
using Reference = Type &
 Reference to a non-constant matrix value.
 
using ConstReference = const Type &
 Reference to a constant matrix value.
 
using Pointer = Type *
 Pointer to a non-constant matrix value.
 
using ConstPointer = const Type *
 Pointer to a constant matrix value.
 
using Iterator = DenseIterator< Type, usePadding >
 Iterator over non-constant elements.
 
using ConstIterator = DenseIterator< const Type, usePadding >
 Iterator over constant elements.
 
using MatrixType = MT
 Type of the matrix.
 

Public Member Functions

template<typename Other , size_t Rows, size_t Cols>
HybridMatrix< Type, M, N, SO > & operator= (const Other(&array)[Rows][Cols])
 Array assignment to all matrix elements. More...
 
template<typename MT , bool SO2>
HybridMatrix< Type, M, N, SO > & operator= (const Matrix< MT, SO2 > &rhs)
 Assignment operator for different matrices. More...
 
template<typename MT , bool SO2>
HybridMatrix< Type, M, N, SO > & operator+= (const Matrix< MT, SO2 > &rhs)
 Addition assignment operator for the addition of a matrix ( $ A+=B $). More...
 
template<typename MT , bool SO2>
HybridMatrix< Type, M, N, SO > & operator-= (const Matrix< MT, SO2 > &rhs)
 Subtraction assignment operator for the subtraction of a matrix ( $ A-=B $). More...
 
template<typename MT , bool SO2>
HybridMatrix< Type, M, N, SO > & operator%= (const Matrix< MT, SO2 > &rhs)
 Schur product assignment operator for the multiplication of a matrix ( $ A\circ=B $). More...
 
template<typename Other >
HybridMatrix< Type, M, N, SO > & scale (const Other &scalar)
 Scaling of the matrix by the scalar value scalar ( $ A*=s $). More...
 
template<typename MT , bool SO2>
DisableIf_< typename HybridMatrix< Type, M, N, SO >::BLAZE_TEMPLATE VectorizedAssign< MT > > assign (const DenseMatrix< MT, SO2 > &rhs)
 Default implementation of the assignment of a dense matrix. More...
 
template<typename MT , bool SO2>
EnableIf_< typename HybridMatrix< Type, M, N, SO >::BLAZE_TEMPLATE VectorizedAssign< MT > > assign (const DenseMatrix< MT, SO2 > &rhs)
 SIMD optimized implementation of the assignment of a dense matrix. More...
 
template<typename MT , bool SO2>
DisableIf_< typename HybridMatrix< Type, M, N, SO >::BLAZE_TEMPLATE VectorizedAddAssign< MT > > addAssign (const DenseMatrix< MT, SO2 > &rhs)
 Default implementation of the addition assignment of a row-major dense matrix. More...
 
template<typename MT , bool SO2>
EnableIf_< typename HybridMatrix< Type, M, N, SO >::BLAZE_TEMPLATE VectorizedAddAssign< MT > > addAssign (const DenseMatrix< MT, SO2 > &rhs)
 SIMD optimized implementation of the addition assignment of a dense matrix. More...
 
template<typename MT , bool SO2>
DisableIf_< typename HybridMatrix< Type, M, N, SO >::BLAZE_TEMPLATE VectorizedSubAssign< MT > > subAssign (const DenseMatrix< MT, SO2 > &rhs)
 Default implementation of the subtraction assignment of a dense matrix. More...
 
template<typename MT , bool SO2>
EnableIf_< typename HybridMatrix< Type, M, N, SO >::BLAZE_TEMPLATE VectorizedSubAssign< MT > > subAssign (const DenseMatrix< MT, SO2 > &rhs)
 SIMD optimized implementation of the subtraction assignment of a dense matrix. More...
 
template<typename MT , bool SO2>
DisableIf_< typename HybridMatrix< Type, M, N, SO >::BLAZE_TEMPLATE VectorizedSchurAssign< MT > > schurAssign (const DenseMatrix< MT, SO2 > &rhs)
 Default implementation of the Schur product assignment of a dense matrix. More...
 
template<typename MT , bool SO2>
EnableIf_< typename HybridMatrix< Type, M, N, SO >::BLAZE_TEMPLATE VectorizedSchurAssign< MT > > schurAssign (const DenseMatrix< MT, SO2 > &rhs)
 SIMD optimized implementation of the Schur product assignment of a dense matrix. More...
 
BLAZE_ALWAYS_INLINE MatrixTypeoperator~ () noexcept
 Conversion operator for non-constant matrices. More...
 
BLAZE_ALWAYS_INLINE const MatrixTypeoperator~ () const noexcept
 Conversion operator for constant matrices. More...
 
Constructors
 HybridMatrix ()
 The default constructor for HybridMatrix. More...
 
 HybridMatrix (size_t m, size_t n)
 Constructor for a matrix of size $ m \times n $. No element initialization is performed! More...
 
 HybridMatrix (size_t m, size_t n, const Type &init)
 Constructor for a homogenous initialization of all $ m \times n $ matrix elements. More...
 
 HybridMatrix (initializer_list< initializer_list< Type > > list)
 List initialization of all matrix elements. More...
 
template<typename Other >
 HybridMatrix (size_t m, size_t n, const Other *array)
 Array initialization of all matrix elements. More...
 
template<typename Other , size_t Rows, size_t Cols>
 HybridMatrix (const Other(&array)[Rows][Cols])
 Array initialization of all matrix elements. More...
 
 HybridMatrix (const HybridMatrix &m)
 The copy constructor for HybridMatrix. More...
 
template<typename MT , bool SO2>
 HybridMatrix (const Matrix< MT, SO2 > &m)
 Conversion constructor from different matrices. More...
 
Data access functions
Reference operator() (size_t i, size_t j) noexcept
 2D-access to the matrix elements. More...
 
ConstReference operator() (size_t i, size_t j) const noexcept
 2D-access to the matrix elements. More...
 
Reference at (size_t i, size_t j)
 Checked access to the matrix elements. More...
 
ConstReference at (size_t i, size_t j) const
 Checked access to the matrix elements. More...
 
Pointer data () noexcept
 Low-level data access to the matrix elements. More...
 
ConstPointer data () const noexcept
 Low-level data access to the matrix elements. More...
 
Pointer data (size_t i) noexcept
 Low-level data access to the matrix elements of row/column i. More...
 
ConstPointer data (size_t i) const noexcept
 Low-level data access to the matrix elements of row/column i. More...
 
Iterator begin (size_t i) noexcept
 Returns an iterator to the first element of row/column i. More...
 
ConstIterator begin (size_t i) const noexcept
 Returns an iterator to the first element of row/column i. More...
 
ConstIterator cbegin (size_t i) const noexcept
 Returns an iterator to the first element of row/column i. More...
 
Iterator end (size_t i) noexcept
 Returns an iterator just past the last element of row/column i. More...
 
ConstIterator end (size_t i) const noexcept
 Returns an iterator just past the last element of row/column i. More...
 
ConstIterator cend (size_t i) const noexcept
 Returns an iterator just past the last element of row/column i. More...
 
Assignment operators
HybridMatrixoperator= (const Type &set)
 Homogenous assignment to all matrix elements. More...
 
HybridMatrixoperator= (initializer_list< initializer_list< Type > > list)
 List assignment to all matrix elements. More...
 
template<typename Other , size_t Rows, size_t Cols>
HybridMatrixoperator= (const Other(&array)[Rows][Cols])
 
HybridMatrixoperator= (const HybridMatrix &rhs)
 Copy assignment operator for HybridMatrix. More...
 
template<typename MT , bool SO2>
HybridMatrixoperator= (const Matrix< MT, SO2 > &rhs)
 
template<typename MT , bool SO2>
HybridMatrixoperator+= (const Matrix< MT, SO2 > &rhs)
 
template<typename MT , bool SO2>
HybridMatrixoperator-= (const Matrix< MT, SO2 > &rhs)
 
template<typename MT , bool SO2>
HybridMatrixoperator%= (const Matrix< MT, SO2 > &rhs)
 
Utility functions
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...
 
constexpr size_t spacing () const noexcept
 Returns the spacing between the beginning of two rows. More...
 
constexpr size_t capacity () const noexcept
 Returns the maximum capacity of the matrix. More...
 
size_t capacity (size_t i) const noexcept
 Returns the current capacity of the specified row/column. More...
 
size_t nonZeros () const
 Returns the total number of non-zero elements in the matrix. More...
 
size_t nonZeros (size_t i) const
 Returns the number of non-zero elements in the specified row/column. More...
 
void reset ()
 Reset to the default initial values. More...
 
void reset (size_t i)
 Reset the specified row/column to the default initial values. More...
 
void clear ()
 Clearing the hybrid matrix. More...
 
void resize (size_t m, size_t n, bool preserve=true)
 Changing the size of the matrix. More...
 
void extend (size_t m, size_t n, bool preserve=true)
 Extending the size of the matrix. More...
 
void swap (HybridMatrix &m) noexcept
 Swapping the contents of two hybrid matrices. More...
 
Numeric functions
HybridMatrixtranspose ()
 In-place transpose of the matrix. More...
 
HybridMatrixctranspose ()
 In-place conjugate transpose of the matrix. More...
 
template<typename Other >
HybridMatrixscale (const Other &scalar)
 
Debugging functions
bool isIntact () const noexcept
 Returns whether the invariants of the hybrid matrix are intact. More...
 
Expression template evaluation functions
template<typename Other >
bool canAlias (const Other *alias) const noexcept
 Returns whether the matrix can alias with the given address alias. More...
 
template<typename Other >
bool isAliased (const Other *alias) const noexcept
 Returns whether the matrix is aliased with the given address alias. More...
 
bool isAligned () const noexcept
 Returns whether the matrix is properly aligned in memory. More...
 
BLAZE_ALWAYS_INLINE SIMDType load (size_t i, size_t j) const noexcept
 Load of a SIMD element of the matrix. More...
 
BLAZE_ALWAYS_INLINE SIMDType loada (size_t i, size_t j) const noexcept
 Aligned load of a SIMD element of the matrix. More...
 
BLAZE_ALWAYS_INLINE SIMDType loadu (size_t i, size_t j) const noexcept
 Unaligned load of a SIMD element of the matrix. More...
 
BLAZE_ALWAYS_INLINE void store (size_t i, size_t j, const SIMDType &value) noexcept
 Store of a SIMD element of the matrix. More...
 
BLAZE_ALWAYS_INLINE void storea (size_t i, size_t j, const SIMDType &value) noexcept
 Aligned store of a SIMD element of the matrix. More...
 
BLAZE_ALWAYS_INLINE void storeu (size_t i, size_t j, const SIMDType &value) noexcept
 Unaligned store of a SIMD element of the matrix. More...
 
BLAZE_ALWAYS_INLINE void stream (size_t i, size_t j, const SIMDType &value) noexcept
 Aligned, non-temporal store of a SIMD element of the matrix. More...
 
template<typename MT , bool SO2>
DisableIf_< VectorizedAssign< MT > > assign (const DenseMatrix< MT, SO2 > &rhs)
 
template<typename MT , bool SO2>
EnableIf_< VectorizedAssign< MT > > assign (const DenseMatrix< MT, SO2 > &rhs)
 
template<typename MT >
void assign (const SparseMatrix< MT, SO > &rhs)
 Default implementation of the assignment of a row-major sparse matrix. More...
 
template<typename MT >
void assign (const SparseMatrix< MT,!SO > &rhs)
 Default implementation of the assignment of a column-major sparse matrix. More...
 
template<typename MT , bool SO2>
DisableIf_< VectorizedAddAssign< MT > > addAssign (const DenseMatrix< MT, SO2 > &rhs)
 
template<typename MT , bool SO2>
EnableIf_< VectorizedAddAssign< MT > > addAssign (const DenseMatrix< MT, SO2 > &rhs)
 
template<typename MT >
void addAssign (const SparseMatrix< MT, SO > &rhs)
 Default implementation of the addition assignment of a row-major sparse matrix. More...
 
template<typename MT >
void addAssign (const SparseMatrix< MT,!SO > &rhs)
 Default implementation of the addition assignment of a column-major sparse matrix. More...
 
template<typename MT , bool SO2>
DisableIf_< VectorizedSubAssign< MT > > subAssign (const DenseMatrix< MT, SO2 > &rhs)
 
template<typename MT , bool SO2>
EnableIf_< VectorizedSubAssign< MT > > subAssign (const DenseMatrix< MT, SO2 > &rhs)
 
template<typename MT >
void subAssign (const SparseMatrix< MT, SO > &rhs)
 Default implementation of the subtraction assignment of a row-major sparse matrix. More...
 
template<typename MT >
void subAssign (const SparseMatrix< MT,!SO > &rhs)
 Default implementation of the subtraction assignment of a column-major sparse matrix. More...
 
template<typename MT , bool SO2>
DisableIf_< VectorizedSchurAssign< MT > > schurAssign (const DenseMatrix< MT, SO2 > &rhs)
 
template<typename MT , bool SO2>
EnableIf_< VectorizedSchurAssign< MT > > schurAssign (const DenseMatrix< MT, SO2 > &rhs)
 
template<typename MT >
void schurAssign (const SparseMatrix< MT, SO > &rhs)
 Default implementation of the Schur product assignment of a row-major sparse matrix. More...
 
template<typename MT >
void schurAssign (const SparseMatrix< MT,!SO > &rhs)
 Default implementation of the Schur product assignment of a column-major sparse matrix. More...
 

Static Public Member Functions

Memory functions
static void * operator new (std::size_t size)
 Class specific implementation of operator new. More...
 
static void * operator new[] (std::size_t size)
 Class specific implementation of operator new[]. More...
 
static void * operator new (std::size_t size, const std::nothrow_t &)
 Class specific implementation of the no-throw operator new. More...
 
static void * operator new[] (std::size_t size, const std::nothrow_t &)
 Class specific implementation of the no-throw operator new[]. More...
 
static void operator delete (void *ptr)
 Class specific implementation of operator delete. More...
 
static void operator delete[] (void *ptr)
 Class specific implementation of operator delete[]. More...
 
static void operator delete (void *ptr, const std::nothrow_t &)
 Class specific implementation of no-throw operator delete. More...
 
static void operator delete[] (void *ptr, const std::nothrow_t &)
 Class specific implementation of no-throw operator delete[]. More...
 

Private Types

enum  : size_t { SIMDSIZE = SIMDTrait<ElementType>::size }
 The number of elements packed within a single SIMD element.
 
enum  : size_t { NN = ( usePadding )?( nextMultiple( N, SIMDSIZE ) ):( N ) }
 Alignment adjustment.
 

Private Attributes

Member variables
AlignedArray< Type, M *NN > v_
 The statically allocated matrix elements. More...
 
size_t m_
 The current number of rows of the matrix.
 
size_t n_
 The current number of columns of the matrix.
 

Detailed Description

template<typename Type, size_t M, size_t N, bool SO = defaultStorageOrder>
class blaze::HybridMatrix< Type, M, N, SO >

Efficient implementation of a dynamically sized matrix with static memory.

The HybridMatrix class template combines the flexibility of a dynamically sized matrix with the efficiency and performance of a fixed-size matrix. It is implemented as a crossing between the blaze::StaticMatrix and the blaze::DynamicMatrix class templates: Similar to the static matrix it uses static stack memory instead of dynamically allocated memory and similar to the dynamic matrix it can be resized (within the extend of the static memory). The type of the elements, the maximum number of rows and columns and the storage order of the matrix can be specified via the four template parameters:

template< typename Type, size_t M, size_t N, bool SO >

Depending on the storage order, the matrix elements are either stored in a row-wise fashion or in a column-wise fashion. Given the 2x3 matrix

\[\left(\begin{array}{*{3}{c}} 1 & 2 & 3 \\ 4 & 5 & 6 \\ \end{array}\right)\]


in case of row-major order the elements are stored in the order

\[\left(\begin{array}{*{6}{c}} 1 & 2 & 3 & 4 & 5 & 6. \\ \end{array}\right)\]

In case of column-major order the elements are stored in the order

\[\left(\begin{array}{*{6}{c}} 1 & 4 & 2 & 5 & 3 & 6. \\ \end{array}\right)\]

The use of HybridMatrix is very natural and intuitive. All operations (addition, subtraction, multiplication, scaling, ...) can be performed on all possible combinations of row-major and column-major dense and sparse matrices with fitting element types. The following example gives an impression of the use of HybridMatrix:

HybridMatrix<double,rowMajor> A( 2, 3 ); // Default constructed, non-initialized, row-major 2x3 matrix
A(0,0) = 1.0; A(0,1) = 2.0; A(0,2) = 3.0; // Initialization of the first row
A(1,0) = 4.0; A(1,1) = 5.0; A(1,2) = 6.0; // Initialization of the second row
HybridMatrix<float,columnMajor> B( 2, 3 ); // Default constructed column-major single precision 2x3 matrix
B(0,0) = 1.0; B(0,1) = 3.0; B(0,2) = 5.0; // Initialization of the first row
B(1,0) = 2.0; B(1,1) = 4.0; B(1,2) = 6.0; // Initialization of the second row
CompressedMatrix<float> C( 2, 3 ); // Empty row-major sparse single precision matrix
HybridMatrix<float> D( 3, 2, 4.0F ); // Directly, homogeneously initialized single precision 3x2 matrix
HybridMatrix<double,rowMajor> E( A ); // Creation of a new row-major matrix as a copy of A
HybridMatrix<double,columnMajor> F; // Creation of a default column-major matrix
E = A + B; // Matrix addition and assignment to a row-major matrix
F = A - C; // Matrix subtraction and assignment to a column-major matrix
F = A * D; // Matrix multiplication between two matrices of different element types
A *= 2.0; // In-place scaling of matrix A
E = 2.0 * B; // Scaling of matrix B
F = D * 2.0; // Scaling of matrix D
E += A - B; // Addition assignment
E -= A + C; // Subtraction assignment
F *= A * D; // Multiplication assignment

Member Enumeration Documentation

◆ anonymous enum

template<typename Type, size_t M, size_t N, bool SO = defaultStorageOrder>
anonymous enum : bool

Compilation flag for SIMD optimization.

The simdEnabled compilation flag indicates whether expressions the matrix is involved in can be optimized via SIMD operations. In case the element type of the matrix is a vectorizable data type, the simdEnabled compilation flag is set to true, otherwise it is set to false.

◆ anonymous enum

template<typename Type, size_t M, size_t N, bool SO = defaultStorageOrder>
anonymous enum : bool

Compilation flag for SMP assignments.

The smpAssignable compilation flag indicates whether the matrix can be used in SMP (shared memory parallel) assignments (both on the left-hand and right-hand side of the assignment).

Constructor & Destructor Documentation

◆ HybridMatrix() [1/8]

template<typename Type , size_t M, size_t N, bool SO>
blaze::HybridMatrix< Type, M, N, SO >::HybridMatrix ( )
inlineexplicit

The default constructor for HybridMatrix.

The size of a default constructed HybridMatrix is initially set to 0.

◆ HybridMatrix() [2/8]

template<typename Type , size_t M, size_t N, bool SO>
blaze::HybridMatrix< Type, M, N, SO >::HybridMatrix ( size_t  m,
size_t  n 
)
inlineexplicit

Constructor for a matrix of size $ m \times n $. No element initialization is performed!

Parameters
mThe number of rows of the matrix.
nThe number of columns of the matrix.
Exceptions
std::invalid_argumentInvalid number of rows for hybrid matrix.
std::invalid_argumentInvalid number of columns for hybrid matrix.

This constructor creates a hybrid matrix of size $ m \times n $, but leaves the elements uninitialized. In case m is larger than the maximum allowed number of rows (i.e. m > M) or n is larger than the maximum allowed number of columns a std::invalid_argument exception is thrown.

◆ HybridMatrix() [3/8]

template<typename Type , size_t M, size_t N, bool SO>
blaze::HybridMatrix< Type, M, N, SO >::HybridMatrix ( size_t  m,
size_t  n,
const Type &  init 
)
inlineexplicit

Constructor for a homogenous initialization of all $ m \times n $ matrix elements.

Parameters
mThe number of rows of the matrix.
nThe number of columns of the matrix.
initThe initial value of the matrix elements.
Exceptions
std::invalid_argumentInvalid number of rows for hybrid matrix.
std::invalid_argumentInvalid number of columns for hybrid matrix.

This constructor creates a hybrid matrix of size $ m \times n $ and initializes all matrix elements with the specified value. In case m is larger than the maximum allowed number of rows (i.e. m > M) or n is larger than the maximum allowed number of columns a std::invalid_argument exception is thrown.

◆ HybridMatrix() [4/8]

template<typename Type , size_t M, size_t N, bool SO>
blaze::HybridMatrix< Type, M, N, SO >::HybridMatrix ( initializer_list< initializer_list< Type > >  list)
inlineexplicit

List initialization of all matrix elements.

Parameters
listThe initializer list.
Exceptions
std::invalid_argumentInvalid number of rows for hybrid matrix.
std::invalid_argumentInvalid number of columns for hybrid matrix.

This constructor provides the option to explicitly initialize the elements of the matrix by means of an initializer list:

{ 4, 5 },
{ 7, 8, 9 } };

The matrix is sized according to the size of the initializer list and all its elements are initialized by the values of the given initializer list. Missing values are initialized as default (as e.g. the value 6 in the example). Note that in case the size of the top-level initializer list exceeds the maximum number of rows or the size of any nested list exceeds the maximum number of columns, a std::invalid_argument exception is thrown.

◆ HybridMatrix() [5/8]

template<typename Type , size_t M, size_t N, bool SO>
template<typename Other >
blaze::HybridMatrix< Type, M, N, SO >::HybridMatrix ( size_t  m,
size_t  n,
const Other *  array 
)
inlineexplicit

Array initialization of all matrix elements.

Parameters
mThe number of rows of the matrix.
nThe number of columns of the matrix.
arrayDynamic array for the initialization.
Exceptions
std::invalid_argumentInvalid number of rows for hybrid matrix.
std::invalid_argumentInvalid number of columns for hybrid matrix.

This constructor offers the option to directly initialize the elements of the matrix with a dynamic array:

int* array = new int[20];
// ... Initialization of the dynamic array
delete[] array;

The matrix is sized according to the given size of the array and initialized with the values from the given array. In case m is larger than the maximum allowed number of rows (i.e. m > M) or n is larger than the maximum allowed number of columns a std::invalid_argument exception is thrown. Note that it is expected that the given array has at least m by n elements. Providing an array with less elements results in undefined behavior!

◆ HybridMatrix() [6/8]

template<typename Type , size_t M, size_t N, bool SO>
template<typename Other , size_t Rows, size_t Cols>
blaze::HybridMatrix< Type, M, N, SO >::HybridMatrix ( const Other(&)  array[Rows][Cols])
inlineexplicit

Array initialization of all matrix elements.

Parameters
array$ M \times N $ dimensional array for the initialization.

This constructor offers the option to directly initialize the elements of the matrix with a static array:

const int init[3][3] = { { 1, 2, 3 },
{ 4, 5 },
{ 7, 8, 9 } };

The matrix is sized according to the size of the array and initialized with the values from the given array. Missing values are initialized with default values (as e.g. the value 6 in the example).

◆ HybridMatrix() [7/8]

template<typename Type , size_t M, size_t N, bool SO>
blaze::HybridMatrix< Type, M, N, SO >::HybridMatrix ( const HybridMatrix< Type, M, N, SO > &  m)
inline

The copy constructor for HybridMatrix.

Parameters
mMatrix to be copied.

The copy constructor is explicitly defined due to the required dynamic memory management and in order to enable/facilitate NRV optimization.

◆ HybridMatrix() [8/8]

template<typename Type , size_t M, size_t N, bool SO>
template<typename MT , bool SO2>
blaze::HybridMatrix< Type, M, N, SO >::HybridMatrix ( const Matrix< MT, SO2 > &  m)
inline

Conversion constructor from different matrices.

Parameters
mMatrix to be copied.
Exceptions
std::invalid_argumentInvalid setup of hybrid matrix.

Member Function Documentation

◆ addAssign() [1/4]

template<typename Type , size_t M, size_t N, bool SO>
template<typename MT >
void blaze::HybridMatrix< Type, M, N, SO >::addAssign ( const SparseMatrix< MT, SO > &  rhs)
inline

Default implementation of the addition assignment of a row-major sparse matrix.

Parameters
rhsThe right-hand side sparse matrix to be added.
Returns
void

This function must NOT be called explicitly! It is used internally for the performance optimized evaluation of expression templates. Calling this function explicitly might result in erroneous results and/or in compilation errors. Instead of using this function use the assignment operator.

◆ addAssign() [2/4]

template<typename Type , size_t M, size_t N, bool SO>
template<typename MT >
void blaze::HybridMatrix< Type, M, N, SO >::addAssign ( const SparseMatrix< MT,!SO > &  rhs)
inline

Default implementation of the addition assignment of a column-major sparse matrix.

Parameters
rhsThe right-hand side sparse matrix to be added.
Returns
void

This function must NOT be called explicitly! It is used internally for the performance optimized evaluation of expression templates. Calling this function explicitly might result in erroneous results and/or in compilation errors. Instead of using this function use the assignment operator.

◆ addAssign() [3/4]

template<typename Type, size_t M, size_t N, bool SO = defaultStorageOrder>
template<typename MT , bool SO2>
DisableIf_<typename HybridMatrix<Type,M,N,SO>::BLAZE_TEMPLATE VectorizedAddAssign<MT> > blaze::HybridMatrix< Type, M, N, SO >::addAssign ( const DenseMatrix< MT, SO2 > &  rhs)
inline

Default implementation of the addition assignment of a row-major dense matrix.

Parameters
rhsThe right-hand side dense matrix to be added.
Returns
void

This function must NOT be called explicitly! It is used internally for the performance optimized evaluation of expression templates. Calling this function explicitly might result in erroneous results and/or in compilation errors. Instead of using this function use the assignment operator.

◆ addAssign() [4/4]

template<typename Type, size_t M, size_t N, bool SO = defaultStorageOrder>
template<typename MT , bool SO2>
EnableIf_<typename HybridMatrix<Type,M,N,SO>::BLAZE_TEMPLATE VectorizedAddAssign<MT> > blaze::HybridMatrix< Type, M, N, SO >::addAssign ( const DenseMatrix< MT, SO2 > &  rhs)
inline

SIMD optimized implementation of the addition assignment of a dense matrix.

Parameters
rhsThe right-hand side dense matrix to be added.
Returns
void

This function must NOT be called explicitly! It is used internally for the performance optimized evaluation of expression templates. Calling this function explicitly might result in erroneous results and/or in compilation errors. Instead of using this function use the assignment operator.

◆ assign() [1/4]

template<typename Type , size_t M, size_t N, bool SO>
template<typename MT >
void blaze::HybridMatrix< Type, M, N, SO >::assign ( const SparseMatrix< MT, SO > &  rhs)
inline

Default implementation of the assignment of a row-major sparse matrix.

Parameters
rhsThe right-hand side sparse matrix to be assigned.
Returns
void

This function must NOT be called explicitly! It is used internally for the performance optimized evaluation of expression templates. Calling this function explicitly might result in erroneous results and/or in compilation errors. Instead of using this function use the assignment operator.

◆ assign() [2/4]

template<typename Type , size_t M, size_t N, bool SO>
template<typename MT >
void blaze::HybridMatrix< Type, M, N, SO >::assign ( const SparseMatrix< MT,!SO > &  rhs)
inline

Default implementation of the assignment of a column-major sparse matrix.

Parameters
rhsThe right-hand side sparse matrix to be assigned.
Returns
void

This function must NOT be called explicitly! It is used internally for the performance optimized evaluation of expression templates. Calling this function explicitly might result in erroneous results and/or in compilation errors. Instead of using this function use the assignment operator.

◆ assign() [3/4]

template<typename Type, size_t M, size_t N, bool SO = defaultStorageOrder>
template<typename MT , bool SO2>
DisableIf_<typename HybridMatrix<Type,M,N,SO>::BLAZE_TEMPLATE VectorizedAssign<MT> > blaze::HybridMatrix< Type, M, N, SO >::assign ( const DenseMatrix< MT, SO2 > &  rhs)
inline

Default implementation of the assignment of a dense matrix.

Parameters
rhsThe right-hand side dense matrix to be assigned.
Returns
void

This function must NOT be called explicitly! It is used internally for the performance optimized evaluation of expression templates. Calling this function explicitly might result in erroneous results and/or in compilation errors. Instead of using this function use the assignment operator.

◆ assign() [4/4]

template<typename Type, size_t M, size_t N, bool SO = defaultStorageOrder>
template<typename MT , bool SO2>
EnableIf_<typename HybridMatrix<Type,M,N,SO>::BLAZE_TEMPLATE VectorizedAssign<MT> > blaze::HybridMatrix< Type, M, N, SO >::assign ( const DenseMatrix< MT, SO2 > &  rhs)
inline

SIMD optimized implementation of the assignment of a dense matrix.

Parameters
rhsThe right-hand side dense matrix to be assigned.
Returns
void

This function must NOT be called explicitly! It is used internally for the performance optimized evaluation of expression templates. Calling this function explicitly might result in erroneous results and/or in compilation errors. Instead of using this function use the assignment operator.

◆ at() [1/2]

template<typename Type , size_t M, size_t N, bool SO>
HybridMatrix< Type, M, N, SO >::Reference blaze::HybridMatrix< Type, M, N, SO >::at ( size_t  i,
size_t  j 
)
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
Reference to the accessed value.
Exceptions
std::out_of_rangeInvalid matrix access index.

In contrast to the subscript operator this function always performs a check of the given access indices.

◆ at() [2/2]

template<typename Type , size_t M, size_t N, bool SO>
HybridMatrix< Type, M, N, SO >::ConstReference blaze::HybridMatrix< Type, M, N, 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
Reference to the accessed value.
Exceptions
std::out_of_rangeInvalid matrix access index.

In contrast to the subscript operator this function always performs a check of the given access indices.

◆ begin() [1/2]

template<typename Type , size_t M, size_t N, bool SO>
HybridMatrix< Type, M, N, SO >::Iterator blaze::HybridMatrix< Type, M, N, SO >::begin ( size_t  i)
inlinenoexcept

Returns an iterator to the first element of row/column i.

Parameters
iThe row/column index.
Returns
Iterator to the first element of row/column i.

This function returns a row/column iterator to the first element of row/column i. In case the storage order is set to rowMajor the function returns an iterator to the first element of row i, in case the storage flag is set to columnMajor the function returns an iterator to the first element of column i.

◆ begin() [2/2]

template<typename Type , size_t M, size_t N, bool SO>
HybridMatrix< Type, M, N, SO >::ConstIterator blaze::HybridMatrix< Type, M, N, SO >::begin ( size_t  i) const
inlinenoexcept

Returns an iterator to the first element of row/column i.

Parameters
iThe row/column index.
Returns
Iterator to the first element of row/column i.

This function returns a row/column iterator to the first element of row/column i. In case the storage order is set to rowMajor the function returns an iterator to the first element of row i, in case the storage flag is set to columnMajor the function returns an iterator to the first element of column i.

◆ canAlias()

template<typename Type , size_t M, size_t N, bool SO>
template<typename Other >
bool blaze::HybridMatrix< Type, M, N, SO >::canAlias ( const Other *  alias) const
inlinenoexcept

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

Parameters
aliasThe alias to be checked.
Returns
true in case the alias corresponds to this matrix, false if not.

This function returns whether the given address can alias with the matrix. In contrast to the isAliased() function this function is allowed to use compile time expressions to optimize the evaluation.

◆ capacity() [1/2]

template<typename Type , size_t M, size_t N, bool SO>
constexpr size_t blaze::HybridMatrix< Type, M, N, SO >::capacity ( ) const
inlinenoexcept

Returns the maximum capacity of the matrix.

Returns
The capacity of the matrix.

◆ capacity() [2/2]

template<typename Type , size_t M, size_t N, bool SO>
size_t blaze::HybridMatrix< Type, M, N, SO >::capacity ( size_t  i) const
inlinenoexcept

Returns the current capacity of the specified row/column.

Parameters
iThe index of the row/column.
Returns
The current capacity of row/column i.

This function returns the current capacity of the specified row/column. In case the storage order is set to rowMajor the function returns the capacity of row i, in case the storage flag is set to columnMajor the function returns the capacity of column i.

◆ cbegin()

template<typename Type , size_t M, size_t N, bool SO>
HybridMatrix< Type, M, N, SO >::ConstIterator blaze::HybridMatrix< Type, M, N, SO >::cbegin ( size_t  i) const
inlinenoexcept

Returns an iterator to the first element of row/column i.

Parameters
iThe row/column index.
Returns
Iterator to the first element of row/column i.

This function returns a row/column iterator to the first element of row/column i. In case the storage order is set to rowMajor the function returns an iterator to the first element of row i, in case the storage flag is set to columnMajor the function returns an iterator to the first element of column i.

◆ cend()

template<typename Type , size_t M, size_t N, bool SO>
HybridMatrix< Type, M, N, SO >::ConstIterator blaze::HybridMatrix< Type, M, N, SO >::cend ( size_t  i) const
inlinenoexcept

Returns an iterator just past the last element of row/column i.

Parameters
iThe row/column index.
Returns
Iterator just past the last element of row/column i.

This function returns an row/column iterator just past the last element of row/column i. In case the storage order is set to rowMajor the function returns an iterator just past the last element of row i, in case the storage flag is set to columnMajor the function returns an iterator just past the last element of column i.

◆ clear()

template<typename Type , size_t M, size_t N, bool SO>
void blaze::HybridMatrix< Type, M, N, SO >::clear ( )
inline

Clearing the hybrid matrix.

Returns
void

After the clear() function, the size of the matrix is 0.

◆ columns()

template<typename Type , size_t M, size_t N, bool SO>
size_t blaze::HybridMatrix< Type, M, N, SO >::columns ( ) const
inlinenoexcept

Returns the current number of columns of the matrix.

Returns
The number of columns of the matrix.

◆ ctranspose()

template<typename Type , size_t M, size_t N, bool SO>
HybridMatrix< Type, M, N, SO > & blaze::HybridMatrix< Type, M, N, SO >::ctranspose ( )
inline

In-place conjugate transpose of the matrix.

Returns
Reference to the transposed matrix.
Exceptions
std::logic_errorImpossible transpose operation.

This function transposes the hybrid matrix in-place. Note that this function can only be used on hybrid matrices whose current dimensions allow an in-place transpose operation. In case the current number of rows is larger than the maximum number of columns or if the current number of columns is larger than the maximum number of rows, an std::logic_error is thrown.

◆ data() [1/4]

template<typename Type , size_t M, size_t N, bool SO>
HybridMatrix< Type, M, N, SO >::Pointer blaze::HybridMatrix< Type, M, N, SO >::data ( )
inlinenoexcept

Low-level data access to the matrix elements.

Returns
Pointer to the internal element storage.

This function returns a pointer to the internal storage of the hybrid matrix. Note that you can NOT assume that all matrix elements lie adjacent to each other! The hybrid matrix may use techniques such as padding to improve the alignment of the data. Whereas the number of elements within a row/column are given by the rows() and columns() member functions, respectively, the total number of elements including padding is given by the spacing() member function.

◆ data() [2/4]

template<typename Type , size_t M, size_t N, bool SO>
HybridMatrix< Type, M, N, SO >::ConstPointer blaze::HybridMatrix< Type, M, N, SO >::data ( ) const
inlinenoexcept

Low-level data access to the matrix elements.

Returns
Pointer to the internal element storage.

This function returns a pointer to the internal storage of the hybrid matrix. Note that you can NOT assume that all matrix elements lie adjacent to each other! The hybrid matrix may use techniques such as padding to improve the alignment of the data. Whereas the number of elements within a row/column are given by the rows() and columns() member functions, respectively, the total number of elements including padding is given by the spacing() member function.

◆ data() [3/4]

template<typename Type , size_t M, size_t N, bool SO>
HybridMatrix< Type, M, N, SO >::Pointer blaze::HybridMatrix< Type, M, N, SO >::data ( size_t  i)
inlinenoexcept

Low-level data access to the matrix elements of row/column i.

Parameters
iThe row/column index.
Returns
Pointer to the internal element storage.

This function returns a pointer to the internal storage for the elements in row/column i.

◆ data() [4/4]

template<typename Type , size_t M, size_t N, bool SO>
HybridMatrix< Type, M, N, SO >::ConstPointer blaze::HybridMatrix< Type, M, N, SO >::data ( size_t  i) const
inlinenoexcept

Low-level data access to the matrix elements of row/column i.

Parameters
iThe row/column index.
Returns
Pointer to the internal element storage.

This function returns a pointer to the internal storage for the elements in row/column i.

◆ end() [1/2]

template<typename Type , size_t M, size_t N, bool SO>
HybridMatrix< Type, M, N, SO >::Iterator blaze::HybridMatrix< Type, M, N, SO >::end ( size_t  i)
inlinenoexcept

Returns an iterator just past the last element of row/column i.

Parameters
iThe row/column index.
Returns
Iterator just past the last element of row/column i.

This function returns an row/column iterator just past the last element of row/column i. In case the storage order is set to rowMajor the function returns an iterator just past the last element of row i, in case the storage flag is set to columnMajor the function returns an iterator just past the last element of column i.

◆ end() [2/2]

template<typename Type , size_t M, size_t N, bool SO>
HybridMatrix< Type, M, N, SO >::ConstIterator blaze::HybridMatrix< Type, M, N, SO >::end ( size_t  i) const
inlinenoexcept

Returns an iterator just past the last element of row/column i.

Parameters
iThe row/column index.
Returns
Iterator just past the last element of row/column i.

This function returns an row/column iterator just past the last element of row/column i. In case the storage order is set to rowMajor the function returns an iterator just past the last element of row i, in case the storage flag is set to columnMajor the function returns an iterator just past the last element of column i.

◆ extend()

template<typename Type , size_t M, size_t N, bool SO>
void blaze::HybridMatrix< Type, M, N, SO >::extend ( size_t  m,
size_t  n,
bool  preserve = true 
)
inline

Extending the size of the matrix.

Parameters
mNumber of additional rows.
nNumber of additional columns.
preservetrue if the old values of the matrix should be preserved, false if not.
Returns
void

This function increases the matrix size by m rows and n columns. In case the resulting number of rows or columns is larger than the maximum number of rows or columns (i.e. if m > M or n > N) a std::invalid_argument exception is thrown. During this operation, all matrix elements are potentially changed. In order to preserve the old matrix values, the preserve flag can be set to true.
Note that new matrix elements are not initialized!

◆ isAliased()

template<typename Type , size_t M, size_t N, bool SO>
template<typename Other >
bool blaze::HybridMatrix< Type, M, N, SO >::isAliased ( const Other *  alias) const
inlinenoexcept

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

Parameters
aliasThe alias to be checked.
Returns
true in case the alias corresponds to this matrix, false if not.

This function returns whether the given address is aliased with the matrix. In contrast to the canAlias() function this function is not allowed to use compile time expressions to optimize the evaluation.

◆ isAligned()

template<typename Type , size_t M, size_t N, bool SO>
bool blaze::HybridMatrix< Type, M, N, SO >::isAligned ( ) const
inlinenoexcept

Returns whether the matrix is properly aligned in memory.

Returns
true in case the matrix is aligned, false if not.

This function returns whether the matrix is guaranteed to be properly aligned in memory, i.e. whether the beginning and the end of each row/column of the matrix are guaranteed to conform to the alignment restrictions of the element type Type.

◆ isIntact()

template<typename Type , size_t M, size_t N, bool SO>
bool blaze::HybridMatrix< Type, M, N, SO >::isIntact ( ) const
inlinenoexcept

Returns whether the invariants of the hybrid matrix are intact.

Returns
true in case the hybrid matrix's invariants are intact, false otherwise.

This function checks whether the invariants of the hybrid matrix are intact, i.e. if its state is valid. In case the invariants are intact, the function returns true, else it will return false.

◆ load()

template<typename Type , size_t M, size_t N, bool SO>
BLAZE_ALWAYS_INLINE HybridMatrix< Type, M, N, SO >::SIMDType blaze::HybridMatrix< Type, M, N, SO >::load ( size_t  i,
size_t  j 
) const
noexcept

Load of a SIMD element 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
The loaded SIMD element.

This function performs a load of a specific SIMD element of the dense matrix. The row index must be smaller than the number of rows and the column index must be smaller then the number of columns. Additionally, the column index (in case of a row-major matrix) or the row index (in case of a column-major matrix) must be a multiple of the number of values inside the SIMD element. This function must NOT be called explicitly! It is used internally for the performance optimized evaluation of expression templates. Calling this function explicitly might result in erroneous results and/or in compilation errors.

◆ loada()

template<typename Type , size_t M, size_t N, bool SO>
BLAZE_ALWAYS_INLINE HybridMatrix< Type, M, N, SO >::SIMDType blaze::HybridMatrix< Type, M, N, SO >::loada ( size_t  i,
size_t  j 
) const
noexcept

Aligned load of a SIMD element 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
The loaded SIMD element.

This function performs an aligned load of a specific SIMD element of the dense matrix. The row index must be smaller than the number of rows and the column index must be smaller than the number of columns. Additionally, the column index (in case of a row-major matrix) or the row index (in case of a column-major matrix) must be a multiple of the number of values inside the SIMD element. This function must NOT be called explicitly! It is used internally for the performance optimized evaluation of expression templates. Calling this function explicitly might result in erroneous results and/or in compilation errors.

◆ loadu()

template<typename Type , size_t M, size_t N, bool SO>
BLAZE_ALWAYS_INLINE HybridMatrix< Type, M, N, SO >::SIMDType blaze::HybridMatrix< Type, M, N, SO >::loadu ( size_t  i,
size_t  j 
) const
noexcept

Unaligned load of a SIMD element 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
The loaded SIMD element.

This function performs an unaligned load of a specific SIMD element of the dense matrix. The row index must be smaller than the number of rows and the column index must be smaller than the number of columns. Additionally, the column index (in case of a row-major matrix) or the row index (in case of a column-major matrix) must be a multiple of the number of values inside the SIMD element. This function must NOT be called explicitly! It is used internally for the performance optimized evaluation of expression templates. Calling this function explicitly might result in erroneous results and/or in compilation errors.

◆ nonZeros() [1/2]

template<typename Type , size_t M, size_t N, bool SO>
size_t blaze::HybridMatrix< Type, M, N, SO >::nonZeros ( ) const
inline

Returns the total number of non-zero elements in the matrix.

Returns
The number of non-zero elements in the matrix.

◆ nonZeros() [2/2]

template<typename Type , size_t M, size_t N, bool SO>
size_t blaze::HybridMatrix< Type, M, N, SO >::nonZeros ( size_t  i) const
inline

Returns the number of non-zero elements in the specified row/column.

Parameters
iThe index of the row/column.
Returns
The number of non-zero elements of row/column i.

This function returns the current number of non-zero elements in the specified row/column. In case the storage order is set to rowMajor the function returns the number of non-zero elements in row i, in case the storage flag is set to columnMajor the function returns the number of non-zero elements in column i.

◆ operator delete() [1/2]

template<typename Type , size_t M, size_t N, bool SO>
void blaze::HybridMatrix< Type, M, N, SO >::operator delete ( void *  ptr)
inlinestatic

Class specific implementation of operator delete.

Parameters
ptrThe memory to be deallocated.
Returns
void

◆ operator delete() [2/2]

template<typename Type , size_t M, size_t N, bool SO>
void blaze::HybridMatrix< Type, M, N, SO >::operator delete ( void *  ptr,
const std::nothrow_t &   
)
inlinestatic

Class specific implementation of no-throw operator delete.

Parameters
ptrThe memory to be deallocated.
Returns
void

◆ operator delete[]() [1/2]

template<typename Type , size_t M, size_t N, bool SO>
void blaze::HybridMatrix< Type, M, N, SO >::operator delete[] ( void *  ptr)
inlinestatic

Class specific implementation of operator delete[].

Parameters
ptrThe memory to be deallocated.
Returns
void

◆ operator delete[]() [2/2]

template<typename Type , size_t M, size_t N, bool SO>
void blaze::HybridMatrix< Type, M, N, SO >::operator delete[] ( void *  ptr,
const std::nothrow_t &   
)
inlinestatic

Class specific implementation of no-throw operator delete[].

Parameters
ptrThe memory to be deallocated.
Returns
void

◆ operator new() [1/2]

template<typename Type , size_t M, size_t N, bool SO>
void * blaze::HybridMatrix< Type, M, N, SO >::operator new ( std::size_t  size)
inlinestatic

Class specific implementation of operator new.

Parameters
sizeThe total number of bytes to be allocated.
Returns
Pointer to the newly allocated memory.
Exceptions
std::bad_allocAllocation failed.

This class-specific implementation of operator new provides the functionality to allocate dynamic memory based on the alignment restrictions of the HybridMatrix class template.

◆ operator new() [2/2]

template<typename Type , size_t M, size_t N, bool SO>
void * blaze::HybridMatrix< Type, M, N, SO >::operator new ( std::size_t  size,
const std::nothrow_t &   
)
inlinestatic

Class specific implementation of the no-throw operator new.

Parameters
sizeThe total number of bytes to be allocated.
Returns
Pointer to the newly allocated memory.
Exceptions
std::bad_allocAllocation failed.

This class-specific implementation of operator new provides the functionality to allocate dynamic memory based on the alignment restrictions of the HybridMatrix class template.

◆ operator new[]() [1/2]

template<typename Type , size_t M, size_t N, bool SO>
void * blaze::HybridMatrix< Type, M, N, SO >::operator new[] ( std::size_t  size)
inlinestatic

Class specific implementation of operator new[].

Parameters
sizeThe total number of bytes to be allocated.
Returns
Pointer to the newly allocated memory.
Exceptions
std::bad_allocAllocation failed.

This class-specific implementation of operator new provides the functionality to allocate dynamic memory based on the alignment restrictions of the HybridMatrix class template.

◆ operator new[]() [2/2]

template<typename Type , size_t M, size_t N, bool SO>
void * blaze::HybridMatrix< Type, M, N, SO >::operator new[] ( std::size_t  size,
const std::nothrow_t &   
)
inlinestatic

Class specific implementation of the no-throw operator new[].

Parameters
sizeThe total number of bytes to be allocated.
Returns
Pointer to the newly allocated memory.
Exceptions
std::bad_allocAllocation failed.

This class-specific implementation of operator new provides the functionality to allocate dynamic memory based on the alignment restrictions of the HybridMatrix class template.

◆ operator%=()

template<typename Type, size_t M, size_t N, bool SO = defaultStorageOrder>
template<typename MT , bool SO2>
HybridMatrix<Type,M,N,SO>& blaze::HybridMatrix< Type, M, N, SO >::operator%= ( const Matrix< MT, SO2 > &  rhs)
inline

Schur product assignment operator for the multiplication of a matrix ( $ A\circ=B $).

Parameters
rhsThe right-hand side matrix for the Schur product.
Returns
Reference to the matrix.
Exceptions
std::invalid_argumentMatrix sizes do not match.

In case the current sizes of the two matrices don't match, a std::invalid_argument exception is thrown.

◆ operator()() [1/2]

template<typename Type , size_t M, size_t N, bool SO>
HybridMatrix< Type, M, N, SO >::Reference blaze::HybridMatrix< Type, M, N, SO >::operator() ( size_t  i,
size_t  j 
)
inlinenoexcept

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
Reference to the accessed value.

This function only performs an index check in case BLAZE_USER_ASSERT() is active. In contrast, the at() function is guaranteed to perform a check of the given access indices.

◆ operator()() [2/2]

template<typename Type , size_t M, size_t N, bool SO>
HybridMatrix< Type, M, N, SO >::ConstReference blaze::HybridMatrix< Type, M, N, SO >::operator() ( size_t  i,
size_t  j 
) const
inlinenoexcept

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
Reference-to-const to the accessed value.

This function only performs an index check in case BLAZE_USER_ASSERT() is active. In contrast, the at() function is guaranteed to perform a check of the given access indices.

◆ operator+=()

template<typename Type, size_t M, size_t N, bool SO = defaultStorageOrder>
template<typename MT , bool SO2>
HybridMatrix<Type,M,N,SO>& blaze::HybridMatrix< Type, M, N, SO >::operator+= ( const Matrix< MT, SO2 > &  rhs)
inline

Addition assignment operator for the addition of a matrix ( $ A+=B $).

Parameters
rhsThe right-hand side matrix to be added to the matrix.
Returns
Reference to the matrix.
Exceptions
std::invalid_argumentMatrix sizes do not match.

In case the current sizes of the two matrices don't match, a std::invalid_argument exception is thrown.

◆ operator-=()

template<typename Type, size_t M, size_t N, bool SO = defaultStorageOrder>
template<typename MT , bool SO2>
HybridMatrix<Type,M,N,SO>& blaze::HybridMatrix< Type, M, N, SO >::operator-= ( const Matrix< MT, SO2 > &  rhs)
inline

Subtraction assignment operator for the subtraction of a matrix ( $ A-=B $).

Parameters
rhsThe right-hand side matrix to be subtracted from the matrix.
Returns
Reference to the matrix.
Exceptions
std::invalid_argumentMatrix sizes do not match.

In case the current sizes of the two matrices don't match, a std::invalid_argument exception is thrown.

◆ operator=() [1/5]

template<typename Type , size_t M, size_t N, bool SO>
HybridMatrix< Type, M, N, SO > & blaze::HybridMatrix< Type, M, N, SO >::operator= ( const Type &  set)
inline

Homogenous assignment to all matrix elements.

Parameters
setScalar value to be assigned to all matrix elements.
Returns
Reference to the assigned matrix.

◆ operator=() [2/5]

template<typename Type , size_t M, size_t N, bool SO>
HybridMatrix< Type, M, N, SO > & blaze::HybridMatrix< Type, M, N, SO >::operator= ( initializer_list< initializer_list< Type > >  list)
inline

List assignment to all matrix elements.

Parameters
listThe initializer list.
Exceptions
std::invalid_argumentInvalid number of rows for hybrid matrix.
std::invalid_argumentInvalid number of columns for hybrid matrix.

This assignment operator offers the option to directly assign to all elements of the matrix by means of an initializer list:

A = { { 1, 2, 3 },
{ 4, 5 },
{ 7, 8, 9 } };

The matrix is resized according to the given initializer list and all its elements are assigned the values from the given initializer list. Missing values are initialized as default (as e.g. the value 6 in the example). Note that in case the size of the top-level initializer list exceeds the maximum number of rows or the size of any nested list exceeds the maximum number of columns, a std::invalid_argument exception is thrown.

◆ operator=() [3/5]

template<typename Type , size_t M, size_t N, bool SO>
HybridMatrix< Type, M, N, SO > & blaze::HybridMatrix< Type, M, N, SO >::operator= ( const HybridMatrix< Type, M, N, SO > &  rhs)
inline

Copy assignment operator for HybridMatrix.

Parameters
rhsMatrix to be copied.
Returns
Reference to the assigned matrix.

Explicit definition of a copy assignment operator for performance reasons.

◆ operator=() [4/5]

template<typename Type, size_t M, size_t N, bool SO = defaultStorageOrder>
template<typename Other , size_t Rows, size_t Cols>
HybridMatrix<Type,M,N,SO>& blaze::HybridMatrix< Type, M, N, SO >::operator= ( const Other(&)  array[Rows][Cols])
inline

Array assignment to all matrix elements.

Parameters
array$ M \times N $ dimensional array for the assignment.
Returns
Reference to the assigned matrix.

This assignment operator offers the option to directly set all elements of the matrix:

const int init[3][3] = { { 1, 2, 3 },
{ 4, 5 },
{ 7, 8, 9 } };
A = init;

The matrix is assigned the values from the given array. Missing values are initialized with default values (as e.g. the value 6 in the example).

◆ operator=() [5/5]

template<typename Type, size_t M, size_t N, bool SO = defaultStorageOrder>
template<typename MT , bool SO2>
HybridMatrix<Type,M,N,SO>& blaze::HybridMatrix< Type, M, N, SO >::operator= ( const Matrix< MT, SO2 > &  rhs)
inline

Assignment operator for different matrices.

Parameters
rhsMatrix to be copied.
Returns
Reference to the assigned matrix.
Exceptions
std::invalid_argumentInvalid assignment to hybrid matrix.

This constructor initializes the matrix as a copy of the given matrix. In case the number of rows of the given matrix is not M or the number of columns is not N, a std::invalid_argument exception is thrown.

◆ operator~() [1/2]

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

Conversion operator for constant matrices.

Returns
Constant reference of the actual type of the matrix.

◆ reset() [1/2]

template<typename Type , size_t M, size_t N, bool SO>
void blaze::HybridMatrix< Type, M, N, SO >::reset ( )
inline

Reset to the default initial values.

Returns
void

◆ reset() [2/2]

template<typename Type , size_t M, size_t N, bool SO>
void blaze::HybridMatrix< Type, M, N, SO >::reset ( size_t  i)
inline

Reset the specified row/column to the default initial values.

Parameters
iThe index of the row/column.
Returns
void

This function resets the values in the specified row/column to their default value. In case the storage order is set to rowMajor the function resets the values in row i, in case the storage order is set to columnMajor the function resets the values in column i. Note that the capacity of the row/column remains unchanged.

◆ resize()

template<typename Type , size_t M, size_t N, bool SO>
void blaze::HybridMatrix< Type, M, N, SO >::resize ( size_t  m,
size_t  n,
bool  preserve = true 
)

Changing the size of the matrix.

Parameters
mThe new number of rows of the matrix.
nThe new number of columns of the matrix.
preservetrue if the old values of the matrix should be preserved, false if not.
Returns
void

This function resizes the matrix using the given size to $ m \times n $. In case the given number of rows m is larger than the maximum number of rows (i.e. if m > M) or in case the given number of columns n is larger than the maximum number of column (i.e. if n > N) a std::invalid_argument exception is thrown. Note that this function may invalidate all existing views (submatrices, rows, columns, ...) on the matrix if it is used to shrink the matrix. Additionally, during this operation all matrix elements are potentially changed. In order to preserve the old matrix values, the preserve flag can be set to true.

Note that in case the number of rows or columns is increased new matrix elements are not initialized! The following example illustrates the resize operation of a $ 2 \times 4 $ matrix to a $ 4 \times 2 $ matrix. The new, uninitialized elements are marked with x:

\[ \left(\begin{array}{*{4}{c}} 1 & 2 & 3 & 4 \\ 5 & 6 & 7 & 8 \\ \end{array}\right) \Longrightarrow \left(\begin{array}{*{2}{c}} 1 & 2 \\ 5 & 6 \\ x & x \\ x & x \\ \end{array}\right) \]

◆ rows()

template<typename Type , size_t M, size_t N, bool SO>
size_t blaze::HybridMatrix< Type, M, N, SO >::rows ( ) const
inlinenoexcept

Returns the current number of rows of the matrix.

Returns
The number of rows of the matrix.

◆ scale()

template<typename Type, size_t M, size_t N, bool SO = defaultStorageOrder>
template<typename Other >
HybridMatrix<Type,M,N,SO>& blaze::HybridMatrix< Type, M, N, SO >::scale ( const Other &  scalar)
inline

Scaling of the matrix by the scalar value scalar ( $ A*=s $).

Parameters
scalarThe scalar value for the matrix scaling.
Returns
Reference to the matrix.

This function scales the matrix by applying the given scalar value scalar to each element of the matrix. For built-in and complex data types it has the same effect as using the multiplication assignment operator:

// ... Resizing and initialization
A *= 4; // Scaling of the matrix
A.scale( 4 ); // Same effect as above

◆ schurAssign() [1/4]

template<typename Type , size_t M, size_t N, bool SO>
template<typename MT >
void blaze::HybridMatrix< Type, M, N, SO >::schurAssign ( const SparseMatrix< MT, SO > &  rhs)
inline

Default implementation of the Schur product assignment of a row-major sparse matrix.

Parameters
rhsThe right-hand side sparse matrix for the Schur product.
Returns
void

This function must NOT be called explicitly! It is used internally for the performance optimized evaluation of expression templates. Calling this function explicitly might result in erroneous results and/or in compilation errors. Instead of using this function use the assignment operator.

◆ schurAssign() [2/4]

template<typename Type , size_t M, size_t N, bool SO>
template<typename MT >
void blaze::HybridMatrix< Type, M, N, SO >::schurAssign ( const SparseMatrix< MT,!SO > &  rhs)
inline

Default implementation of the Schur product assignment of a column-major sparse matrix.

Parameters
rhsThe right-hand side sparse matrix for the Schur product.
Returns
void

This function must NOT be called explicitly! It is used internally for the performance optimized evaluation of expression templates. Calling this function explicitly might result in erroneous results and/or in compilation errors. Instead of using this function use the assignment operator.

◆ schurAssign() [3/4]

template<typename Type, size_t M, size_t N, bool SO = defaultStorageOrder>
template<typename MT , bool SO2>
DisableIf_<typename HybridMatrix<Type,M,N,SO>::BLAZE_TEMPLATE VectorizedSchurAssign<MT> > blaze::HybridMatrix< Type, M, N, SO >::schurAssign ( const DenseMatrix< MT, SO2 > &  rhs)
inline

Default implementation of the Schur product assignment of a dense matrix.

Parameters
rhsThe right-hand side dense matrix for the Schur product
Returns
void

This function must NOT be called explicitly! It is used internally for the performance optimized evaluation of expression templates. Calling this function explicitly might result in erroneous results and/or in compilation errors. Instead of using this function use the assignment operator.

◆ schurAssign() [4/4]

template<typename Type, size_t M, size_t N, bool SO = defaultStorageOrder>
template<typename MT , bool SO2>
EnableIf_<typename HybridMatrix<Type,M,N,SO>::BLAZE_TEMPLATE VectorizedSchurAssign<MT> > blaze::HybridMatrix< Type, M, N, SO >::schurAssign ( const DenseMatrix< MT, SO2 > &  rhs)
inline

SIMD optimized implementation of the Schur product assignment of a dense matrix.

Parameters
rhsThe right-hand side dense matrix for the Schur product.
Returns
void

This function must NOT be called explicitly! It is used internally for the performance optimized evaluation of expression templates. Calling this function explicitly might result in erroneous results and/or in compilation errors. Instead of using this function use the assignment operator.

◆ spacing()

template<typename Type , size_t M, size_t N, bool SO>
constexpr size_t blaze::HybridMatrix< Type, M, N, SO >::spacing ( ) const
inlinenoexcept

Returns the spacing between the beginning of two rows.

Returns
The spacing between the beginning of two rows.

This function returns the spacing between the beginning of two rows, i.e. the total number of elements of a row.

◆ store()

template<typename Type , size_t M, size_t N, bool SO>
BLAZE_ALWAYS_INLINE void blaze::HybridMatrix< Type, M, N, SO >::store ( size_t  i,
size_t  j,
const SIMDType value 
)
noexcept

Store of a SIMD element 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].
valueThe SIMD element to be stored.
Returns
void

This function performs a store of a specific SIMD element of the dense matrix. The row index must be smaller than the number of rows and the column index must be smaller than the number of columns. Additionally, the column index (in case of a row-major matrix) or the row index (in case of a column-major matrix) must be a multiple of the number of values inside the SIMD element. This function must NOT be called explicitly! It is used internally for the performance optimized evaluation of expression templates. Calling this function explicitly might result in erroneous results and/or in compilation errors.

◆ storea()

template<typename Type , size_t M, size_t N, bool SO>
BLAZE_ALWAYS_INLINE void blaze::HybridMatrix< Type, M, N, SO >::storea ( size_t  i,
size_t  j,
const SIMDType value 
)
noexcept

Aligned store of a SIMD element 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].
valueThe SIMD element to be stored.
Returns
void

This function performs an aligned store of a specific SIMD element of the dense matrix. The row index must be smaller than the number of rows and the column index must be smaller than the number of columns. Additionally, the column index (in case of a row-major matrix) or the row index (in case of a column-major matrix) must be a multiple of the number of values inside the SIMD element. This function must NOT be called explicitly! It is used internally for the performance optimized evaluation of expression templates. Calling this function explicitly might result in erroneous results and/or in compilation errors.

◆ storeu()

template<typename Type , size_t M, size_t N, bool SO>
BLAZE_ALWAYS_INLINE void blaze::HybridMatrix< Type, M, N, SO >::storeu ( size_t  i,
size_t  j,
const SIMDType value 
)
noexcept

Unaligned store of a SIMD element 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].
valueThe SIMD element to be stored.
Returns
void

This function performs an unaligned store of a specific SIMD element of the dense matrix. The row index must be smaller than the number of rows and the column index must be smaller than the number of columns. Additionally, the column index (in case of a row-major matrix) or the row index (in case of a column-major matrix) must be a multiple of the number of values inside the SIMD element. This function must NOT be called explicitly! It is used internally for the performance optimized evaluation of expression templates. Calling this function explicitly might result in erroneous results and/or in compilation errors.

◆ stream()

template<typename Type , size_t M, size_t N, bool SO>
BLAZE_ALWAYS_INLINE void blaze::HybridMatrix< Type, M, N, SO >::stream ( size_t  i,
size_t  j,
const SIMDType value 
)
noexcept

Aligned, non-temporal store of a SIMD element 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].
valueThe SIMD element to be stored.
Returns
void

This function performs an aligned, non-temporal store of a specific SIMD element of the dense matrix. The row index must be smaller than the number of rows and the column index must be smaller than the number of columns. Additionally, the column index (in case of a row-major matrix) or the row index (in case of a column-major matrix) must be a multiple of the number of values inside the SIMD element. This function must NOT be called explicitly! It is used internally for the performance optimized evaluation of expression templates. Calling this function explicitly might result in erroneous results and/or in compilation errors.

◆ subAssign() [1/4]

template<typename Type , size_t M, size_t N, bool SO>
template<typename MT >
void blaze::HybridMatrix< Type, M, N, SO >::subAssign ( const SparseMatrix< MT, SO > &  rhs)
inline

Default implementation of the subtraction assignment of a row-major sparse matrix.

Parameters
rhsThe right-hand side sparse matrix to be subtracted.
Returns
void

This function must NOT be called explicitly! It is used internally for the performance optimized evaluation of expression templates. Calling this function explicitly might result in erroneous results and/or in compilation errors. Instead of using this function use the assignment operator.

◆ subAssign() [2/4]

template<typename Type , size_t M, size_t N, bool SO>
template<typename MT >
void blaze::HybridMatrix< Type, M, N, SO >::subAssign ( const SparseMatrix< MT,!SO > &  rhs)
inline

Default implementation of the subtraction assignment of a column-major sparse matrix.

Parameters
rhsThe right-hand side sparse matrix to be subtracted.
Returns
void

This function must NOT be called explicitly! It is used internally for the performance optimized evaluation of expression templates. Calling this function explicitly might result in erroneous results and/or in compilation errors. Instead of using this function use the assignment operator.

◆ subAssign() [3/4]

template<typename Type, size_t M, size_t N, bool SO = defaultStorageOrder>
template<typename MT , bool SO2>
DisableIf_<typename HybridMatrix<Type,M,N,SO>::BLAZE_TEMPLATE VectorizedSubAssign<MT> > blaze::HybridMatrix< Type, M, N, SO >::subAssign ( const DenseMatrix< MT, SO2 > &  rhs)
inline

Default implementation of the subtraction assignment of a dense matrix.

Parameters
rhsThe right-hand side dense matrix to be subtracted.
Returns
void

This function must NOT be called explicitly! It is used internally for the performance optimized evaluation of expression templates. Calling this function explicitly might result in erroneous results and/or in compilation errors. Instead of using this function use the assignment operator.

◆ subAssign() [4/4]

template<typename Type, size_t M, size_t N, bool SO = defaultStorageOrder>
template<typename MT , bool SO2>
EnableIf_<typename HybridMatrix<Type,M,N,SO>::BLAZE_TEMPLATE VectorizedSubAssign<MT> > blaze::HybridMatrix< Type, M, N, SO >::subAssign ( const DenseMatrix< MT, SO2 > &  rhs)
inline

SIMD optimized implementation of the subtraction assignment of a dense matrix.

Parameters
rhsThe right-hand side dense matrix to be subtracted.
Returns
void

This function must NOT be called explicitly! It is used internally for the performance optimized evaluation of expression templates. Calling this function explicitly might result in erroneous results and/or in compilation errors. Instead of using this function use the assignment operator.

◆ swap()

template<typename Type , size_t M, size_t N, bool SO>
void blaze::HybridMatrix< Type, M, N, SO >::swap ( HybridMatrix< Type, M, N, SO > &  m)
inlinenoexcept

Swapping the contents of two hybrid matrices.

Parameters
mThe matrix to be swapped.
Returns
void

◆ transpose()

template<typename Type , size_t M, size_t N, bool SO>
HybridMatrix< Type, M, N, SO > & blaze::HybridMatrix< Type, M, N, SO >::transpose ( )
inline

In-place transpose of the matrix.

Returns
Reference to the transposed matrix.
Exceptions
std::logic_errorImpossible transpose operation.

This function transposes the hybrid matrix in-place. Note that this function can only be used on hybrid matrices whose current dimensions allow an in-place transpose operation. In case the current number of rows is larger than the maximum number of columns or if the current number of columns is larger than the maximum number of rows, an std::logic_error is thrown.

Member Data Documentation

◆ v_

template<typename Type, size_t M, size_t N, bool SO = defaultStorageOrder>
AlignedArray<Type,M*NN> blaze::HybridMatrix< Type, M, N, SO >::v_
private

The statically allocated matrix elements.

Access to the matrix elements is gained via the function call operator. In case of row-major order the memory layout of the elements is

\[\left(\begin{array}{*{5}{c}} 0 & 1 & 2 & \cdots & N-1 \\ N & N+1 & N+2 & \cdots & 2 \cdot N-1 \\ \vdots & \vdots & \vdots & \ddots & \vdots \\ M \cdot N-N & M \cdot N-N+1 & M \cdot N-N+2 & \cdots & M \cdot N-1 \\ \end{array}\right)\]

.


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