All Classes Namespaces Files Functions Variables Typedefs Enumerations Enumerator Friends Macros Groups Pages
Public Types | Public Member Functions | Private Types | List of all members
blaze::HybridVector< typename, size_t, bool > Class Template Reference

Efficient implementation of a dynamically sized vector with static memory.The HybridVector class template combines the flexibility of a dynamically sized vector with the efficiency and performance of a fixed-size vector. It is implemented as a crossing between the blaze::StaticVector and the blaze::DynamicVector class templates: Similar to the static vector it uses static stack memory instead of dynamically allocated memory and similar to the dynamic vector it can be resized (within the extend of the static memory). The type of the elements, the maximum number of elements and the transpose flag of the vector can be specified via the three template parameters: More...

#include <HybridVector.h>

Inherits blaze::DenseVector< HybridVector< Type, N, TF >, TF >.

Public Types

enum  { vectorizable = IsVectorizable<Type>::value }
 Compilation flag for intrinsic optimization. More...
 
enum  { smpAssignable = 0 }
 Compilation flag for SMP assignments. More...
 
typedef HybridVector< Type, N, TF > This
 Type of this HybridVector instance.
 
typedef This ResultType
 Result type for expression template evaluations.
 
typedef HybridVector< Type, N,!TF > TransposeType
 Transpose type for expression template evaluations.
 
typedef Type ElementType
 Type of the vector elements.
 
typedef IT::Type IntrinsicType
 Intrinsic type of the vector elements.
 
typedef const Type & ReturnType
 Return type for expression template evaluations.
 
typedef const HybridVectorCompositeType
 Data type for composite expression templates.
 
typedef Type & Reference
 Reference to a non-constant vector value.
 
typedef const Type & ConstReference
 Reference to a constant vector value.
 
typedef DenseIterator< Type > Iterator
 Iterator over non-constant elements.
 
typedef DenseIterator< const Type > ConstIterator
 Iterator over constant elements.
 
typedef VT VectorType
 Type of the vector.
 

Public Member Functions

template<typename Other , size_t M>
HybridVector< Type, N, TF > & operator= (const Other(&array)[M])
 Array assignment to all vector elements. More...
 
template<typename VT >
HybridVector< Type, N, TF > & operator= (const Vector< VT, TF > &rhs)
 Assignment operator for different vectors. More...
 
template<typename VT >
HybridVector< Type, N, TF > & operator+= (const Vector< VT, TF > &rhs)
 Addition assignment operator for the addition of a vector ( $ \vec{a}+=\vec{b} $). More...
 
template<typename VT >
HybridVector< Type, N, TF > & operator-= (const Vector< VT, TF > &rhs)
 Subtraction assignment operator for the subtraction of a vector ( $ \vec{a}-=\vec{b} $). More...
 
template<typename VT >
HybridVector< Type, N, TF > & operator*= (const Vector< VT, TF > &rhs)
 Multiplication assignment operator for the multiplication of a vector ( $ \vec{a}*=\vec{b} $). More...
 
template<typename Other >
EnableIf< IsNumeric< Other >
, HybridVector< Type, N, TF >
>::Type & 
operator*= (Other rhs)
 Multiplication assignment operator for the multiplication between a vector and a scalar value ( $ \vec{a}*=s $). More...
 
template<typename Other >
EnableIf< IsNumeric< Other >
, HybridVector< Type, N, TF >
>::Type & 
operator/= (Other rhs)
 Division assignment operator for the division of a vector by a scalar value ( $ \vec{a}/=s $). More...
 
template<typename Other >
HybridVector< Type, N, TF > & scale (Other scalar)
 Scaling of the vector by the scalar value scalar ( $ \vec{a}*=s $). More...
 
template<typename VT >
DisableIf< typename
HybridVector< Type, N, TF >
::BLAZE_TEMPLATE
VectorizedAssign< VT > >::Type 
assign (const DenseVector< VT, TF > &rhs)
 Default implementation of the assignment of a dense vector. More...
 
template<typename VT >
EnableIf< typename
HybridVector< Type, N, TF >
::BLAZE_TEMPLATE
VectorizedAssign< VT > >::Type 
assign (const DenseVector< VT, TF > &rhs)
 Intrinsic optimized implementation of the assignment of a dense vector. More...
 
template<typename VT >
DisableIf< typename
HybridVector< Type, N, TF >
::BLAZE_TEMPLATE
VectorizedAddAssign< VT >
>::Type 
addAssign (const DenseVector< VT, TF > &rhs)
 Default implementation of the addition assignment of a dense vector. More...
 
template<typename VT >
EnableIf< typename
HybridVector< Type, N, TF >
::BLAZE_TEMPLATE
VectorizedAddAssign< VT >
>::Type 
addAssign (const DenseVector< VT, TF > &rhs)
 Intrinsic optimized implementation of the addition assignment of a dense vector. More...
 
template<typename VT >
DisableIf< typename
HybridVector< Type, N, TF >
::BLAZE_TEMPLATE
VectorizedSubAssign< VT >
>::Type 
subAssign (const DenseVector< VT, TF > &rhs)
 Default implementation of the subtraction assignment of a dense vector. More...
 
template<typename VT >
EnableIf< typename
HybridVector< Type, N, TF >
::BLAZE_TEMPLATE
VectorizedSubAssign< VT >
>::Type 
subAssign (const DenseVector< VT, TF > &rhs)
 Intrinsic optimized implementation of the subtraction assignment of a dense vector. More...
 
template<typename VT >
DisableIf< typename
HybridVector< Type, N, TF >
::BLAZE_TEMPLATE
VectorizedMultAssign< VT >
>::Type 
multAssign (const DenseVector< VT, TF > &rhs)
 Default implementation of the multiplication assignment of a dense vector. More...
 
template<typename VT >
EnableIf< typename
HybridVector< Type, N, TF >
::BLAZE_TEMPLATE
VectorizedMultAssign< VT >
>::Type 
multAssign (const DenseVector< VT, TF > &rhs)
 Intrinsic optimized implementation of the multiplication assignment of a dense vector. More...
 
VectorTypeoperator~ ()
 Conversion operator for non-constant vectors. More...
 
const VectorTypeoperator~ () const
 Conversion operator for constant vectors. More...
 
Constructors
 HybridVector ()
 The default constructor for HybridVector. More...
 
 HybridVector (size_t n)
 Constructor for a vector of size n. More...
 
 HybridVector (size_t n, const Type &init)
 Constructor for a homogeneous initialization of all n vector elements. More...
 
template<typename Other >
 HybridVector (size_t n, const Other *array)
 Array initialization of all vector elements. More...
 
template<typename Other , size_t M>
 HybridVector (const Other(&array)[M])
 Array initialization of all vector elements. More...
 
 HybridVector (const HybridVector &v)
 The copy constructor for HybridVector. More...
 
template<typename VT >
 HybridVector (const Vector< VT, TF > &v)
 Conversion constructor from different vectors. More...
 
Data access functions
Reference operator[] (size_t index)
 Subscript operator for the direct access to the vector elements. More...
 
ConstReference operator[] (size_t index) const
 Subscript operator for the direct access to the vector elements. More...
 
Type * data ()
 Low-level data access to the vector elements. More...
 
const Type * data () const
 Low-level data access to the vector elements. More...
 
Iterator begin ()
 Returns an iterator to the first element of the hybrid vector. More...
 
ConstIterator begin () const
 Returns an iterator to the first element of the hybrid vector. More...
 
ConstIterator cbegin () const
 Returns an iterator to the first element of the hybrid vector. More...
 
Iterator end ()
 Returns an iterator just past the last element of the hybrid vector. More...
 
ConstIterator end () const
 Returns an iterator just past the last element of the hybrid vector. More...
 
ConstIterator cend () const
 Returns an iterator just past the last element of the hybrid vector. More...
 
Assignment operators
template<typename Other , size_t M>
HybridVectoroperator= (const Other(&array)[M])
 
HybridVectoroperator= (const Type &rhs)
 Homogenous assignment to all vector elements. More...
 
HybridVectoroperator= (const HybridVector &rhs)
 Copy assignment operator for HybridVector. More...
 
template<typename VT >
HybridVectoroperator= (const Vector< VT, TF > &rhs)
 
template<typename VT >
HybridVectoroperator+= (const Vector< VT, TF > &rhs)
 
template<typename VT >
HybridVectoroperator-= (const Vector< VT, TF > &rhs)
 
template<typename VT >
HybridVectoroperator*= (const Vector< VT, TF > &rhs)
 
template<typename Other >
EnableIf< IsNumeric< Other >
, HybridVector >::Type & 
operator*= (Other rhs)
 
template<typename Other >
EnableIf< IsNumeric< Other >
, HybridVector >::Type & 
operator/= (Other rhs)
 
Utility functions
size_t size () const
 Returns the current size/dimension of the vector. More...
 
size_t capacity () const
 Returns the maximum capacity of the vector. More...
 
size_t nonZeros () const
 Returns the number of non-zero elements in the vector. More...
 
void reset ()
 Reset to the default initial values. More...
 
void clear ()
 Clearing the vector. More...
 
void resize (size_t n, bool preserve=true)
 Changing the size of the vector. More...
 
void extend (size_t n, bool preserve=true)
 Extending the size of the vector. More...
 
template<typename Other >
HybridVectorscale (Other scalar)
 
void swap (HybridVector &v)
 Swapping the contents of two hybrid vectors. More...
 
Expression template evaluation functions
template<typename Other >
bool canAlias (const Other *alias) const
 Returns whether the vector can alias with the given address alias. More...
 
template<typename Other >
bool isAliased (const Other *alias) const
 Returns whether the vector is aliased with the given address alias. More...
 
bool isAligned () const
 Returns whether the vector is properly aligned in memory. More...
 
IntrinsicType load (size_t index) const
 Aligned load of an intrinsic element of the vector. More...
 
IntrinsicType loadu (size_t index) const
 Unaligned load of an intrinsic element of the vector. More...
 
void store (size_t index, const IntrinsicType &value)
 Aligned store of an intrinsic element of the vector. More...
 
void storeu (size_t index, const IntrinsicType &value)
 Unaligned store of an intrinsic element of the vector. More...
 
void stream (size_t index, const IntrinsicType &value)
 Aligned, non-temporal store of an intrinsic element of the vector. More...
 
template<typename VT >
DisableIf< VectorizedAssign
< VT > >::Type 
assign (const DenseVector< VT, TF > &rhs)
 
template<typename VT >
EnableIf< VectorizedAssign< VT >
>::Type 
assign (const DenseVector< VT, TF > &rhs)
 
template<typename VT >
void assign (const SparseVector< VT, TF > &rhs)
 Default implementation of the assignment of a sparse vector. More...
 
template<typename VT >
DisableIf< VectorizedAddAssign
< VT > >::Type 
addAssign (const DenseVector< VT, TF > &rhs)
 
template<typename VT >
EnableIf< VectorizedAddAssign
< VT > >::Type 
addAssign (const DenseVector< VT, TF > &rhs)
 
template<typename VT >
void addAssign (const SparseVector< VT, TF > &rhs)
 Default implementation of the addition assignment of a sparse vector. More...
 
template<typename VT >
DisableIf< VectorizedSubAssign
< VT > >::Type 
subAssign (const DenseVector< VT, TF > &rhs)
 
template<typename VT >
EnableIf< VectorizedSubAssign
< VT > >::Type 
subAssign (const DenseVector< VT, TF > &rhs)
 
template<typename VT >
void subAssign (const SparseVector< VT, TF > &rhs)
 Default implementation of the subtraction assignment of a sparse vector. More...
 
template<typename VT >
DisableIf
< VectorizedMultAssign< VT >
>::Type 
multAssign (const DenseVector< VT, TF > &rhs)
 
template<typename VT >
EnableIf< VectorizedMultAssign
< VT > >::Type 
multAssign (const DenseVector< VT, TF > &rhs)
 
template<typename VT >
void multAssign (const SparseVector< VT, TF > &rhs)
 Default implementation of the multiplication assignment of a sparse vector. More...
 

Private Types

enum  { NN = N + ( IT::size - ( N % IT::size ) ) % IT::size }
 Alignment adjustment.
 
typedef IntrinsicTrait< Type > IT
 Intrinsic trait for the vector element type.
 

Private Attributes

Member variables
AlignedArray< Type, NN > v_
 The statically allocated vector elements. More...
 
size_t size_
 The current size/dimension of the vector.
 

Detailed Description

template<typename, size_t, bool>
class blaze::HybridVector< typename, size_t, bool >

Efficient implementation of a dynamically sized vector with static memory.

The HybridVector class template combines the flexibility of a dynamically sized vector with the efficiency and performance of a fixed-size vector. It is implemented as a crossing between the blaze::StaticVector and the blaze::DynamicVector class templates: Similar to the static vector it uses static stack memory instead of dynamically allocated memory and similar to the dynamic vector it can be resized (within the extend of the static memory). The type of the elements, the maximum number of elements and the transpose flag of the vector can be specified via the three template parameters:

template< typename Type, size_t N, bool TF >

These contiguously stored elements can be directly accessed with the subscript operator. The numbering of the vector elements is

\[\left(\begin{array}{*{4}{c}} 0 & 1 & \cdots & N-1 \\ \end{array}\right)\]

The use of HybridVector is very natural and intuitive. All operations (addition, subtraction, multiplication, scaling, ...) can be performed on all possible combinations of dense and sparse vectors with fitting element types. The following example gives an impression of the use of a 2-dimensional HybridVector:

HybridVector<double,2UL> a; // Default initialized 2D vector
a[0] = 1.0; // Initialization of the first element
a[1] = 2.0; // Initialization of the second element
HybridVector<double,2UL> b( 3.0, 2.0 ); // Directly initialized 2D vector
CompressedVector<float> c( 4.0F ); // Empty single precision vector
HybridVector<double,2UL> d; // Default constructed hybrid vector
StaticMatrix<double,2UL,2UL> A; // Default constructed static row-major matrix
d = a + b; // Vector addition between vectors of equal element type
d = a - c; // Vector subtraction between a dense and sparse vector with different element types
d = a * b; // Component-wise vector multiplication
a *= 2.0; // In-place scaling of vector
d = a * 2.0; // Scaling of vector a
d = 2.0 * a; // Scaling of vector a
d += a - b; // Addition assignment
d -= a + c; // Subtraction assignment
d *= a * b; // Multiplication assignment
double scalar = trans( a ) * b; // Scalar/dot/inner product between two vectors
A = a * trans( b ); // Outer product between two vectors

In order to provide maximum performance HybridVector is guaranteed to be properly aligned in memory based on the alignment restrictions of the specified element type. Note however that this enforced alignment can cause problems when HybridVector is used in conjunction with a std::vector or a similar container:

std::vector< VectorType > vec( 10 ); // Potentially causes segmentation faults

The core problem is that the allocated dynamic memory of the container potentially does not satisfy the alignment restrictions of the HybridVector. For that reason, the AlignedAllocator can be used:

std::vector< VectorType, AllocatorType > vec( 10 ); // Guaranteed to work

Member Enumeration Documentation

template<typename , size_t , bool >
anonymous enum

Compilation flag for intrinsic optimization.

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

template<typename , size_t , bool >
anonymous enum

Compilation flag for SMP assignments.

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

Constructor & Destructor Documentation

template<typename Type , size_t N, bool TF>
blaze::HybridVector< Type, N, TF >::HybridVector ( )
inlineexplicit

The default constructor for HybridVector.

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

template<typename Type , size_t N, bool TF>
blaze::HybridVector< Type, N, TF >::HybridVector ( size_t  n)
inlineexplicit

Constructor for a vector of size n.

Parameters
nThe size of the vector.
Exceptions
std::invalid_argumentInvalid size for hybrid vector.

This constructor creates a hybrid vector of size n and initializes all vector elements to the default value (for instance 0 for integral types). In case n is larger than the maximum allowed number of elements (i.e. n > N) a std::invalid_argument exception is thrown.

template<typename Type , size_t N, bool TF>
blaze::HybridVector< Type, N, TF >::HybridVector ( size_t  n,
const Type &  init 
)
inlineexplicit

Constructor for a homogeneous initialization of all n vector elements.

Parameters
nThe size of the vector.
initThe initial value of the vector elements.
Exceptions
std::invalid_argumentInvalid size for hybrid vector.

This constructor creates a hybrid vector of size n and initializes all vector elements with the specified value. In case n is larger than the maximum allowed number of elements (i.e. n > N) a std::invalid_argument exception is thrown.

template<typename Type , size_t N, bool TF>
template<typename Other >
blaze::HybridVector< Type, N, TF >::HybridVector ( size_t  n,
const Other *  array 
)
inlineexplicit

Array initialization of all vector elements.

Parameters
nThe size of the vector.
arrayDynamic array for the initialization.

This assignment operator offers the option to directly initialize the elements of the vector with a dynamic array:

double* array = new double[6];
// ... Initialization of the dynamic array
delete[] array;

The vector is sized according to the size of the array and initialized with the values from the given array. In case the size of the given array exceeds the maximum size of the hybrid vector (i.e. is larger than N), a std::invalid_argument exception is thrown.
Note that it is expected that the given array has at least n elements. Providing an array with less elements results in undefined behavior!

template<typename Type , size_t N, bool TF>
template<typename Other , size_t M>
blaze::HybridVector< Type, N, TF >::HybridVector ( const Other(&)  array[M])
inlineexplicit

Array initialization of all vector elements.

Parameters
arrayM-dimensional array for the initialization.

This assignment operator offers the option to directly initialize the elements of the vector with a static array:

const double init[2] = { 1.0, 2.0 };

The vector is sized according to the size of the array and initialized with the values from the given array. This constructor only works for arrays with a size smaller-or-equal than the maximum number of elements of the hybrid vector (i.e. M <= N). The attempt to use a larger array will result in a compile time error.

template<typename Type , size_t N, bool TF>
blaze::HybridVector< Type, N, TF >::HybridVector ( const HybridVector< typename, size_t, bool > &  v)
inline

The copy constructor for HybridVector.

Parameters
vVector to be copied.

The copy constructor is explicitly defined in order to enable/facilitate NRV optimization.

template<typename Type , size_t N, bool TF>
template<typename VT >
blaze::HybridVector< Type, N, TF >::HybridVector ( const Vector< VT, TF > &  v)
inline

Conversion constructor from different vectors.

Parameters
vVector to be copied.
Exceptions
std::invalid_argumentInvalid setup of hybrid vector.

This constructor initializes the hybrid vector from the given vector. In case the size of the given vector exceeds the maximum size of the hybrid vector (i.e. is larger than N), a std::invalid_argument exception is thrown.

Member Function Documentation

template<typename Type , size_t N, bool TF>
template<typename VT >
void blaze::HybridVector< Type, N, TF >::addAssign ( const SparseVector< VT, TF > &  rhs)
inline

Default implementation of the addition assignment of a sparse vector.

Parameters
rhsThe right-hand side sparse vector 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.

template<typename , size_t , bool >
template<typename VT >
DisableIf< typename HybridVector<Type,N,TF>::BLAZE_TEMPLATE VectorizedAddAssign<VT> >::Type blaze::HybridVector< typename, size_t, bool >::addAssign ( const DenseVector< VT, TF > &  rhs)
inline

Default implementation of the addition assignment of a dense vector.

Parameters
rhsThe right-hand side dense vector 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.

template<typename , size_t , bool >
template<typename VT >
EnableIf< typename HybridVector<Type,N,TF>::BLAZE_TEMPLATE VectorizedAddAssign<VT> >::Type blaze::HybridVector< typename, size_t, bool >::addAssign ( const DenseVector< VT, TF > &  rhs)
inline

Intrinsic optimized implementation of the addition assignment of a dense vector.

Parameters
rhsThe right-hand side dense vector 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.

template<typename Type , size_t N, bool TF>
template<typename VT >
void blaze::HybridVector< Type, N, TF >::assign ( const SparseVector< VT, TF > &  rhs)
inline

Default implementation of the assignment of a sparse vector.

Parameters
rhsThe right-hand side sparse vector 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.

template<typename , size_t , bool >
template<typename VT >
DisableIf< typename HybridVector<Type,N,TF>::BLAZE_TEMPLATE VectorizedAssign<VT> >::Type blaze::HybridVector< typename, size_t, bool >::assign ( const DenseVector< VT, TF > &  rhs)
inline

Default implementation of the assignment of a dense vector.

Parameters
rhsThe right-hand side dense vector 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.

template<typename , size_t , bool >
template<typename VT >
EnableIf< typename HybridVector<Type,N,TF>::BLAZE_TEMPLATE VectorizedAssign<VT> >::Type blaze::HybridVector< typename, size_t, bool >::assign ( const DenseVector< VT, TF > &  rhs)
inline

Intrinsic optimized implementation of the assignment of a dense vector.

Parameters
rhsThe right-hand side dense vector 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.

template<typename Type , size_t N, bool TF>
HybridVector< Type, N, TF >::Iterator blaze::HybridVector< Type, N, TF >::begin ( )
inline

Returns an iterator to the first element of the hybrid vector.

Returns
Iterator to the first element of the hybrid vector.
template<typename Type , size_t N, bool TF>
HybridVector< Type, N, TF >::ConstIterator blaze::HybridVector< Type, N, TF >::begin ( ) const
inline

Returns an iterator to the first element of the hybrid vector.

Returns
Iterator to the first element of the hybrid vector.
template<typename Type , size_t N, bool TF>
template<typename Other >
bool blaze::HybridVector< Type, N, TF >::canAlias ( const Other *  alias) const
inline

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

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

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

template<typename Type , size_t N, bool TF>
size_t blaze::HybridVector< Type, N, TF >::capacity ( ) const
inline

Returns the maximum capacity of the vector.

Returns
The capacity of the vector.
template<typename Type , size_t N, bool TF>
HybridVector< Type, N, TF >::ConstIterator blaze::HybridVector< Type, N, TF >::cbegin ( ) const
inline

Returns an iterator to the first element of the hybrid vector.

Returns
Iterator to the first element of the hybrid vector.
template<typename Type , size_t N, bool TF>
HybridVector< Type, N, TF >::ConstIterator blaze::HybridVector< Type, N, TF >::cend ( ) const
inline

Returns an iterator just past the last element of the hybrid vector.

Returns
Iterator just past the last element of the hybrid vector.
template<typename Type , size_t N, bool TF>
void blaze::HybridVector< Type, N, TF >::clear ( )
inline

Clearing the vector.

Returns
void

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

template<typename Type , size_t N, bool TF>
Type * blaze::HybridVector< Type, N, TF >::data ( )
inline

Low-level data access to the vector elements.

Returns
Pointer to the internal element storage.

This function returns a pointer to the internal storage of the hybrid vector.

template<typename Type , size_t N, bool TF>
const Type * blaze::HybridVector< Type, N, TF >::data ( ) const
inline

Low-level data access to the vector elements.

Returns
Pointer to the internal element storage.

This function returns a pointer to the internal storage of the hybrid vector.

template<typename Type , size_t N, bool TF>
HybridVector< Type, N, TF >::Iterator blaze::HybridVector< Type, N, TF >::end ( )
inline

Returns an iterator just past the last element of the hybrid vector.

Returns
Iterator just past the last element of the hybrid vector.
template<typename Type , size_t N, bool TF>
HybridVector< Type, N, TF >::ConstIterator blaze::HybridVector< Type, N, TF >::end ( ) const
inline

Returns an iterator just past the last element of the hybrid vector.

Returns
Iterator just past the last element of the hybrid vector.
template<typename Type , size_t N, bool TF>
void blaze::HybridVector< Type, N, TF >::extend ( size_t  n,
bool  preserve = true 
)
inline

Extending the size of the vector.

Parameters
nNumber of additional vector elements.
preservetrue if the old values of the vector should be preserved, false if not.
Returns
void

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

template<typename Type , size_t N, bool TF>
template<typename Other >
bool blaze::HybridVector< Type, N, TF >::isAliased ( const Other *  alias) const
inline

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

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

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

template<typename Type , size_t N, bool TF>
bool blaze::HybridVector< Type, N, TF >::isAligned ( ) const
inline

Returns whether the vector is properly aligned in memory.

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

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

template<typename Type , size_t N, bool TF>
HybridVector< Type, N, TF >::IntrinsicType blaze::HybridVector< Type, N, TF >::load ( size_t  index) const
inline

Aligned load of an intrinsic element of the vector.

Parameters
indexAccess index. The index must be smaller than the number of vector elements.
Returns
The loaded intrinsic element.

This function performs an aligned load of a specific intrinsic element of the dense vector. The index must be smaller than the number of vector elements and it must be a multiple of the number of values inside the intrinsic 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.

template<typename Type , size_t N, bool TF>
HybridVector< Type, N, TF >::IntrinsicType blaze::HybridVector< Type, N, TF >::loadu ( size_t  index) const
inline

Unaligned load of an intrinsic element of the vector.

Parameters
indexAccess index. The index must be smaller than the number of vector elements.
Returns
The loaded intrinsic element.

This function performs an unaligned load of a specific intrinsic element of the dense vector. The index must be smaller than the number of vector elements and it must be a multiple of the number of values inside the intrinsic 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.

template<typename Type , size_t N, bool TF>
template<typename VT >
void blaze::HybridVector< Type, N, TF >::multAssign ( const SparseVector< VT, TF > &  rhs)
inline

Default implementation of the multiplication assignment of a sparse vector.

Parameters
rhsThe right-hand side sparse vector to be multiplied.
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.

template<typename , size_t , bool >
template<typename VT >
DisableIf< typename HybridVector<Type,N,TF>::BLAZE_TEMPLATE VectorizedMultAssign<VT> >::Type blaze::HybridVector< typename, size_t, bool >::multAssign ( const DenseVector< VT, TF > &  rhs)
inline

Default implementation of the multiplication assignment of a dense vector.

Parameters
rhsThe right-hand side dense vector to be multiplied.
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.

template<typename , size_t , bool >
template<typename VT >
EnableIf< typename HybridVector<Type,N,TF>::BLAZE_TEMPLATE VectorizedMultAssign<VT> >::Type blaze::HybridVector< typename, size_t, bool >::multAssign ( const DenseVector< VT, TF > &  rhs)
inline

Intrinsic optimized implementation of the multiplication assignment of a dense vector.

Parameters
rhsThe right-hand side dense vector to be multiplied.
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.

template<typename Type , size_t N, bool TF>
size_t blaze::HybridVector< Type, N, TF >::nonZeros ( ) const
inline

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

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

Note that the number of non-zero elements is always less than or equal to the current size of the vector.

template<typename , size_t , bool >
template<typename VT >
HybridVector<Type,N,TF>& blaze::HybridVector< typename, size_t, bool >::operator*= ( const Vector< VT, TF > &  rhs)
inline

Multiplication assignment operator for the multiplication of a vector ( $ \vec{a}*=\vec{b} $).

Parameters
rhsThe right-hand side vector to be multiplied with the vector.
Returns
Reference to the vector.
Exceptions
std::invalid_argumentVector sizes do not match.

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

template<typename , size_t , bool >
template<typename Other >
EnableIf< IsNumeric<Other>, HybridVector<Type,N,TF> >::Type& blaze::HybridVector< typename, size_t, bool >::operator*= ( Other  rhs)
inline

Multiplication assignment operator for the multiplication between a vector and a scalar value ( $ \vec{a}*=s $).

Parameters
rhsThe right-hand side scalar value for the multiplication.
Returns
Reference to the vector.
template<typename , size_t , bool >
template<typename VT >
HybridVector<Type,N,TF>& blaze::HybridVector< typename, size_t, bool >::operator+= ( const Vector< VT, TF > &  rhs)
inline

Addition assignment operator for the addition of a vector ( $ \vec{a}+=\vec{b} $).

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

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

template<typename , size_t , bool >
template<typename VT >
HybridVector<Type,N,TF>& blaze::HybridVector< typename, size_t, bool >::operator-= ( const Vector< VT, TF > &  rhs)
inline

Subtraction assignment operator for the subtraction of a vector ( $ \vec{a}-=\vec{b} $).

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

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

template<typename , size_t , bool >
template<typename Other >
EnableIf< IsNumeric<Other>, HybridVector<Type,N,TF> >::Type& blaze::HybridVector< typename, size_t, bool >::operator/= ( Other  rhs)
inline

Division assignment operator for the division of a vector by a scalar value ( $ \vec{a}/=s $).

Parameters
rhsThe right-hand side scalar value for the division.
Returns
Reference to the vector.

Note: A division by zero is only checked by an user assert.

template<typename Type , size_t N, bool TF>
HybridVector< Type, N, TF > & blaze::HybridVector< Type, N, TF >::operator= ( const Type &  rhs)
inline

Homogenous assignment to all vector elements.

Parameters
rhsScalar value to be assigned to all vector elements.
Returns
Reference to the assigned vector.
template<typename Type , size_t N, bool TF>
HybridVector< Type, N, TF > & blaze::HybridVector< Type, N, TF >::operator= ( const HybridVector< typename, size_t, bool > &  rhs)
inline

Copy assignment operator for HybridVector.

Parameters
rhsVector to be copied.
Returns
Reference to the assigned vector.

Explicit definition of a copy assignment operator for performance reasons.

template<typename , size_t , bool >
template<typename Other , size_t M>
HybridVector<Type,N,TF>& blaze::HybridVector< typename, size_t, bool >::operator= ( const Other(&)  array[M])
inline

Array assignment to all vector elements.

Parameters
arrayM-dimensional array for the assignment.
Returns
Reference to the assigned vector.

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

const double init[2] = { 1.0, 2.0 };
v = init;

The vector is sized according to the size of the array and assigned the values of the given array. This assignment operator only works for arrays with a size smaller-or-equal than the maximum number of elements of the hybrid vector. (i.e. M<= N). The attempt to use a larger array will result in a compile time error.

template<typename , size_t , bool >
template<typename VT >
HybridVector<Type,N,TF>& blaze::HybridVector< typename, size_t, bool >::operator= ( const Vector< VT, TF > &  rhs)
inline

Assignment operator for different vectors.

Parameters
rhsVector to be copied.
Returns
Reference to the assigned vector.
Exceptions
std::invalid_argumentInvalid assignment to hybrid vector.

This constructor initializes the vector as a copy of the given vector. In case the size of the given vector is larger than N, a std::invalid_argument exception is thrown.

template<typename Type , size_t N, bool TF>
HybridVector< Type, N, TF >::Reference blaze::HybridVector< Type, N, TF >::operator[] ( size_t  index)
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
Reference to the accessed value.

In case BLAZE_USER_ASSERT() is active, this operator performs an index check.

template<typename Type , size_t N, bool TF>
HybridVector< Type, N, TF >::ConstReference blaze::HybridVector< Type, N, 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
Reference-to-const to the accessed value.

In case BLAZE_USER_ASSERT() is active, this operator performs an index check.

template<typename , bool >
VectorType& blaze::Vector< typename, bool >::operator~ ( )
inlineinherited

Conversion operator for non-constant vectors.

Returns
Reference of the actual type of the vector.
template<typename , bool >
const VectorType& blaze::Vector< typename, bool >::operator~ ( ) const
inlineinherited

Conversion operator for constant vectors.

Returns
Const reference of the actual type of the vector.
template<typename Type , size_t N, bool TF>
void blaze::HybridVector< Type, N, TF >::reset ( )
inline

Reset to the default initial values.

Returns
void
template<typename Type , size_t N, bool TF>
void blaze::HybridVector< Type, N, TF >::resize ( size_t  n,
bool  preserve = true 
)
inline

Changing the size of the vector.

Parameters
nThe new size of the vector.
preservetrue if the old values of the vector should be preserved, false if not.
Returns
void
Exceptions
std::invalid_argumentInvalid size for hybrid vector.

This function resizes the vector to the given size n. In case the given size n is larger than the maximum number of vector elements (i.e. if n > N) a std::invalid_argument exception is thrown. During this operation, all vector elements are potentially changed. In order to preserve the old vector values, the preserve flag can be set to true.
Note that in case the size of the vector is increased new vector elements are not initialized! This is illustrated by the following example, which demonstrates the resizing of a vector of size 2 to a vector of size 4. The new, uninitialized elements are marked with x:

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

template<typename , size_t , bool >
template<typename Other >
HybridVector<Type,N,TF>& blaze::HybridVector< typename, size_t, bool >::scale ( Other  scalar)
inline

Scaling of the vector by the scalar value scalar ( $ \vec{a}*=s $).

Parameters
scalarThe scalar value for the vector scaling.
Returns
Reference to the vector.
template<typename Type , size_t N, bool TF>
size_t blaze::HybridVector< Type, N, TF >::size ( ) const
inline

Returns the current size/dimension of the vector.

Returns
The size of the vector.
template<typename Type , size_t N, bool TF>
void blaze::HybridVector< Type, N, TF >::store ( size_t  index,
const IntrinsicType value 
)
inline

Aligned store of an intrinsic element of the vector.

Parameters
indexAccess index. The index must be smaller than the number of vector elements.
valueThe intrinsic element to be stored.
Returns
void

This function performs an aligned store of a specific intrinsic element of the dense vector. The index must be smaller than the number of vector elements and it must be a multiple of the number of values inside the intrinsic 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.

template<typename Type , size_t N, bool TF>
void blaze::HybridVector< Type, N, TF >::storeu ( size_t  index,
const IntrinsicType value 
)
inline

Unaligned store of an intrinsic element of the vector.

Parameters
indexAccess index. The index must be smaller than the number of vector elements.
valueThe intrinsic element to be stored.
Returns
void

This function performs an unaligned store of a specific intrinsic element of the dense vector. The index must be smaller than the number of vector elements and it must be a multiple of the number of values inside the intrinsic 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.

template<typename Type , size_t N, bool TF>
void blaze::HybridVector< Type, N, TF >::stream ( size_t  index,
const IntrinsicType value 
)
inline

Aligned, non-temporal store of an intrinsic element of the vector.

Parameters
indexAccess index. The index must be smaller than the number of vector elements.
valueThe intrinsic element to be stored.
Returns
void

This function performs an aligned, non-temporal store of a specific intrinsic element of the dense vector. The index must be smaller than the number of vector elements and it must be a multiple of the number of values inside the intrinsic 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.

template<typename Type , size_t N, bool TF>
template<typename VT >
void blaze::HybridVector< Type, N, TF >::subAssign ( const SparseVector< VT, TF > &  rhs)
inline

Default implementation of the subtraction assignment of a sparse vector.

Parameters
rhsThe right-hand side sparse vector 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.

template<typename , size_t , bool >
template<typename VT >
DisableIf< typename HybridVector<Type,N,TF>::BLAZE_TEMPLATE VectorizedSubAssign<VT> >::Type blaze::HybridVector< typename, size_t, bool >::subAssign ( const DenseVector< VT, TF > &  rhs)
inline

Default implementation of the subtraction assignment of a dense vector.

Parameters
rhsThe right-hand side dense vector 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.

template<typename , size_t , bool >
template<typename VT >
EnableIf< typename HybridVector<Type,N,TF>::BLAZE_TEMPLATE VectorizedSubAssign<VT> >::Type blaze::HybridVector< typename, size_t, bool >::subAssign ( const DenseVector< VT, TF > &  rhs)
inline

Intrinsic optimized implementation of the subtraction assignment of a dense vector.

Parameters
rhsThe right-hand side dense vector 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.

template<typename Type , size_t N, bool TF>
void blaze::HybridVector< Type, N, TF >::swap ( HybridVector< typename, size_t, bool > &  v)
inline

Swapping the contents of two hybrid vectors.

Parameters
vThe vector to be swapped.
Returns
void
Exceptions
no-throwguarantee.

Member Data Documentation

template<typename , size_t , bool >
AlignedArray<Type,NN> blaze::HybridVector< typename, size_t, bool >::v_
private

The statically allocated vector elements.

Access to the vector values is gained via the subscript operator. The order of the elements is

\[\left(\begin{array}{*{4}{c}} 0 & 1 & \cdots & N-1 \\ \end{array}\right)\]


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