C++ API Reference for Intel® Data Analytics Acceleration Library 2020 Update 1

List of all members
SubtensorDescriptor< DataType > Class Template Reference

Class with descriptor of the subtensor retrieved from Tensor getSubTensor function. More...

Class Declaration

template<typename DataType>
class daal::data_management::interface1::SubtensorDescriptor< DataType >

Base class that manages buffer memory for read/write operations required by tensors.

Deprecated:
This item will be removed in a future release.

Constructor & Destructor Documentation

Constructor for an empty subtensor descriptor

Member Function Documentation

void freeBuffer ( )
inlineprotected

Frees the buffer

size_t* getFixedDimNums ( ) const
inline

Gets values at which dimensions are fixed

Returns
An array of values
size_t getFixedDims ( ) const
inline

Gets the number of first dimension with fixed values

Returns
The number of first dimension with fixed values
bool getInplaceFlag ( ) const
inline

Returns subtensor inplace flag

Returns
Inplace flag
const TensorOffsetLayout* getLayout ( ) const
inline

Returns subtensor layout

Returns
Layout
size_t getNumberOfDims ( ) const
inline

Returns the number of dimensions of the subtensor

Returns
Number of dimensions of the subtensor
DataType* getPtr ( ) const
inline

Gets a pointer to the buffer for the subtensor

Returns
Pointer to the buffer for the subtensor
size_t getRangeDimIdx ( ) const
inline

Gets values for the next dimension after fixed to get data from

Returns
Value of the dimension
size_t getRangeDimNum ( ) const
inline

Gets range for dimension values to get data from

Returns
Range for dimension values to get data from
size_t getRWFlag ( ) const
inline

Returns a flag specifying read/write access to the subtensor

Returns
Flag specifying read/write access to the subtensor
services::SharedPtr<DataType> getSharedPtr ( ) const
inline

Gets a pointer to the buffer for the subtensor

Returns
Pointer to the buffer for the subtensor
size_t getSize ( ) const
inline

Returns the full size of the subtensor in number of elements

Returns
The full size of the subtensor in number of elements
size_t* getSubtensorDimSizes ( ) const
inline

Returns the array with sizes of dimensions of the subtensor

Returns
Array with sizes of the dimensions of the subtensor
void reset ( )
inline

Reset internal values and pointers to zero values

bool resizeBuffer ( )
inline

Returns true if memory of (_subtensorSize) size is allocated successfully

bool saveOffsetLayout ( const TensorOffsetLayout layout)
inline

Saves subtensor offset layout

Parameters
[in]layoutoffset layout
bool saveOffsetLayoutCopy ( const TensorOffsetLayout layout)
inline

Saves subtensor offset layout copy

Parameters
[in]layoutoffset layout
size_t setDetails ( size_t  tensorNDims,
const size_t *  tensorDimNums,
size_t  nFixedDims,
const size_t *  fixedDimNums,
size_t  rangeDimIdx,
size_t  rangeDimNum,
int  rwFlag 
)
inline

Sets subtensor parameters

Parameters
[in]tensorNDimsThe number of dimensions with fixed values
[in]tensorDimNumsValues of the tensor dmensions
[in]nFixedDimsThe number of first dimension with fixed values
[in]fixedDimNumsValues at which dimensions are fixed
[in]rangeDimIdxValues for the next dimension after fixed to get data from
[in]rangeDimNumRange for dimension values to get data from
[in]rwFlagFlag specifying read/write access to the subtensor
Returns
Subtensor size
void setPtr ( DataType *  ptr)
inline

Sets data pointer to use for in-place calculation

Parameters
[in]ptrpointer
Deprecated:
This item will be removed in a future release.

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

For more complete information about compiler optimizations, see our Optimization Notice.