C++ API Reference for Intel® Data Analytics Acceleration Library 2019 Update 5

Classes | Enumerations
Numeric Tensors

Contains classes for a data management component responsible for representation of data in the n-dimensions numeric format. More...

Classes

class  HomogenTensor< DataType >
 Class that provides methods to access data stored as a contiguous array of homogeneous data in rows-major format. More...
 
singleton  SubtensorDescriptor< DataType >
 Class with descriptor of the subtensor retrieved from Tensor getSubTensor function. More...
 
class  TensorIface
 Abstract interface class for a data management component responsible for representation of data in the numeric format. This class declares the most general methods for data access. More...
 
class  TensorLayoutIface
 Abstract interface class for a data management component responsible for representation of data layout in the tensor. This class declares the most general methods for data access. More...
 
class  TensorLayout
 Class for a data management component responsible for representation of data layout in the tensor. This class implements the most general methods for data layout. More...
 
class  TensorOffsetLayout
 Class for a data management component responsible for representation of data layout in the HomogenTensor. More...
 
class  DenseTensorIface
 Abstract interface class for a data management component responsible for accessing data in the numeric format. This class declares specific methods to access Tensor data in a dense homogeneous form. More...
 
class  Tensor
 Class for a data management component responsible for representation of data in the n-dimensions numeric format. This class implements the most general methods for data access. More...
 

Enumerations

enum  MemoryStatus { notAllocated = 0, userAllocated = 1, internallyAllocated = 2 }
 Enumeration to specify the status of memory related to the Numeric Table. More...
 
enum  AllocationFlag { doNotAllocate = 0, notAllocate = 0, doAllocate = 1 }
 Enumeration to specify whether the Numeric Table must allocate memory. More...
 

Enumeration Type Documentation

enum AllocationFlag

Enumerator
doNotAllocate 

Memory will not be allocated by Tensor

notAllocate 

Memory will not be allocated by Tensor

Deprecated:
This item will be removed in a future release. Use doNotAllocate instead.
doAllocate 

Memory will be allocated by Tensor when needed

enum MemoryStatus

Enumerator
notAllocated 

No memory allocated

userAllocated 

Memory allocated on user side

internallyAllocated 

Memory allocated and managed by Tensor

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