A model of a tensor. More...

#include <iolink/view/tensor/TensorView.h>

Inheritance diagram for iolink::TensorView:

Public Member Functions
virtual	~TensorView ()=default
	Virtual destructor to support polymorphism.

virtual TensorCapabilitySet	capabilities () const =0
	Get the capabilities of the tensor.

bool	support (TensorCapabilitySet capabilities) const
	Shortcut to check if capabilities are supported by the tensor.

virtual const VectorXu64 &	shape () const =0
	Get the shape of the tensor.

size_t	dimensionCount () const
	Shortcut to get the number of dimensions of the tensor.

virtual DataType	dtype () const =0
	Get the data type of the tensor.

std::string	toString () const
	Return a string representation.

virtual void	read (const RegionXu64 &region, void *dst) const
	Read data from a given region into a destination buffer.

template<typename T >
T	at (const VectorXu64 &index) const
	Get data from a specific index in the tensor.

virtual void	write (const RegionXu64 &region, const void *src)
	Write data from a source buffer into a given region of the tensor.

template<typename T >
void	setAt (const VectorXu64 &index, T value)
	Set a value at a specific index in the tensor.

virtual void	reshape (const VectorXu64 &shape, DataType dtype)

virtual const Indexer &	indexer () const
	Get the indexer for the tensor underlying buffer.

virtual size_t	bufferSize () const
	Get the size in bytes of the underlying buffer.

virtual void *	buffer ()
	Get a mutable pointer to the underlying memory buffer.

virtual const void *	buffer () const
	Get an immutable pointer to the underlying memory buffer.

Public Member Functions inherited from iolink::View
virtual std::shared_ptr< Extension >	extension (size_t id)
	Get an extension of the view, corresponding to the given identifier.

template<typename T >
std::shared_ptr< T >	extension ()
	Get an extension of the view, corresponding to the given type.

std::shared_ptr< ViewOriginExtension >	viewOrigin ()
	Shortcut to access to ViewOrigin extension.

Detailed Description

A model of a tensor.

Tensors are simple multidimensional arrays that can be used to represent data in a structured way. They can be used to represent images, audio, text, or any other type of data that can be represented as a multidimensional array.

Member Function Documentation

◆ at()

template<typename T >

T iolink::TensorView::at ( const VectorXu64 & index ) const

inline

Get data from a specific index in the tensor.

Template Parameters

T	The type of the data to read from the tensor.

Parameters

index The index of the tensor element to read.

Returns: The value at the specified index

Warning: No verification is made to check if the template type suits the tensor's data type, so use this method only when you're absolutely sure that the type matches the tensor's data type.

◆ buffer() [1/2]

virtual void * iolink::TensorView::buffer ( )

virtual

Get a mutable pointer to the underlying memory buffer.

Returns: A pointer to the tensor's underlying memory buffer, allowing direct access to the data.

◆ buffer() [2/2]

virtual const void * iolink::TensorView::buffer ( ) const

virtual

Get an immutable pointer to the underlying memory buffer.

Returns: A pointer to the tensor's underlying memory buffer, allowing read-only access.

◆ bufferSize()

virtual size_t iolink::TensorView::bufferSize ( ) const

virtual

Get the size in bytes of the underlying buffer.

Returns: The size in bytes of the tensor's underlying memory buffer.

◆ capabilities()

virtual TensorCapabilitySet iolink::TensorView::capabilities ( ) const

pure virtual

Get the capabilities of the tensor.

Returns: A set of capabilities that the tensor supports.

See also: TensorCapability; TensorCapabilitySet

◆ dimensionCount()

size_t iolink::TensorView::dimensionCount ( ) const

inline

Shortcut to get the number of dimensions of the tensor.

Returns: The number of dimensions of the tensor, which is equal to the size of the shape vector.

◆ dtype()

virtual DataType iolink::TensorView::dtype ( ) const

pure virtual

Get the data type of the tensor.

Returns: The data type of the tensor, which describes the type of elements stored in the tensor.

◆ indexer()

virtual const Indexer & iolink::TensorView::indexer ( ) const

virtual

Get the indexer for the tensor underlying buffer.

Returns: An Indexer that describes the memory layout of the tensor's data.

◆ read()

virtual void iolink::TensorView::read	(	const RegionXu64 &	region,
		void *	dst
	)		const

virtual

Read data from a given region into a destination buffer.

Parameters

region	The region of the tensor to read from.
dst	The destination buffer where the read data will be copied.

Exceptions

NotImplemented	If the implementation does not support reading.
InvalidArgument	If the region is not compatible with the tensor's shape.

◆ setAt()

template<typename T >

void iolink::TensorView::setAt	(	const VectorXu64 &	index,
		T	value
	)

inline

Set a value at a specific index in the tensor.

Template Parameters

T	The type of the value to set in the tensor.

Parameters

index	The index of the tensor element to set.
value	The value to set at the specified index.

Warning: No verification is made to check if the template type suits the tensor's data type, so use this method only when you're absolutely sure that the type matches the tensor's data type.

◆ shape()

virtual const VectorXu64 & iolink::TensorView::shape ( ) const

pure virtual

Get the shape of the tensor.

Returns: A vector representing the shape of the tensor, where each element corresponds to the size of the tensor along that dimension.

◆ support()

bool iolink::TensorView::support ( TensorCapabilitySet capabilities ) const

inline

Shortcut to check if capabilities are supported by the tensor.

Parameters

capabilities The set of capabilities to check.

Returns: true if the tensor supports the given capabilities, false otherwise.

◆ write()

virtual void iolink::TensorView::write	(	const RegionXu64 &	region,
		const void *	src
	)

virtual

Write data from a source buffer into a given region of the tensor.

Parameters

region	The region of the tensor to write to.
src	The source buffer containing the data that will be written.

Exceptions

NotImplemented	If the implementation does not support writing.
InvalidArgument	If the region is not compatible with the tensor's shape.

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

iolink/public/include/iolink/view/tensor/TensorView.h

Public Member Functions

Detailed Description

Member Function Documentation

◆ at()

◆ buffer() [1/2]

◆ buffer() [2/2]

◆ bufferSize()

◆ capabilities()

◆ dimensionCount()

◆ dtype()

◆ indexer()

◆ read()

◆ setAt()

◆ shape()

◆ support()

◆ write()