The primary data structure in NEML2 for working with labeled tensor views. More...

Detailed Description

template<class Derived, TorchSize D>
class neml2::LabeledTensor< Derived, D >

The primary data structure in NEML2 for working with labeled tensor views.

Each LabeledTensor consists of one BatchTensor and one or more LabeledAxis. The LabeledTensor<D> is templated on the base dimension \(D\). LabeledTensor handles the creation, modification, and accessing of labeled tensors.

Template Parameters

D	The number of base dimensions

#include <LabeledTensor.h>

Classes
struct	variable_type
	Template setup for appropriate variable types. More...

Public Member Functions
	LabeledTensor ()=default
	Default constructor.

	LabeledTensor (const torch::Tensor &tensor, TorchSize batch_dim, const std::vector< const LabeledAxis * > &axes)
	Construct from a Tensor with batch dim and vector of `LabeledAxis`

	LabeledTensor (const BatchTensor &tensor, const std::vector< const LabeledAxis * > &axes)
	Construct from a BatchTensor with vector of `LabeledAxis`

	LabeledTensor (const Derived &other)
	Copy constructor.

void	operator= (const Derived &other)
	Assignment operator.

	operator BatchTensor () const
	A potentially dangerous implicit conversion.

	operator torch::Tensor () const
	A potentially dangerous implicit conversion.

Derived	clone (torch::MemoryFormat memory_format=torch::MemoryFormat::Contiguous) const
	Clone this LabeledTensor.

template<typename T >
void	copy_ (const T &other)
	Copy the value from another tensor.

Derived	detach () const
	Return a copy without gradient graphs.

void	detach_ ()
	Detach from gradient graphs.

void	zero_ ()
	Zero out this tensor.

torch::TensorOptions	options () const
	Get the tensor options.

TorchSize	batch_dim () const
	Return the number of batch dimensions.

TorchSize	base_dim () const
	Return the number of base dimensions.

TorchShapeRef	batch_sizes () const
	Return the batch size.

TorchShapeRef	base_sizes () const
	Return the base size.

const std::vector< const LabeledAxis * > &	axes () const
	Get all the labeled axes.

const LabeledAxis &	axis (TorchSize i=0) const
	Get a specific labeled axis.

template<typename... S>
TorchSlice	slice_indices (S &&... names) const
	How to slice the tensor given the names on each axis.

TorchShapeRef	storage_size () const
	The shape of the entire LabeledTensor.

template<typename... S>
TorchShape	storage_size (S &&... names) const
	The shape of a sub-block specified by the names on each dimension.

template<typename... S>
BatchTensor	operator() (S &&... names) const

Derived	slice (TorchSize i, const std::string &name) const
	Slice the tensor on the given dimension by a single variable or sub-axis.

template<typename... S>
Derived	block (S &&... names) const
	Get the sub-block labeled by the given sub-axis names.

Derived	batch_index (TorchSlice indices) const
	Get a batch.

void	batch_index_put (TorchSlice indices, const torch::Tensor &other)
	Set a index sliced on the batch dimensions to a value.

BatchTensor	base_index (TorchSlice indices) const
	Return an index sliced on the batch dimensions.

void	base_index_put (TorchSlice indices, const torch::Tensor &other)
	Set a index sliced on the batch dimensions to a value.

template<typename T , typename... S>
variable_type< T >::type	get (S &&... names) const
	Get and interpret the view as an object.

template<typename T , typename... S>
variable_type< T >::type	get_list (S &&... names) const
	Get and interpret the view as a list of objects.

template<typename T , typename... S>
void	set (const BatchTensorBase< T > &value, S &&... names)
	Set and interpret the input as an object.

template<typename T , typename... S>
void	set_list (const BatchTensorBase< T > &value, S &&... names)
	Set and interpret the input as a list of objects.

Derived	operator- () const
	Negation.

Derived	to (const torch::TensorOptions &options) const
	Change tensor options.


const BatchTensor &	tensor () const

BatchTensor &	tensor ()

Static Public Member Functions
static Derived	empty (TorchShapeRef batch_shape, const std::vector< const LabeledAxis * > &axes, const torch::TensorOptions &options=default_tensor_options())
	Setup new empty storage.

static Derived	empty_like (const Derived &other)
	Setup new empty storage like another LabeledTensor.

static Derived	zeros (TorchShapeRef batch_shape, const std::vector< const LabeledAxis * > &axes, const torch::TensorOptions &options=default_tensor_options())
	Setup new storage with zeros.

static Derived	zeros_like (const Derived &other)
	Setup new storage with zeros like another LabeledTensor.

Protected Attributes
BatchTensor	_tensor
	The tensor.

std::vector< const LabeledAxis * >	_axes
	The labeled axes of this tensor.

Constructor & Destructor Documentation

◆ LabeledTensor() [1/4]

template<class Derived , TorchSize D>

LabeledTensor ( )

default

Default constructor.

◆ LabeledTensor() [2/4]

template<class Derived , TorchSize D>

LabeledTensor	(	const torch::Tensor &	tensor,
		TorchSize	batch_dim,
		const std::vector< const LabeledAxis * > &	axes )

Construct from a Tensor with batch dim and vector of LabeledAxis

◆ LabeledTensor() [3/4]

template<class Derived , TorchSize D>

LabeledTensor	(	const BatchTensor &	tensor,
		const std::vector< const LabeledAxis * > &	axes )

Construct from a BatchTensor with vector of LabeledAxis

◆ LabeledTensor() [4/4]

template<class Derived , TorchSize D>

LabeledTensor ( const Derived & other )

Copy constructor.

Member Function Documentation

◆ axes()

template<class Derived , TorchSize D>

const std::vector< const LabeledAxis * > & axes ( ) const

inline

Get all the labeled axes.

◆ axis()

template<class Derived , TorchSize D>

const LabeledAxis & axis ( TorchSize i = 0 ) const

inline

Get a specific labeled axis.

◆ base_dim()

template<class Derived , TorchSize D>

TorchSize base_dim ( ) const

Return the number of base dimensions.

◆ base_index()

template<class Derived , TorchSize D>

BatchTensor base_index ( TorchSlice indices ) const

Return an index sliced on the batch dimensions.

◆ base_index_put()

template<class Derived , TorchSize D>

void base_index_put	(	TorchSlice	indices,
		const torch::Tensor &	other )

Set a index sliced on the batch dimensions to a value.

◆ base_sizes()

template<class Derived , TorchSize D>

TorchShapeRef base_sizes ( ) const

Return the base size.

◆ batch_dim()

template<class Derived , TorchSize D>

TorchSize batch_dim ( ) const

Return the number of batch dimensions.

◆ batch_index()

template<class Derived , TorchSize D>

Derived batch_index ( TorchSlice indices ) const

Get a batch.

◆ batch_index_put()

template<class Derived , TorchSize D>

void batch_index_put	(	TorchSlice	indices,
		const torch::Tensor &	other )

Set a index sliced on the batch dimensions to a value.

◆ batch_sizes()

template<class Derived , TorchSize D>

TorchShapeRef batch_sizes ( ) const

Return the batch size.

◆ block()

template<class Derived , TorchSize D>

template<typename... S>

Derived block ( S &&... names ) const

Get the sub-block labeled by the given sub-axis names.

◆ clone()

template<class Derived , TorchSize D>

Derived clone ( torch::MemoryFormat memory_format = torch::MemoryFormat::Contiguous ) const

Clone this LabeledTensor.

◆ copy_()

template<class Derived , TorchSize D>

template<typename T >

void copy_ ( const T & other )

Copy the value from another tensor.

◆ detach()

template<class Derived , TorchSize D>

Derived detach ( ) const

Return a copy without gradient graphs.

◆ detach_()

template<class Derived , TorchSize D>

void detach_ ( )

Detach from gradient graphs.

◆ empty()

template<class Derived , TorchSize D>

Derived empty	(	TorchShapeRef	batch_shape,
		const std::vector< const LabeledAxis * > &	axes,
		const torch::TensorOptions &	options = default_tensor_options() )

static

Setup new empty storage.

◆ empty_like()

template<class Derived , TorchSize D>

Derived empty_like ( const Derived & other )

static

Setup new empty storage like another LabeledTensor.

◆ get()

template<class Derived , TorchSize D>

template<typename T , typename... S>

variable_type< T >::type get ( S &&... names ) const

inline

Get and interpret the view as an object.

◆ get_list()

template<class Derived , TorchSize D>

template<typename T , typename... S>

variable_type< T >::type get_list ( S &&... names ) const

inline

Get and interpret the view as a list of objects.

◆ operator BatchTensor()

template<class Derived , TorchSize D>

operator BatchTensor ( ) const

A potentially dangerous implicit conversion.

◆ operator torch::Tensor()

template<class Derived , TorchSize D>

operator torch::Tensor ( ) const

A potentially dangerous implicit conversion.

◆ operator()()

template<class Derived , TorchSize D>

template<typename... S>

BatchTensor operator() ( S &&... names ) const

Return a labeled view into the tensor. No reshaping is done.

◆ operator-()

template<class Derived , TorchSize D>

Derived operator- ( ) const

Negation.

◆ operator=()

template<class Derived , TorchSize D>

void operator= ( const Derived & other )

Assignment operator.

◆ options()

template<class Derived , TorchSize D>

torch::TensorOptions options ( ) const

inline

Get the tensor options.

◆ set()

template<class Derived , TorchSize D>

template<typename T , typename... S>

void set	(	const BatchTensorBase< T > &	value,
		S &&...	names )

inline

Set and interpret the input as an object.

◆ set_list()

template<class Derived , TorchSize D>

template<typename T , typename... S>

void set_list	(	const BatchTensorBase< T > &	value,
		S &&...	names )

inline

Set and interpret the input as a list of objects.

◆ slice()

template<class Derived , TorchSize D>

Derived slice	(	TorchSize	i,
		const std::string &	name ) const

Slice the tensor on the given dimension by a single variable or sub-axis.

◆ slice_indices()

template<class Derived , TorchSize D>

template<typename... S>

TorchSlice slice_indices ( S &&... names ) const

How to slice the tensor given the names on each axis.

◆ storage_size() [1/2]

template<class Derived , TorchSize D>

TorchShapeRef storage_size ( ) const

The shape of the entire LabeledTensor.

◆ storage_size() [2/2]

template<class Derived , TorchSize D>

template<typename... S>

TorchShape storage_size ( S &&... names ) const

The shape of a sub-block specified by the names on each dimension.

◆ tensor() [1/2]

template<class Derived , TorchSize D>

BatchTensor & tensor ( )

inline

◆ tensor() [2/2]

template<class Derived , TorchSize D>

const BatchTensor & tensor ( ) const

inline

Get the underlying tensor

◆ to()

template<class Derived , TorchSize D>

Derived to ( const torch::TensorOptions & options ) const

Change tensor options.

◆ zero_()

template<class Derived , TorchSize D>

void zero_ ( )

Zero out this tensor.

◆ zeros()

template<class Derived , TorchSize D>

Derived zeros	(	TorchShapeRef	batch_shape,
		const std::vector< const LabeledAxis * > &	axes,
		const torch::TensorOptions &	options = default_tensor_options() )

static

Setup new storage with zeros.

◆ zeros_like()

template<class Derived , TorchSize D>

Derived zeros_like ( const Derived & other )

static

Setup new storage with zeros like another LabeledTensor.

Member Data Documentation

◆ _axes

template<class Derived , TorchSize D>

std::vector<const LabeledAxis *> _axes

protected

The labeled axes of this tensor.

◆ _tensor

template<class Derived , TorchSize D>

BatchTensor _tensor

protected

The tensor.

Detailed Description

Classes

Public Member Functions

Static Public Member Functions

Protected Attributes

Constructor & Destructor Documentation

◆ LabeledTensor() [1/4]

◆ LabeledTensor() [2/4]

◆ LabeledTensor() [3/4]

◆ LabeledTensor() [4/4]

Member Function Documentation

◆ axes()

◆ axis()

◆ base_dim()

◆ base_index()

◆ base_index_put()

◆ base_sizes()

◆ batch_dim()

◆ batch_index()

◆ batch_index_put()

◆ batch_sizes()

◆ block()

◆ clone()

◆ copy_()

◆ detach()

◆ detach_()

◆ empty()

◆ empty_like()

◆ get()

◆ get_list()

◆ operator BatchTensor()

◆ operator torch::Tensor()

◆ operator()()

◆ operator-()

◆ operator=()

◆ options()

◆ set()

◆ set_list()

◆ slice()

◆ slice_indices()

◆ storage_size() [1/2]

◆ storage_size() [2/2]

◆ tensor() [1/2]

◆ tensor() [2/2]

◆ to()

◆ zero_()

◆ zeros()

◆ zeros_like()

Member Data Documentation

◆ _axes

◆ _tensor