TMVA::DNN::TTensorBatch< Architecture_t > Class Template Reference

template<typename Architecture_t>
class TMVA::DNN::TTensorBatch< Architecture_t >

TTensorBatch.

Class representing training batches consisting of a vector of matrices as input data and a matrix of output data. The input and output data can be accessed using the GetInput() and GetOutput() member functions.

Template Parameters

Architecture_t

The underlying architecture.

Definition at line 59 of file TensorDataLoader.h.

Public Types
using	Matrix_t = typename Architecture_t::Matrix_t
using	Tensor_t = typename Architecture_t::Tensor_t

Public Member Functions
	TTensorBatch (const TTensorBatch &)=default
	TTensorBatch (Tensor_t &, Matrix_t &, Matrix_t &)
	TTensorBatch (TTensorBatch &&)=default
Tensor_t &	GetInput ()
	Return the tensor representing the input data.
Matrix_t &	GetOutput ()
	Return the matrix representing the output data.
Matrix_t &	GetWeights ()
	Return the matrix holding the event weights.
TTensorBatch &	operator= (const TTensorBatch &)=default
TTensorBatch &	operator= (TTensorBatch &&)=default

Private Attributes
Tensor_t	fInputTensor
	The input tensor batch, one matrix one input.
Matrix_t	fOutputMatrix
	The output matrix representing the ground truth.
Matrix_t	fWeightMatrix
	The event/example weights.

#include <TMVA/DNN/TensorDataLoader.h>

Member Typedef Documentation

Matrix_t

template<typename Architecture_t >

using TMVA::DNN::TTensorBatch< Architecture_t >::Matrix_t = typename Architecture_t::Matrix_t

Definition at line 61 of file TensorDataLoader.h.

Tensor_t

template<typename Architecture_t >

using TMVA::DNN::TTensorBatch< Architecture_t >::Tensor_t = typename Architecture_t::Tensor_t

Definition at line 62 of file TensorDataLoader.h.

Constructor & Destructor Documentation

TTensorBatch() [1/3]

template<typename Architecture_t >

TMVA::DNN::TTensorBatch< Architecture_t >::TTensorBatch	(	Tensor_t &	inputTensor,
		Matrix_t &	outputMatrix,
		Matrix_t &	weightMatrix
	)

Definition at line 198 of file TensorDataLoader.h.

TTensorBatch() [2/3]

template<typename Architecture_t >

TMVA::DNN::TTensorBatch< Architecture_t >::TTensorBatch ( const TTensorBatch< Architecture_t > &  )

default

TTensorBatch() [3/3]

template<typename Architecture_t >

TMVA::DNN::TTensorBatch< Architecture_t >::TTensorBatch ( TTensorBatch< Architecture_t > &&  )

default

Member Function Documentation

GetInput()

template<typename Architecture_t >

Tensor_t & TMVA::DNN::TTensorBatch< Architecture_t >::GetInput ( )

inline

Return the tensor representing the input data.

Definition at line 77 of file TensorDataLoader.h.

GetOutput()

template<typename Architecture_t >

Matrix_t & TMVA::DNN::TTensorBatch< Architecture_t >::GetOutput ( )

inline

Return the matrix representing the output data.

Definition at line 79 of file TensorDataLoader.h.

GetWeights()

template<typename Architecture_t >

Matrix_t & TMVA::DNN::TTensorBatch< Architecture_t >::GetWeights ( )

inline

Return the matrix holding the event weights.

Definition at line 81 of file TensorDataLoader.h.

operator=() [1/2]

template<typename Architecture_t >

TTensorBatch & TMVA::DNN::TTensorBatch< Architecture_t >::operator= ( const TTensorBatch< Architecture_t > &  )

default

operator=() [2/2]

template<typename Architecture_t >

TTensorBatch & TMVA::DNN::TTensorBatch< Architecture_t >::operator= ( TTensorBatch< Architecture_t > &&  )

default

Member Data Documentation

fInputTensor

template<typename Architecture_t >

Tensor_t TMVA::DNN::TTensorBatch< Architecture_t >::fInputTensor

private

The input tensor batch, one matrix one input.

Definition at line 65 of file TensorDataLoader.h.

fOutputMatrix

template<typename Architecture_t >

Matrix_t TMVA::DNN::TTensorBatch< Architecture_t >::fOutputMatrix

private

The output matrix representing the ground truth.

Definition at line 66 of file TensorDataLoader.h.

fWeightMatrix

template<typename Architecture_t >

Matrix_t TMVA::DNN::TTensorBatch< Architecture_t >::fWeightMatrix

private

The event/example weights.

Definition at line 67 of file TensorDataLoader.h.