#include <TensorContraction.h>

Public Member Functions
template<typename Index , typename Scalar >
EIGEN_ALWAYS_INLINE void	operator() (const internal::blas_data_mapper< Scalar, Index, ColMajor > &output_mapper, const TensorContractionParams &params, Index i, Index j, Index num_rows, Index num_cols) const

Member Function Documentation

◆ operator()()

template<typename Index , typename Scalar >

EIGEN_ALWAYS_INLINE void Eigen::NoOpOutputKernel::operator()	(	const internal::blas_data_mapper< Scalar, Index, ColMajor > &	output_mapper,
		const TensorContractionParams &	params,
		Index	i,
		Index	j,
		Index	num_rows,
		Index	num_cols
	)		const

inline

Tensor contraction evaluator calls this kernel after finishing each block of output matrix. Output blocks belong to the 2-dimensional output tensor.

TensorContractionParams contains contraction dimensions information required to map output 2-d space into the expected output tensor space (potentially higher dimensional).

Parameters

[in]	output_mapper	Access to output tensor memory
[in]	params	Tensor contraction parameters
[in]	i	Index of a first row available through output_mapper
[in]	j	Index of a first column available through output_mapper
[in]	num_rows	Number of available rows
[in]	num_cols	Number of available columns

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

core/util/algorithms/eigen-3.4.0/unsupported/Eigen/CXX11/src/Tensor/TensorContraction.h

Public Member Functions

Member Function Documentation

◆ operator()()