tensor_utilities.h

 
#ifndef TENSORUTILITIES_H
#define TENSORUTILITIES_H
 
// System includes
 
#include <algorithm>
#include <fstream>
#include <iostream>
#include <limits>
#include <math.h>
#include <vector>
#include <numeric>
 
// OpenNN includes
 
#include "config.h"
 
#include "../eigen/Eigen/Dense"
 
using namespace std;
using namespace Eigen;
 
namespace OpenNN
{
 
void initialize_sequential(Tensor<type, 1>&);
 
void multiply_rows(Tensor<type, 2>&, const Tensor<type, 1>&);
void divide_columns(Tensor<type, 2>&, const Tensor<type, 1>&);
 
bool is_zero(const Tensor<type, 1>&);
bool is_zero(const Tensor<type,1>&, const type&);
bool is_constant(const Tensor<type, 1>&);
 
bool is_equal(const Tensor<type, 2>&, const type&, const type& = type(0));
 
bool are_equal(const Tensor<type, 1>&, const Tensor<type, 1>&, const type& = type(0));
bool are_equal(const Tensor<type, 2>&, const Tensor<type, 2>&, const type& = type(0));
 
bool is_false(const Tensor<bool, 1>&);
 
bool is_binary(const Tensor<type, 2>&);
 
void save_csv(const Tensor<type,2>&, const string&);
 
Tensor<Index, 1> calculate_rank_greater(const Tensor<type, 1>&);
Tensor<Index, 1> calculate_rank_less(const Tensor<type, 1>&);
 
void scrub_missing_values(Tensor<type, 2>&, const type&);
 
type l1_norm(const ThreadPoolDevice*, const Tensor<type, 1>&);
void l1_norm_gradient(const ThreadPoolDevice*, const Tensor<type, 1>&, Tensor<type, 1>&);
void l1_norm_hessian(const ThreadPoolDevice*, const Tensor<type, 1>&, Tensor<type, 2>&);
 
type l2_norm(const ThreadPoolDevice*, const Tensor<type, 1>&);
void l2_norm_gradient(const ThreadPoolDevice*, const Tensor<type, 1>&, Tensor<type, 1>&);
void l2_norm_hessian(const ThreadPoolDevice*, const Tensor<type, 1>&, Tensor<type, 2>&);
 
Tensor<type, 2> kronecker_product(const Tensor<type, 1>&, const Tensor<type, 1>&);
 
void sum_diagonal(Tensor<type, 2>&, const type&);
 
Tensor<type, 1> perform_Householder_QR_decomposition(const Tensor<type, 2>&, const Tensor<type, 1>&);
 
void fill_submatrix(const Tensor<type, 2>& matrix,
          const Tensor<Index, 1>& rows_indices,
          const Tensor<Index, 1>& columns_indices, type* submatrix);
 
Index count_NAN(const Tensor<type, 1>&);
 
void check_size(const Tensor<type, 1>&, const Index&, const string&);
 
void check_dimensions(const Tensor<type, 2>&, const Index&, const Index&, const string&);
 
void check_columns_number(const Tensor<type, 2>&, const Index&, const string&);
 
bool is_less_than(const Tensor<type, 1>&, const type&);
 
 
Tensor<type, 2> assemble_vector_vector(const Tensor<type, 1>&, const Tensor<type, 1>&);
Tensor<type, 2> assemble_vector_matrix(const Tensor<type, 1>&, const Tensor<type, 2>&);
Tensor<type, 2> assemble_matrix_vector(const Tensor<type, 2>&, const Tensor<type, 1>&);
Tensor<type, 2> assemble_matrix_matrix(const Tensor<type, 2>&, const Tensor<type, 2>&);
 
}
 
#endif