ldaplusplus::numpy_format::NumpyOutput< Scalar > Class Template Reference

#include <NumpyFormat.hpp>

Public Member Functions
	NumpyOutput (const Scalar *data, std::vector< size_t > shape, bool fortran)
template<int Rows, int Cols, int Options>
	NumpyOutput (const Eigen::Matrix< Scalar, Rows, Cols, Options > &matrix)
std::string	dtype () const
const char *	data () const
const std::vector< size_t > &	shape () const
bool	fortran_contiguous () const

Friends
std::ostream &	operator<< (std::ostream &os, const NumpyOutput &data)

Detailed Description

template<typename Scalar>
class ldaplusplus::numpy_format::NumpyOutput< Scalar >

NumpyOutput provides an easy way to serialize a contiguous array using the numpy format.

Examples:

MatrixXf x = MatrixXf::Random(10, 20);
std::cout << numpy_format::NumpyOutput<float>(x);

Constructor & Destructor Documentation

template<typename Scalar >

ldaplusplus::numpy_format::NumpyOutput< Scalar >::NumpyOutput ( const Scalar *  data, std::vector< size_t >  shape, bool  fortran  )

inline

Construct a NumpyOutput object from raw data.

Parameters

data	The contiguous in memory data
shape	The logical array shape of the data
fortran	Wether the data is row major (false) or column major (true) order

template<typename Scalar >

template<int Rows, int Cols, int Options>

ldaplusplus::numpy_format::NumpyOutput< Scalar >::NumpyOutput ( const Eigen::Matrix< Scalar, Rows, Cols, Options > &  matrix )

inline

Construct a NumpyOutput object from any Eigen Matrix.

Parameters

matrix

The matrix containing the data

Member Function Documentation

template<typename Scalar >

const char* ldaplusplus::numpy_format::NumpyOutput< Scalar >::data ( ) const

inline

Return the data as a constant character array.

template<typename Scalar >

std::string ldaplusplus::numpy_format::NumpyOutput< Scalar >::dtype ( ) const

inline

Return the numpy dtype for the data including the endianess.

template<typename Scalar >

bool ldaplusplus::numpy_format::NumpyOutput< Scalar >::fortran_contiguous ( ) const

inline

Wether the data is fortran contiguous (column major) or C contiguous (row major).

template<typename Scalar >

const std::vector<size_t>& ldaplusplus::numpy_format::NumpyOutput< Scalar >::shape ( ) const

inline

Return a vector containing the shape of the numpy array (rows, columns, etc).

Friends And Related Function Documentation

template<typename Scalar >

std::ostream& operator<< ( std::ostream &  os, const NumpyOutput< Scalar > &  data  )

friend

This operator implements the npy format for a block of NumpyOutput of type Scalar.

---

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

• include/ldaplusplus/NumpyFormat.hpp