kmeans::Matrix< Index_, Data_ > Class Template Reference

Interface for matrix data. More...

#include <Matrix.hpp>

Inheritance diagram for kmeans::Matrix< Index_, Data_ >:

Public Member Functions
virtual Index_	num_observations () const =0
virtual std::size_t	num_dimensions () const =0
virtual std::unique_ptr< RandomAccessExtractor< Index_, Data_ > >	new_extractor () const =0
virtual std::unique_ptr< ConsecutiveAccessExtractor< Index_, Data_ > >	new_extractor (Index_ start, Index_ length) const =0
virtual std::unique_ptr< IndexedAccessExtractor< Index_, Data_ > >	new_extractor (const Index_ *sequence, std::size_t length) const =0
auto	new_known_extractor () const
auto	new_known_extractor (Index_ start, Index_ length) const
auto	new_known_extractor (const Index_ *sequence, std::size_t length) const

Detailed Description

template<typename Index_, typename Data_>
class kmeans::Matrix< Index_, Data_ >

Interface for matrix data.

This defines the expectations for a matrix of observation-level data to be used in Initialize::run() and Refine::run(). Each matrix should support extraction of the vector of coordinates for each observation.

Template Parameters

Index_	Integer type of the observation indices.
Data_	Numeric type of the data.

Member Function Documentation

new_extractor() [1/3]

template<typename Index_ , typename Data_ >

virtual std::unique_ptr< RandomAccessExtractor< Index_, Data_ > > kmeans::Matrix< Index_, Data_ >::new_extractor ( ) const

pure virtual

Returns

A new random-access extractor.

Implemented in kmeans::SimpleMatrix< Index_, Data_ >.

new_extractor() [2/3]

template<typename Index_ , typename Data_ >

virtual std::unique_ptr< IndexedAccessExtractor< Index_, Data_ > > kmeans::Matrix< Index_, Data_ >::new_extractor ( const Index_ * sequence, std::size_t length ) const

pure virtual

Parameters

[in]	sequence	Pointer to an array of sorted and unique indices of observations, to be accessed in the provided order. It is assumed that the vector will not be deallocated before the destruction of the returned IndexedAccessExtractor.
	length	Number of observations in sequence.

Returns

A new indexed-access extractor.

Implemented in kmeans::SimpleMatrix< Index_, Data_ >.

new_extractor() [3/3]

template<typename Index_ , typename Data_ >

virtual std::unique_ptr< ConsecutiveAccessExtractor< Index_, Data_ > > kmeans::Matrix< Index_, Data_ >::new_extractor ( Index_ start, Index_ length ) const

pure virtual

Parameters

start	Start of the contiguous block of observations to be accessed consecutively.
length	Length of the contiguous block of observations to be accessed consecutively.

Returns

A new consecutive-access extractor.

Implemented in kmeans::SimpleMatrix< Index_, Data_ >.

new_known_extractor() [1/3]

template<typename Index_ , typename Data_ >

auto kmeans::Matrix< Index_, Data_ >::new_known_extractor ( ) const

inline

Returns

A new random-access extractor.

Subclasses may override this method to return a pointer to a specific RandomAccessExtractor subclass. This is used for devirtualization in other kmeans functions. If no override is provided, new_extractor() is called instead.

new_known_extractor() [2/3]

template<typename Index_ , typename Data_ >

auto kmeans::Matrix< Index_, Data_ >::new_known_extractor ( const Index_ * sequence, std::size_t length ) const

inline

Parameters

sequence	Pointer to a sorted and unique array of observations, see the equivalent new_extractor() overload.
length	Length of the array, see the equivalent new_extractor() overload.

Returns

A new indexed-access extractor.

Subclasses may override this method to return a pointer to a specific IndexedAccessExtractor subclass. This is used for devirtualization in other kmeans functions. If no override is provided, the relevant new_extractor() overload is called instead.

new_known_extractor() [3/3]

template<typename Index_ , typename Data_ >

auto kmeans::Matrix< Index_, Data_ >::new_known_extractor ( Index_ start, Index_ length ) const

inline

Parameters

start	Start of the contiguous block, see the equivalent new_extractor() overload.
length	Length of the contiguous block, see the equivalent new_extractor() overload.

Returns

A new consecutive-access extractor.

Subclasses may override this method to return a pointer to a specific ConsecutiveAccessExtractor subclass. This is used for devirtualization in other kmeans functions. If no override is provided, the relevant new_extractor() overload is called instead.

num_dimensions()

template<typename Index_ , typename Data_ >

virtual std::size_t kmeans::Matrix< Index_, Data_ >::num_dimensions ( ) const

pure virtual

Returns

Number of dimensions.

Implemented in kmeans::SimpleMatrix< Index_, Data_ >.

num_observations()

template<typename Index_ , typename Data_ >

virtual Index_ kmeans::Matrix< Index_, Data_ >::num_observations ( ) const

pure virtual

Returns

Number of observations.

Implemented in kmeans::SimpleMatrix< Index_, Data_ >.

---

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

• kmeans/Matrix.hpp