Xapian::KMeans Class Reference

Kmeans clusterer: This clusterer implements the K-Means clustering algorithm. More...

#include <cluster.h>

Inheritance diagram for Xapian::KMeans:

Collaboration diagram for Xapian::KMeans:

Public Member Functions
	KMeans (unsigned int k_, unsigned int max_iters_=0)
	Constructor specifying number of clusters and maximum iterations. More...
ClusterSet	cluster (const MSet &mset) override
	Implements the KMeans clustering algorithm. More...
void	set_stopper (const Xapian::Stopper *stop=NULL)
	Set the Xapian::Stopper object to be used for identifying stopwords. More...
std::string	get_description () const override
	Return a string describing this object. More...
Public Member Functions inherited from Xapian::Clusterer
virtual	~Clusterer ()
	Destructor. More...
Clusterer *	release ()
	Start reference counting this object. More...
const Clusterer *	release () const
	Start reference counting this object. More...
Public Member Functions inherited from Xapian::Internal::opt_intrusive_base
	opt_intrusive_base (const opt_intrusive_base &)
opt_intrusive_base &	operator= (const opt_intrusive_base &)
	opt_intrusive_base ()
	Construct object which is initially not reference counted. More...
virtual	~opt_intrusive_base ()
void	ref () const
void	unref () const

Private Member Functions
void	initialise_clusters (ClusterSet &cset, Xapian::doccount num_of_points)
	Initialise 'k' clusters by selecting 'k' centroids and assigning them to different clusters. More...
void	initialise_points (const MSet &source)
	Initialise the Points to be fed into the Clusterer with the MSet object 'source'. More...

Private Attributes
std::vector< Point >	points
	Contains the initialised points that are to be clustered. More...
unsigned int	k
	Specifies that the clusterer needs to form 'k' clusters. More...
unsigned int	max_iters
	Specifies the maximum number of iterations that KMeans will have. More...
Xapian::Internal::opt_intrusive_ptr< const Xapian::Stopper >	stopper
	Pointer to stopper object for identifying stopwords. More...

Additional Inherited Members
Public Attributes inherited from Xapian::Internal::opt_intrusive_base
unsigned	_refs
	Reference count. More...
Protected Member Functions inherited from Xapian::Internal::opt_intrusive_base
void	release () const
	Start reference counting. More...

Detailed Description

Kmeans clusterer: This clusterer implements the K-Means clustering algorithm.

Definition at line 596 of file cluster.h.

Constructor & Destructor Documentation

KMeans()

Xapian::KMeans::KMeans ( unsigned int  k_, unsigned int  max_iters_ = 0  )

explicit

Constructor specifying number of clusters and maximum iterations.

Parameters

k_	Number of required clusters
max_iters_	The maximum number of iterations for which KMeans will run if it doesn't converge

Member Function Documentation

cluster()

ClusterSet Xapian::KMeans::cluster ( const MSet &  mset )

overridevirtual

Implements the KMeans clustering algorithm.

Parameters

mset	MSet object containing the documents that are to be clustered

Implements Xapian::Clusterer.

get_description()

std::string Xapian::KMeans::get_description ( ) const

overridevirtual

Return a string describing this object.

Implements Xapian::Clusterer.

initialise_clusters()

void Xapian::KMeans::initialise_clusters ( ClusterSet &  cset, Xapian::doccount  num_of_points  )

private

Initialise 'k' clusters by selecting 'k' centroids and assigning them to different clusters.

Parameters

cset	ClusterSet object to be initialised by assigning centroids to each cluster
num_of_points	Number of points passed to clusterer

initialise_points()

void Xapian::KMeans::initialise_points ( const MSet &  source )

private

Initialise the Points to be fed into the Clusterer with the MSet object 'source'.

The TF-IDF weights for the documents are calculated and stored within the Points to be used later during distance calculations

Parameters

source

MSet object containing the documents which will be used to create document vectors that are represented as Point objects

set_stopper()

void Xapian::KMeans::set_stopper ( const Xapian::Stopper *  stop = NULL )

inline

Set the Xapian::Stopper object to be used for identifying stopwords.

Stopwords are discarded while calculating term frequency for terms.

Parameters

stop	The Stopper object to set (default NULL, which means no stopwords)

Definition at line 651 of file cluster.h.

Member Data Documentation

k

unsigned int Xapian::KMeans::k

private

Specifies that the clusterer needs to form 'k' clusters.

Definition at line 601 of file cluster.h.

max_iters

unsigned int Xapian::KMeans::max_iters

private

Specifies the maximum number of iterations that KMeans will have.

Definition at line 604 of file cluster.h.

points

std::vector<Point> Xapian::KMeans::points

private

Contains the initialised points that are to be clustered.

Definition at line 598 of file cluster.h.

stopper

Xapian::Internal::opt_intrusive_ptr<const Xapian::Stopper> Xapian::KMeans::stopper

private

Pointer to stopper object for identifying stopwords.

Definition at line 607 of file cluster.h.

---

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

• include/xapian/cluster.h