When an empty cluster is detected, this class takes the point furthest from the centroid of the cluster with maximum variance as a new cluster. More...

Public Member Functions
	MaxVarianceNewCluster ()
	Default constructor required by EmptyClusterPolicy. More...

template<typename MetricType , typename MatType >
size_t	EmptyCluster (const MatType &data, const size_t emptyCluster, const arma::mat &oldCentroids, arma::mat &newCentroids, arma::Col< size_t > &clusterCounts, MetricType &metric, const size_t iteration)
	Take the point furthest from the centroid of the cluster with maximum variance to be a new cluster. More...

template<typename Archive >
void	Serialize (Archive &ar, const unsigned int version)
	Serialize the object. More...

Private Member Functions
template<typename MetricType , typename MatType >
void	Precalculate (const MatType &data, const arma::mat &oldCentroids, arma::Col< size_t > &clusterCounts, MetricType &metric)
	Called when we are on a new iteration. More...

Private Attributes
arma::Row< size_t >	assignments
	Cached assignments for each point. More...

size_t	iteration
	Index of iteration for which variance is cached. More...

arma::vec	variances
	Cached variances for each cluster. More...

Detailed Description

When an empty cluster is detected, this class takes the point furthest from the centroid of the cluster with maximum variance as a new cluster.

Definition at line 26 of file max_variance_new_cluster.hpp.

Constructor & Destructor Documentation

mlpack::kmeans::MaxVarianceNewCluster::MaxVarianceNewCluster ( )

inline

Default constructor required by EmptyClusterPolicy.

Definition at line 30 of file max_variance_new_cluster.hpp.

References EmptyCluster(), iteration, and Serialize().

Member Function Documentation

template<typename MetricType , typename MatType >

size_t mlpack::kmeans::MaxVarianceNewCluster::EmptyCluster	(	const MatType &	data,
		const size_t	emptyCluster,
		const arma::mat &	oldCentroids,
		arma::mat &	newCentroids,
		arma::Col< size_t > &	clusterCounts,
		MetricType &	metric,
		const size_t	iteration
	)

Take the point furthest from the centroid of the cluster with maximum variance to be a new cluster.

Template Parameters

MatType Type of data (arma::mat or arma::sp_mat).

Parameters

data	Dataset on which clustering is being performed.
emptyCluster	Index of cluster which is empty.
oldCentroids	Centroids of each cluster (one per column), at the start of the iteration.
newCentroids	Centroids of each cluster (one per column), at the end of the iteration. This will be modified!
clusterCounts	Number of points in each cluster.
assignments	Cluster assignments of each point.

Returns: Number of points changed.

Referenced by MaxVarianceNewCluster().

template<typename MetricType , typename MatType >

void mlpack::kmeans::MaxVarianceNewCluster::Precalculate	(	const MatType &	data,
		const arma::mat &	oldCentroids,
		arma::Col< size_t > &	clusterCounts,
		MetricType &	metric
	)