Distance Class

Provides methods to summarize distances within and between sets of multidimensional points.

Inheritance Hierarchy

SystemObject
Novacta.AnalyticsDistance

Namespace: Novacta.Analytics
Assembly: Novacta.Analytics (in Novacta.Analytics.dll) Version: 2.0.0

Syntax

C++

Copy

public static class Distance

Public NotInheritable Class Distance

public ref class Distance abstract sealed

[<AbstractClassAttribute>]
[<SealedAttribute>]
type Distance =  class end

The Distance type exposes the following members.

Methods

	Name	Description
	AverageLinkage	Computes the average linkage between the specified clusters.
	CentroidLinkage	Computes the centroid linkage between the specified clusters.
	CompleteDiameter	Computes the complete diameter of the specified cluster.
	CompleteLinkage	Computes the complete linkage between the specified clusters.
	Euclidean	Computes the Euclidean distance matrix of the specified cluster.
	SingleLinkage	Computes the single linkage between the specified clusters.

Top

Remarks

Sets of multidimensional points are referred to as clusters.

Distances within a cluster

The Euclidean distances among points in a cluster are returned by method Euclidean(DoubleMatrix) as a distance matrix. The complete diameter of a cluster, i.e. the maximal Euclidean distance among two cluster points, is instead computed by method CompleteDiameter(DoubleMatrix).

Distances between clusters

Distances between points belonging to different clusters can be summarized in terms of linkages. Method SingleLinkage computes the shortest distance between pairs of individuals belonging to different clusters, while CompleteLinkage and AverageLinkage compute the maximum and the average of those distances, respectively. Method CentroidLinkage the distance between the mean points (centroids) of the specified clusters.

Reference

Novacta.Analytics Namespace