UniquePoints: Unique Points

In Mthrun/DatabionicSwarm: Swarm Intelligence for Self-Organized Clustering

Unique Points

Description

return only the unique points in Datapoints

Usage

UniquePoints(Datapoints, Cls, Eps=1e-10)

Arguments

Datapoints	[1:n,1:d] numeric matrix of Datapoints points of dimension d, the points are in the rows
Cls	[1:n] numeric vector of classes for each datapoint.
Eps	Optional,scalar above zero that defines minimum non-identical euclidean distance between two points

Details

Euclidean distance is computed and used within. Setting Eps to a very small number results in the identification of unique data points. Setting epsilon to a higher number results in the definition of mesh points within an d-dimensional R-ball graph.

Value

List with

Unique	[1:k,1:d] Datapoints points without duplicate points
UniqueInd	[1:k] index vector such that Unique == Datapoints[UniqueInd,], it has k non-consecutive numbers or labels, each label defines a row number within Datapoints[1:n,1:d] of a unique data point
Uniq2DatapointsInd	[1:n] index vector. It has k unique index numbers representing the arbitrary labels. Each labels is mapped uniquely to a point in Unique. Logically in a way such that Datapoints == Unique[Uniq2DatapointsInd,] (will not work directly in R this way)
NewUniqueInd	[1:k] index vector stating the index of the newly defined datastructure Unique.
NewUniq2DataIdx	[1:k] index vector such that Unique[NewUniq2DataIdx,] == Datapoints[Uniq2DatapointsInd,], it has n non-consecutive numbers or labels, each label defines a row number within Unique[1:k,1:d] of a unique data point
IsDuplicate	[1:n,1:n] matrix,for i!=j IsDuplicate[i,j]== 1 if Datapoints[i,] == Datapoints[j,] IsDuplicate[i,i]==0
Eps	Numeric stating the neighborhood radius around unique points.

Author(s)

Michael Thrun

Examples

Datapoints  = rbind(c(0,0), c(1,1), c(2,2))
Datapoints2 = rbind(Datapoints, Datapoints+0.001)
Datapoints3 = rbind(Datapoints2, c(1,1)-0.001)

Datapoints  = rbind(c(0,0), c(0,0.015), c(0,0.01), c(0,0.015))

V1 = UniquePoints(Datapoints = Datapoints, Eps = 0.01)
V2 = UniquePoints(Datapoints = Datapoints2, Eps = 0.01)
V3 = UniquePoints(Datapoints = Datapoints3, Eps = 0.01)

Mthrun/DatabionicSwarm documentation built on Nov. 2, 2023, 6:51 a.m.