reAssign: Re-Assign Observations Trimmed Prior to Trimmed k-Means...

In QuantileGH: Quantile Least Mahalanobis Distance Estimator for Tukey g-&-h Mixture

Re-Assign Observations Trimmed Prior to Trimmed k-Means Clustering

Description

Re-assign the observations, which are trimmed in the trimmed k-means algorithm, back to the closest cluster as determined by the smallest Mahalanobis distance.

Usage

reAssign(x, ...)

## S3 method for class 'tkmeans'
reAssign(x, ...)

Arguments

x	a tkmeans object
...	potential parameters, currently not in use.

Details

Given the tkmeans input, the mahalanobis distance is computed between each trimmed observation and each cluster. Each trimmed observation is assigned to the closest cluster (i.e., with the smallest Mahalanobis distance).

Value

Function reAssign.tkmeans returns an 'reAssign_tkmeans' object, which inherits from tkmeans class.

Note

Either kmeans or tkmeans is slow for big x.

Examples

library(tclust)
data(geyser2)
clus = tkmeans(geyser2, k = 3L, alpha = .03)
plot(clus, main = 'Before Re-Assigning')
plot(reAssign(clus), main = 'After Re-Assigning')

QuantileGH documentation built on May 29, 2024, 12:14 p.m.