PFclustering: Power Fuzzy Clustering

In FCPS: Fundamental Clustering Problems Suite

Power Fuzzy Clustering

Description

Performs Power Fuzzy Clustering [Nguyen et al, 2026]

Usage

PFclustering(
  Data,
  ClusterNo,
  Fuzzifier = 2,
  DistanceMethod = "Euclidean",
  NoIterations = 100,
  PlotIt = FALSE,
  ...
)

Arguments

Data	Numeric data set with n observations and d features.
ClusterNo	Integer or NULL. Number of clusters to search for.
Fuzzifier	Optional, Numeric scalar, strictly greater than 1. Fuzzifier parameter m. Default is 2.
DistanceMethod	Optional, Character scalar. Distance metric to use. Choices are "Minkowski", "Mahalanobis", or "Euclidean".
NoIterations	Optional, , Integer scalar. Maximum number of iterations. Default is 100.
PlotIt	Optional, Logical. If TRUE, plots the clustering result via ClusterPlotMDS(). Default is FALSE.
...	Optional, Further parameters passed to pfclust::PFC(), depending on the selected distance. These may include, for example: q: distance exponent, must be strictly greater than 0 p: Minkowski exponent alpha, beta: Mahalanobis parameters threshold: convergence tolerance

Details

Please see [Nguyen et al, 2026].

Value

A list with components:

• Cls: numeric vector of length nrow(Data) containing cluster assignments.

• Object: object returned by pfclust::PFC().

Author(s)

Michael Thrun

References

P. T. Nguyen, C. Tortora and A. Punzo: “Power Fuzzy Clustering: Flexible Distance Metrics and Inclusion of Covariates,” IEEE Transactions on Fuzzy Systems, Vol. 34 (7), pp. 2171–2182, doi: 10.1109/TFUZZ.2026.3683998, 2026.

Examples

data('Hepta')

out=PFclustering(Hepta$Data,ClusterNo=7,
PlotIt=FALSE,NoIterations=10)

FCPS documentation built on July 15, 2026, 5:08 p.m.