ds.nbclust: Determines optimal number of clusters for dataset

In FlorianSchw/dsClusterAnalysisClient: DataSHIELD client side functions for the dsClusterAnalysisClient package

Determines optimal number of clusters for dataset

Description

This function is similar to the R function 'NBClust' from the NBClust package

Usage

ds.nbclust(
  df.name = NULL,
  diss = NULL,
  distance = "euclidean",
  min.nc = 2,
  max.nc = 15,
  method = NULL,
  index = "all",
  alphaBeale = 0.1,
  seed = 123,
  datasources = NULL
)

Arguments

df.name	is a string character of the data set and can be either a matrix or data frame
diss	is a dissimilarity structure which will be calculated according to the distance method
distance	specifies the method for the distance matrix calculation and can be either 'euclidean', 'maximum', 'manhattan', 'canberra', 'binary' or 'minkowski'
min.nc	specifies the minimum number of clusters
max.nc	specifies the maximum number of clusters
method	describes the clustering method and can be either "ward.D2", "single", "complete", "average", "mcquitty", "median", "centroid", "kmeans" or "ward.D"
index	describes the clustering index and can be either "kl", "ch", "hartigan", "ccc", "scott", "marriot", "trcovw", "tracew", "friedman", "rubin", "cindex", "db", "silhouette", "duda", "pseudot2", "beale", "ratkowsky", "ball", "ptbiserial", "gap", "frey", "mcclain", "gamma", "gplus", "tau", "dunn", "hubert", "sdindex", "dindex", "sdbw", "all" or "alllong"
alphaBeale	value for "beale" clustering index
seed	is an integer for random start point
datasources	is a DSConnection object

Details

The function uses partitioning methods to find optimal numbers of clusters for a given dataset.

Value

a summary suggesting the optimal number of clusters

Author(s)

Florian Schwarz for the German Institute of Human Nutrition

FlorianSchw/dsClusterAnalysisClient documentation built on Feb. 8, 2025, 10:32 a.m.