auto_cluster: Cluster Analysis combining Hierarquical and Kmeans cluster

In neale-eldash/pd: Day-2-day functions in R

Description

This function runs as Cluster Analysis. The first step is to run a Hierarquical Cluster, and then use centroids as starting point for the Kmeans Cluster. This function outputs the cluster ids for each line in the dataframe and a summary of the group sizes. Details of the analysis for each step are:

• Hierarquical: Uses Euclidean distance and Ward's method. See more details in hclust.

• Kmeans: Uses default options. See more details in kmeans.

Usage

auto_cluster(df = NULL, grps = 3:6, name = "kmeans", iter.max = 100)

Arguments

df	A dataframe containing the variables to be used in the analysis.
grps	An array or number with the number of groups that should be created.
name	A string with the name of the variables that will be created.
iter.max	A number indicating the maximum number of iterations for the Kmeans cluster.

Value

A list with three components:

• grps(dataframe): with the variables identifying the cluster each observation belongs too.

• summary(dataframe): summary of the number of observations per cluster.

• details(dataframe): details for each iteration of the kmeans algorithm.

Examples

df.cluster <- auto_cluster(df=df,grps=3:6,name='kmeans')

neale-eldash/pd documentation built on June 26, 2021, 10:47 a.m.