ipfCluster {ipft}

Creates clusters using the specified method

Description

Creates clusters using the the specified method and assigns a cluster id to each cluster

Usage

ipfCluster(data, method = "k-means", k = NULL, grid = NULL, ...)

Arguments

data

a data frame

method

the method to use to clusterize the data. Implemented methods are: 'k-means' for k-means algorithm. Requires parameter k. 'grid' for clustering based on grid partition. Requires parameter grid. 'AP' for affinity propagation algorithm.

k

parameter k

grid

a vector with the grid size for the 'grid' method

...

additional parameters for k-means, apcluster and apclusterK for 'k-means' method additional parameters see: https://stat.ethz.ch/R-manual/R-devel/library/stats/html/kmeans.html for 'apcluster'AP' method additional parameters see: https://cran.r-project.org/web/packages/apcluster/index.html

Value

A list with: clusters -> a numeric vector with the ids of the clusters centers -> a data frame with the centers of the clusters

Examples

    clusters <- ipfCluster(head(ipftrain, 20)[, 169:170], k = 4)

    clusters <- ipfCluster(head(ipftrain[, grep('^wap', names(ipftrain))], 20),
    method = 'AP')$clusters