nroKohonen: Self-organizing map

In Numero: Statistical Framework to Define Subgroups in Complex Datasets

Self-organizing map

Description

Interpolates the initial district profiles of a self-organizing map based on pre-determined seed profiles.

Usage

nroKohonen(seeds, radius = 3, smoothness = 1.0)

Arguments

seeds	A matrix or a data frame of K rows and N columns.
radius	Map radius.
smoothness	Rigidity of the map to adapt to regional differences.

Value

A list of named elements: centroids contains the N-dimensional district profiles, and topology is an H x 6 matrix that contains the 2D spatial layout for the map districts: the first two columns (X, Y) indicate the positions of districts in Cartesian coordinates, the other four columns (RADIUS1, RADIUS2, ANGLE1, ANGLE2) define the perimeter of the district areas for visualisation on a circular map.

Additional parameters are stored as attributes in topology.

The function is named after Teuvo Kohonen, the inventor of the self-organizing map.

See Also

Please see nroKmeans() to create the seeds.

Examples

# Import data.
fname <- system.file("extdata", "finndiane.txt", package = "Numero")
dataset <- read.delim(file = fname)

# Prepare training data.
trvars <- c("CHOL", "HDL2C", "TG", "CREAT", "uALB")
trdata <- scale.default(dataset[,trvars]) 

# K-means clustering.
km <- nroKmeans(data = trdata)

# Self-organizing map.
sm <- nroKohonen(seeds = km)
print(head(sm$centroids))
print(head(sm$topology))

Numero documentation built on Sept. 17, 2024, 5:09 p.m.