nroTrain {Numero}

Train self-organizing map

Description

Iterative algorithm to adapt a self-organizing map (SOM) to a set of multivariable data.

Usage

nroTrain(map, data, subsample = NULL, balance = 0, message = NULL)

Arguments

map

A list object as returned by nroKohonen().

data

A matrix or a data frame.

subsample

Number of rows used during a single training cycle.

balance

Penalty parameter for variation in the numbers of resident samples across disctricts, see nroKmeans().

message

If positive, progress information is printed at the specified interval in seconds.

Details

The map is fitted according to columns that are found both in the SOM centroids and the input data.

If subsample is less than the number of data rows, a random subset of the specified size is used for each training cycle. By default, subsample is set automatically depending on the size of the dataset.

Value

A copy of the list object map, where the element centroids is updated according to the data patterns. The quantization errors during training are stored in the element history. The subsampling parameter that was used during training is stored in the element subsample.

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)

# Train with full data.
sm <- nroKohonen(seeds = km)
sm <- nroTrain(map = sm, data = trdata, subsample = nrow(trdata))
print(sm$history)

# Train with subsampling.
sm <- nroKohonen(seeds = km)
sm <- nroTrain(map = sm, data = trdata, subsample = 200)
print(sm$history)