nroPermute: Permutation analysis of map layout

Description

Estimate the dynamic range and statistical significance for regional patterns on a self-organizing maps using permutations.

Usage

nroPermute(som, districts, data, n = 10000, clip = 5.0, message = NA)

Arguments

som

A list object in the format from nroTrain().

districts

An integer vector of M best matching districts.

data

A numeric vector of M values or an M x N matrix (or data frame), where M is the number of data points and N is the number of variables.

Maximum number of permutations.

clip

Range parameter for outlier clipping (standard deviations from the median).

message

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

Value

A data frame with eight columns. For example, P.z is a parametric estimate for statistical significance, P.freq is the frequency-based estimate for statistical signicance, and Z is the estimated z-score of how far the observed map plane is from the average randomly generated layout. N.data indicates how many data values were used and N.cycles tells the number of completed permutations. AMPLITUDE is a dynamic range modifier for colors that can be used in nroColorize().

The output also contains the attribute 'zbase' that indicates the normalization factor for the color amplitudes.

Details

The input argument som must contain the map topology and the centroid profiles as returned by the functions nroKmeans(), nroKohonen(), or nroTrain().

The input argument districts must contain integers between 1 and K, where K is the number map units. Any other values will be ignored.

Training variables and data points are detected by the column names of som$centroids, the attribute "variables" in districts and the names of elements in districts.

References

Gao S, Mutter S, Casey AE, M<U+00E4>kinen V-P (2018) Numero: a statistical framework to define multivariable subgroups in complex population-based datasets, Int J Epidemiology, https://doi.org/10.1093/ije/dyy113

Examples

Run this code

# NOT RUN {
# Import data.
fname <- system.file("extdata", "finndiane.txt", package = "Numero")
dataset <- read.delim(file = fname)

# Set row names.
rownames(dataset) <- paste("r", 1:nrow(dataset), sep="")

# 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)
sm <- nroTrain(som = sm, data = trdata)

# Assign data points into districts.
matches <- nroMatch(centroids = sm, data = trdata)

# Estimate statistics for cholesterol
chol <- nroPermute(som = sm, districts = matches, data = dataset$CHOL)
print(chol[,c("TRAINING", "Z", "P.z", "P.freq")])

# Estimate statistics.
stats <- nroPermute(som = sm, districts = matches, data = dataset)
print(stats[,c("TRAINING", "Z", "P.z", "P.freq")])
# }

Run the code above in your browser using DataCamp Workspace