restlos (version 0.2-2)

flood: The Flood Algorithm

Description

The function determines a robust subsample utilizing self-organizing maps (SOM).

Usage

flood(data, Nx=10, Ny=10, rlen=2000)

Arguments

data

At least a two-dimensional data matrix is required. Number of observations needs to be greater than number of dimensions.

Nx

Size of the SOM-net in x direction. Default is 10.

Ny

Size of the SOM-net in y direction. Default is 10.

rlen

Number of iterations during SOM learn process. Default is 2000.

Value

som.results

SOM results as delivered by som.

som.neigh

A matrix showing for every neuron (first column) the index off the neighboring neurons (columns 2-5).

umatrix

The U-matrix shows the U-value for every neuron.

winneuron

Vector of length n giving the index of the nearest neuron (Euclidean distance).

lib

List of all basins found. Index of neurons. Smallest subsample of size (n+d+1)/2.

lin

List of all neighboring neurons per basin. Index of neurons. Smallest subsample of size (n+d+1)/2.

geb

Number of associated data points per basin. Smallest subsample of size (n+d+1)/2.

l

Internal value necessary for plotting.

fafh

Data for plotting the flood area flood height curve.

fafh.lib

Internal data necessary for plotting extented flooding.

fafh.drin

Internal data necessary for plotting extented flooding.

drin

Robust subsample of minimal size.

Details

The function first calls the som function within the som-package. The results are subsequently used to determine a robust subsample. Arguments Nx, Ny and rlen are passed to som. These arguments should be selected depending on the size of the data set (number of observations/dimensions). The larger the data set the larger the net size and the number of iterations should be. Note: At the moment only rectangular and quadratic SOM nets are supported.

References

Liebscher, S., Kirschstein, T., and Becker, C. (2012): The Flood Algorithm - A Multivariate, Self-Organizing-Map-Based, Robust Location and Covariance Estimator, Statistics and Computing, 22(1), 325-336, DOI: 10.1007/s11222-011-9250-3.

Examples

# flood(halle)