wccxyf: Supervised mapping of spectra with self-organising maps

Description

Supervised self-organising maps for mapping high-dimensional spectra or patterns to 2D; instead of Euclidean distance, the weighted cross correlation (WCC) similarity measure is used. Modelled after the SOM function in package 'class'. wccxyf takes 'continous' patterns, i.e. datapoints are equidistant.

At this point, no facilities are implemented for growing networks or k-means-like fine-tuning of the maps, such as in function wccsom.

Usage

wccxyf(data, Y, grid=somgrid(), rlen = 100, alpha = c(0.05, 0.01), radius = quantile(nhbrdist, 0.67), xweight = 0.5, trwidth = 20, toroidal = FALSE, keep.data = TRUE)

Arguments

data

Spectra or patterns to be mapped: a matrix, with each row representing a compound.

Property for each pattern, either a numerical vector or matrix, or a class matrix. In the latter case, the Tanimoto distance is used for Y; in all other cases (also for combinations of numerical and class properties) the Euclidean distance is used.

grid

A grid for the representatives: see 'somgrid'.

rlen

the number of times the complete data set will be presented to the network.

alpha

a vector of two numbers indicating the amount of change. Default is to decline linearly from 0.05 to 0.01 over rlen updates.

radius

the initial radius of the neighbourhood to be used for each update: the decrease is exponential over rlen updates in such a way that after one-third of the updates only the winning unit is updated. The default is to start with a value that covers 2/3 of all units.

xweight

weight of X matrix in determining the distances of objects to units.

trwidth

width of the triangle function used in the WCC measure, given in the number of data points.

toroidal

if TRUE, then the edges of the map are joined. Note that in a toroidal hexagonal map, the number of rows must be even.

keep.data

store training data and their mapping in the network.

Value

grid: the grid, an object of class '"somgrid"'.
changes: vector of mean average deviations from code vectors
codes: a matrix of code vectors.
trwdth: the triangle width used for the WCC measure
acors: autocorrelations of the code vectors.
toroidal: setting of parameter 'toroidal'.
FineTune: setting of parameter 'FineTune'.
unit.classif: mapping of training data: a vector of unit numbers. Only if keep.data equals TRUE.
wccs: WCC values of all training data, compared to the best matching codebook vector. Only if keep.data equals TRUE.
data.acors: WAC values for training data. Only if keep.data equals TRUE.

References

FIXME: this page is a copy of wccsom, should be edited further

Examples

Run this code

## Not run: 
# data(degelder)
# gr <- somgrid(5, 5, "hexagonal")
# set.seed(7)
# x <- wccxyf(degelder$patterns, degelder$properties[,"cell.vol"],
#             grid=gr, trwidth=20, rlen=100)
# ## End(Not run)

Run the code above in your browser using DataLab