Build a self-organizing map
SOM(data, xdim = 10, ydim = 10, zdim = NULL, rlen = 10,
alphaA = c(0.05, 0.01), radiusA = stats::quantile(nhbrdist, 0.67) *
c(1, 0), alphaB = alphaA * c(-negAlpha, -0.01 * negAlpha),
radiusB = negRadius * radiusA, init = FALSE,
initf = Initialize_PCA, distf = 2, codes = NULL,
importance = NULL, nhbr.method = "maximum", negRadius = 1.33,
negAlpha = 0.1, noMapping = F)Matrix containing the training data
Width of the grid
Hight of the grid
Depth of the grid, causes grid to be 3D
Number of times to loop over the training data for each MST
Start and end learning rate
Start and end radius
Start and end learning rate for the second radius
Start and end radius (make sure it's larger than radiusA)
Initialize cluster centers in a non-random way
Use the given initialization function if init==T (default: Initialize_PCA)
Distance function (1=manhattan, 2=euclidean, 3=chebyshev)
Cluster centers to start with
array with numeric values. Parameters will be scaled according to importance
Way of computing grid distances, passed as method= to dist() function. Default 'maximum' (square neighborhoods); use 'euclidean' for round neighborhoods.
easy way to set radiusB as a multiple of default radius (use lower value for higher dimensions)
the same for alphaB
If true, do not produce mapping (default F). Useful for online/streaming use.
A map, which is a list containing all parameter settings and results
FlowSOM::SOM