as.kernelmatrix: Turn a standard matrix into a Kernel matrix

In yasomi: Yet Another Self Organising Map Implementation

Description

This function simply adds the "kernelmatrix" class to a standard matrix for subsequent use for a Kernel Self-Organising Map. Optionally performs some sanity check on the matrix to verify that it represents a non negative kernel matrix.

Usage

as.kernelmatrix(data, ...)

Arguments

data	an object of class "matrix"
...	not used

Details

The current version of the function only checks for symmetry of the data. Future versions will provide optional non negativity tests.

Value

An object of class "kernelmatrix" which is in fact a matrix.

Author(s)

Fabrice Rossi

See Also

See batchsom.kernelmatrix for the main use of Kernel matrices

Examples

## the linear kernel applied to the Iris dataset
data(iris)
# scaling and kernel computation
data <- as.kernelmatrix(tcrossprod(scale(iris[1:4])))

yasomi documentation built on May 2, 2019, 5:59 p.m.