powered by
The data matrix is scaled and sphered so it is orthonormal. The Cholesky decomposition is used.
sphereX(X)
X rectangular data matrix
X
sphered matrix
scale, NNPredict
scale
NNPredict
# NOT RUN { data(longley) longley.x <- data.matrix(longley[, 1:6]) sphereX(longley.x) # }
Run the code above in your browser using DataLab