Normalizes data in K dimensions using principal curves such that afterward the data cluster (approximately linearly) along the diagonal (in K dimensions).
# S3 method for matrix
normalizePrincipalCurve(x, ..., center=TRUE, returnFit=FALSE)
Returns an NxK matrix
.
An NxK matrix
where the columns represent the
(K >= 2) dimensions.
Additional arguments passed to
fitPrincipalCurve
used for fitting the model.
If TRUE
, normalized data is centered such that the median
signal in each dimension is at zero.
If TRUE
, the fitted principal curve parameters are
returned as an attribute.
Henrik Bengtsson
[1] Hastie, T. and Stuetzle, W, Principal Curves, JASA, 1989.
fitPrincipalCurve
and
backtransformPrincipalCurve
.