SFA_EXECUTE Apply the learned functions to input data.
OUT = SFA_EXECUTE(HDL, DATA[, PRJ, NCOMP]) applies the SFA
functions learned by the SFA obejct referenced by HDL to the input
data DATA and returns their output. (The learning phase has of
course to be finished, i.e. the SFA-step 'sfa' has to be completed).
The execution is completed in 4 steps:
1. projection on the input principal components (dimensionality
reduction)
2. expansion (if necessary)
3. projection on the whitened (expanded) space
4. projection on the slow functions
If PRJ is defined and nonzero, the preprocessing step 1 is skipped
(for SFA2 objects only).
If NCOMP is defined, it specifies the number of functions to be used.
See also SFA_STEP