Orthogonal PLS-DA (from ropls)
CalculateImpVarCutoff(mSetObj, spe.thresh, lev.thresh)
Input the name of the created mSetObj (see InitDataObjects)
alpha threshold, less is better, default less than 5 percentile based chi-square note: spe and leverage are vectors, not a single value, but a list to store the result note: the last model is Model.res, no spe Calculate leverage cutoff based on permutation Calculate the reference distribution of leverages note: leverage.perm is a list with each member in a 3 column matrix
leverage threshold, the higher better, default more than 95 percentile of permuted leverage