In some cases when user does not limit number of PLS components manually, PLS results in following error message:
Error in crossval.regmodel(model, x, y, cv, cal.fun = pls.cal) :
In addition: Warning message:
In sqrt(as.numeric(c)) : NaNs produced
The problem most probably is related to singular matrix problem in SIMPLS algorithm in one of the cross-validation step. Check if number of components is adjusted correctly in cross-validation.
In some cases when user does not limit number of PLS components manually, PLS results in following error message:
The problem most probably is related to singular matrix problem in SIMPLS algorithm in one of the cross-validation step. Check if number of components is adjusted correctly in cross-validation.