calc_pls_args: Calculate PLSR - Arguments

Description Arguments Details Note See Also Examples


The following parameters can be used in the ... argument in function getap, also within function gdmm, to override the values in the analysis procedure file and so to modify the calculation of PLSR models - see examples.

gdmm(dataset, ap=getap(...))



Logical. If used in getap, if PLSR models should be calculated with a given dataset.


NULL or character vector. Which variables should be used to regress on. Set to NULL for using all numerical variables to regress on, or provide a character vector with the column names of numerical variables to use those for regression in the PLSR.


NULL or integer length one. The number of components used in PLSR. Set to NULL for automatic detection, or provide an integer to use this number of components in the PLSR.


Character. Which crossvalidation to use. Possible values are:

  • "def" Read in the default value from settings.r (parameter plsr_calc_typeOfCrossvalid)

  • A numeric length one for this n-fold crossvalidation. The default is to always exclude resp. include consecutive scans together.

  • A valid name of a class variable for performing a crossvalidation based on the grouping defined by this variable. For a class variable containing e.g. four different levels, a 4-fold crossvalidation with always all members of one group being excluded is performed. This is overruling any grouping that would result from the consecutive scans, please see below.

  • "LOO" for a leave-one-out crossvalidation

If a vector with the same length as the vector in pls.regOn is provided, each element of pls.valid is used for crossvalidating the corresponding element in pls.regOn. Any of the above mentioned input types can be mixed, so the input could be e.g. pls.valid <- c("C_FooBar", 10, "C_BarFoo", 10). The corresponding pls.regOn input for this would then be e.g. pls.regOn <- c("Y_FooBar", "Y_FooBar", "Y_BarFoo", "Y_BarFoo"). Please note that via the parameter plsr_calc_CV_consecsTogether in the settings file you can select if for crossvalidation the consecutive scans (i.e. the scans with the same sample number) should always be excluded or included together. The default is to always exclude resp. include the consecutive scans of a single sample together.


Logical. If a plsr-specific box-plot based outlier-detection algorithm should be used on the data of a first plsr model to determine the outliers that then will be excluded in the final plsr model. Possible values are:

  • "def" Read in the default value from settings.r (parameter plsr_calc_excludePlsrOutliers)

  • TRUE for excluding plsr specific outliers

  • FALSE for not performing the plsr specific outlier exclusion

If a vector with the same length as the vector in pls.regOn is provided, each element of pls.exOut is used to perform the corresponding outlier-detection (or not) for each element in pls.regOn.


For a list of all parameters that can be used in the ... argument in getap and in the plot functions please see anproc_file.


Calculation of PLSR models is done with the function plsr.

See Also


Other Calc. arguments: calc_NNET_args, calc_SVM_args, calc_aqg_args, calc_discrimAnalysis_args, calc_pca_args, calc_randomForest_args, calc_sim_args, split_dataset

Other PLSR documentation: plot_pls,aquap_cube-method, plot_pls_args, plot_pls_indepPred


## Not run: 
 dataset <- gfd()
 cube <- gdmm(dataset, ap=getap(pls.regOn="Y_Temp"))
 cube <- gdmm(dataset, ap=getap(pls.ncomp=5, pls.regOn="Y_Foo"))

## End(Not run)

bpollner/aquap2 documentation built on Jan. 30, 2019, 9:08 a.m.