Skip to contents

This is the S3 specialization of the function stats::prcomp() for QTS samples.


# S3 method for qts_sample
prcomp(x, M = 5, fit = FALSE, ...)



An object of class qts_sample.


An integer value specifying the number of principal component to compute. Defaults to 5L.


A boolean specifying whether the resulting prcomp_qts object should store a reconstruction of the sample from the retained PCs. Defaults to FALSE.


Arguments passed to or from other methods.


An object of class prcomp_qts which is a list with the following components:

  • tpca: An object of class MFPCAfit as produced by the function MFPCA::MFPCA(),

  • var_props: A numeric vector storing the percentage of variance explained by each PC,

  • total_variance: A numeric value storing the total variance of the sample,

  • mean_qts: An object of class qts containing the mean QTS (used for centering the QTS sample before projecting it to the tangent space),

  • principal_qts: A list of qtss containing the required principal components.


The mean_qts component of the resulting object is the QTS used for centering. It it part of the prcomp_qts object because it is needed to reconstruct the sample from the retained PCs. The prcomp_qts object also contains the total variance of the sample and the percentage of variance explained by each PC.


res_pca <- prcomp(vespa64$igp[1:16])
#>  The maximum number of principal component is 15.