PcaHubert-class {rrcov}R Documentation

Class "PcaHubert" - ROBust method for Principal Components Analysis

Description

The ROBPCA algorithm was proposed by Hubert et al (2005) and stays for 'ROBust method for Principal Components Analysis'. It is resistant to outliers in the data. The robust loadings are computed using projection-pursuit techniques and the MCD method. Therefore ROBPCA can be applied to both low and high-dimensional data sets. In low dimensions, the MCD method is applied.

Objects from the Class

Objects can be created by calls of the form new("PcaHubert", ...) but the usual way of creating PcaHubert objects is a call to the function PcaHubert which serves as a constructor.

Slots

alpha:
Object of class "numeric" the fraction of outliers the algorithm should resist - this is the argument alpha
quan:
Object of class "numeric" The quantile h used throughout the algorithm
call:
Object of class "language" The (matched) function call.
center:
Object of class "vector" the center of the data
loadings:
Object of class "matrix" the matrix of variable loadings (i.e., a matrix whose columns contain the eigenvectors)
eigenvalues:
Object of class "vector" the eigenvalues
scores:
Object of class "matrix" the scores - the value of the projected on the space of the principal components data (the centred (and scaled if requested) data multiplied by the loadings matrix) is returned. Hence, cov(scores) is the diagonal matrix diag(eigenvalues)
k:
Object of class "numeric" number of (choosen) principal components
sd:
Object of class "Uvector" Score distances within the robust PCA subspace
od:
Object of class "Uvector" Orthogonal distances to the robust PCA subspace
cutoff.sd:
Object of class "numeric" Cutoff value for the score distances
cutoff.od:
Object of class "numeric" Cutoff values for the orthogonal distances
flag:
Object of class "Uvector" The observations whose score distance is larger than cutoff.sd or whose orthogonal distance is larger than cutoff.od can be considered as outliers and receive a flag equal to zero. The regular observations receive a flag 1
n.obs:
Object of class "numeric" the number of observations

Extends

Class "PcaRobust", directly. Class "Pca", by class "PcaRobust", distance 2.

Methods

getQuan
signature(obj = "PcaHubert"): Returns the quantile used throughout the algorithm

Author(s)

Valentin Todorov valentin.todorov@chello.at

See Also

PcaRobust-class, Pca-class, PcaClassic, PcaClassic-class

Examples

showClass("PcaHubert")

[Package rrcov version 0.5-01 Index]