sPCA_rSVD {ltsspca}

Sparse Principal Component Analysis via Regularized Singular Value Decompsition (sPCA-rSVD)

Description

the function that computes sPCA_rSVD

Usage

sPCA_rSVD(x, k, method = "hard", center = FALSE, scale = FALSE,
  l.search = NULL, ls.min = 1)

Arguments

x

the input data matrix

k

the maximal number of PC's to seach for in the initial stage

method

threshold method used in the algorithm; If method = "hard" (defauls), the hard threshold function is used; if method = "soft", the soft threshold function is used; if method = "scad", the scad threshold function is used

center

if center = TRUE the data will be centered by the columnwise means; default is center = FALSE

scale

if scale = TRUE the data will be scaled by the columnwise standard deviations; default is scaled = FALSE

l.search

a list of length kmax which contains the search grids chosen by the user; default is NULL

ls.min

the smallest grid step when searching for the sparsity of each PC; default is 1

Value

an object of class "sPCA_rSVD" is returned

loadings

the sparse loading matrix estimated with sPCA_rSVD

scores

the estimated score matrix

eigenvalues

the estimated eigenvalues

spca.it

the list that contains the results of sPCA_rSVD when searching for the individual PCs

ls

the list that contains the final search grid for each PC direction

References

Shen, H. and Huang, J. (2008), “Sparse principal component anlysis via regularized low rank matrix decomposition”, Journal of Multivariate Analysis, 99, 1015–1034.

Shen, D., Shen, H., and Marron, J. (2013). “Consistency of sparse PCA in high dimensional low sample size context”, Journal of Multivariate Analysis, 115, 315–333.

Examples

## Not run: 
nonrobM <- sPCA_rSVD(x = x, k = 2, center =  T, scale = F)

## End(Not run)