mixOmics (version 5.2.0)

nipals: Non-linear Iterative Partial Least Squares (NIPALS) algorithm

Description

This function performs NIPALS algorithm, i.e. the singular-value decomposition (SVD) of a data table that can contain missing values.

Usage

nipals(X, ncomp = 1, reconst = FALSE, max.iter = 500, tol = 1e-09)

Arguments

X

real matrix or data frame whose SVD decomposition is to be computed. It can contain missing values.

ncomp

integer, the number of components to keep. If missing ncomp=ncol(X).

reconst

logical that specify if nipals must perform the reconstitution of the data using the ncomp components.

max.iter

integer, the maximum number of iterations.

tol

a positive real, the tolerance used in the iterative algorithm.

Value

• The returned value is a list with components:
• eig
  vector containing the pseudosingular values of X, of length ncomp.
• t
  matrix whose columns contain the left singular vectors of X.
• p
  matrix whose columns contain the right singular vectors of X.
• rec
  matrix obtained by the reconstitution of the data using the ncomp components.

encoding

latin1

Details

The NIPALS algorithm (Non-linear Iterative Partial Least Squares) has been developed by H. Wold at first for PCA and later-on for PLS. It is the most commonly used method for calculating the principal components of a data set. It gives more numerically accurate results when compared with the SVD of the covariance matrix, but is slower to calculate.

This algorithm allows to realize SVD with missing data, without having to delete the rows with missing data or to estimate the missing data.

References

Tenenhaus, M. (1998). La regression PLS: theorie et pratique. Paris: Editions Technic.

Wold H. (1966). Estimation of principal components and related models by iterative least squares. In: Krishnaiah, P. R. (editors), Multivariate Analysis. Academic Press, N.Y., 391-420.

Wold H. (1975). Path models with latent variables: The NIPALS approach. In: Blalock H. M. et al. (editors). Quantitative Sociology: International perspectives on mathematical and statistical model building. Academic Press, N.Y., 307-357.

See Also

svd, princomp, prcomp, eigen and http://www.mixOmics.org for more details.

Examples

## Hilbert matrix
hilbert <- function(n) { i <- 1:n; 1 / outer(i - 1, i, "+") }
X.na <- X <- hilbert(9)[, 1:6]

## Hilbert matrix with missing data
idx.na <- matrix(sample(c(0, 1, 1, 1, 1), 36, replace = TRUE), ncol = 6)
X.na[idx.na == 0] <- NA
X.rec <- nipals(X.na, reconst = TRUE)$rec
round(X, 2)
round(X.rec, 2)