rrcov3way (version 0.1-18)

do3Scale: Centering and scaling

Description

Centering and/or normalization of a three way array or a matricized array across one mode (modes indicated by "A", "B" or "C").

Usage

# S3 method for tucker3
do3Scale(x, renorm.mode = c("A", "B", "C"), …)
    # S3 method for parafac
do3Scale(x, renorm.mode = c("A", "B", "C"), …)
    # S3 method for default
do3Scale(x, center = FALSE, scale = FALSE, 
        center.mode = c("A", "B", "C", "AB", "AC", "BC", "ABC"), 
        scale.mode = c("B", "A", "C"), 
        only.data=TRUE, …)

Arguments

x

Three dimensional array of order (I x J x K) or a matrix (or data.frame coerced to a matrix) of order (I x JK) containing the matricized array (frontal slices)

center

Whether and how to center the data. Can be NULL, logical TRUE or FALSE, function or a numeric vector with length corresponding to the number of columns in the corresponding mode. If center=TRUE, mean() is used; default is center=FALSE.

scale

Whether and how to scale the data. Can be NULL, logical TRUE or FALSE, function or a numeric vector with length corresponding to the number of columns in the corresponding mode. If scale=TRUE, sd() is used; default is scale=FALSE.

center.mode

Across which mode to center. Default is center.mode="A"

scale.mode

Within which mode to scale. Default is scale.mode="B"

renorm.mode

Within which mode to renormalize a Parafac or Tucker3 solution. See in Details how this is performed for the different models. Default is renorm.mode="A"

only.data

Whether to return only the centered/scaled data or also the center and the scale themselves. Default is only.data=TRUE

…

potential further arguments passed to lower level functions.

Value

A named list, consisting of the centered and/or scaled data, a center vector, a scale vector and the mode in which the data were centered/scaled.

References

Kiers, H.A.L. (2000).Towards a standardizrd notationand terminology in multiway analysis. Journal of Chemometrics, 14:105-122.

Kroonenberg, P.M. (1983).Three-mode principal component analysis: Theory and applications (Vol. 2), DSWO press.

Examples

# NOT RUN {
    data(elind)
    (x1 <- do3Scale(elind, center=TRUE, scale=TRUE))
    (x2 <- do3Scale(elind, center=TRUE, scale=TRUE, center.mode="B"))
    (x3 <- do3Scale(elind, center=TRUE, scale=TRUE, center.mode="C", scale.mode="C"))
# }

Run the code above in your browser using DataCamp Workspace