Pareto Scaling Leaves variables unscaled. Optional centering.

Pareto Scaling Leaves variables unscaled. Optional centering.

Usage

pareto_scaling(center = FALSE)

Arguments

center

Logical. If TRUE, variables are mean-centered before scaling.

Value

A list with elements:

X

Numeric matrix containing the scaled data.

prep

List describing the preprocessing, including centering and scaling parameters (center, scale, X_mean, X_sd).

Details

Scales variables by the square root of their standard deviation.

See also

Other scaling_strategies: unscaled(), uv_scaling()

Examples

paritalUV <- pareto_scaling(center=TRUE)
X <- matrix(c(10,10, 0,0, 0, 10, 0, 1000), ncol=4)
X_scaled <- prep_X(paritalUV, X)
str(X_scaled)
#> List of 2
#>  $ X   : num [1:2, 1:4] 0 0 0 0 -1.88 ...
#>  $ prep:List of 4
#>   ..$ center: logi TRUE
#>   ..$ scale : chr "Pareto"
#>   ..$ X_mean: num [1:4] 10 0 5 500
#>   ..$ X_sd  : num [1:4] 0 0 7.07 707.11
X_scaled$X
#>      [,1] [,2]      [,3]      [,4]
#> [1,]    0    0 -1.880302 -18.80302
#> [2,]    0    0  1.880302  18.80302