Apply a fitted normalization to a numeric vector

Given a normalize_x_fit object produced by find_best_normalization(), re-winsorizes x using the same settings, applies the fitted transformation, and standardizes using the fitted center and scale.

Using the fitted ECDF / parameters rather than re-fitting ensures that new observations (e.g. from a different year) are scored on the same scale as the training data.

Usage

apply_normalization(x, fit, verbose = FALSE)

Arguments

x

Numeric vector to score.

fit

A normalize_x_fit object returned by find_best_normalization().

verbose

Logical. Print a brief summary. Default FALSE.

Value

A numeric vector of the same length as x with transformation metadata attached as attributes (transform_type, transform_label, transform_note, guess_reason, center, scale, range, winsorize, offset, fit_object).

See also

find_best_normalization(), normalize_x()

Examples

x_train <- c( rnorm(200), rep(0, 30) )
x_new   <- rnorm(50)

# fit on training data, score new data with same parameters
fit  <- find_best_normalization( x_train, range = "np", verbose = FALSE )
z    <- apply_normalization( x_new, fit = fit )