Calculate robust M-scores for numeric variables

Calculate median/MAD-based M-scores for selected numeric variables and return both the transformed data and the parameters needed to review or reuse the transformation.

Usage

CreateMScoreObject(
  df,
  variables = NULL,
  names_prefix = "M_",
  RetainLabels = TRUE,
  RenameLabels = TRUE,
  center = TRUE,
  scale = TRUE,
  constant = 1.4826
)

Arguments

df

A data frame.

variables

Character vector of numeric variables to transform. If NULL, numeric variables are detected with getNumVars().

names_prefix

Prefix for generated M-score columns.

RetainLabels

Logical; keep variable labels when possible.

RenameLabels

Logical; rename generated labels when labels are retained.

center

Logical; subtract the median before scaling.

scale

Logical; divide by the median absolute deviation.

constant

Scaling constant passed to stats::mad().

Value

An object of class "MScoreObj", a list with:

• MScores: data frame of M-score variables only

• DataWithM: original df plus M-score variables

• Parameters: data frame with Variable, N, Median, and MAD

• Center: logical flag used

• Scale: logical flag used

• Constant: MAD scaling constant used