Normalize numeric and factor variables to a 0–1 scale

This function rescales variables to the 0 to 1 range. It supports flipping of scales and preserves variable labels.

Usage

normalize(
  data,
  vars = everything(),
  flip = FALSE,
  keep_all = TRUE,
  outside_scale = c("Don't know", "Prefer not to say")
)

Arguments

data

A data frame containing variables to normalize.

vars

Variables to normalize. Supports tidyselect syntax (e.g., starts_with(), c(var1, var2)).

flip

Logical; if TRUE, reverses the direction of valid response levels before scaling.

keep_all

Logical; if TRUE, returns full dataset; if FALSE, returns only normalized variables.

outside_scale

Character vector of response levels (typically DK/PNTS) that lie outside of the scale and should be treated as neutral midpoint (0.5).

Value

A data frame with normalized variables.

• If keep_all = TRUE, the full dataset is returned with normalized variables.

• If keep_all = FALSE, only normalized variables are returned.

Details

• Numeric variables are min–max scaled to 0 to 1.

• Factor variables are converted to numeric order and evenly spaced between 0 and 1.

• Levels listed in outside_scale (e.g., Don't know) are assigned a value of 0.5.

Examples

# Normalize Q1 and Q2 with flipping the scale order
normalize(data = data, vars = c(Q2, Q3), flip = TRUE)
#> Error: These packages are required but not installed: tidyselect, dplyr