IQR-Based Top/Bottom Coding for a Single Variable

Internal utility that applies top and/or bottom coding to a numeric vector using an IQR rule on the log() transformation.

Usage

iqr_top_bottom_coding(
  x,
  w = NULL,
  times = 3,
  type = c("type_4", "type_2"),
  one_sided = NULL
)

Arguments

x

Numeric vector. Variable to be top/bottom coded.

w

Optional numeric vector of weights.

times

Numeric. Multiplier for IQR to determine bounds (default is 3).

type

Character. Quantile estimation type, passed to compute_weighted_percentiles function (e.g., "type_4", "type_2").

one_sided

Character. "top", "bottom", or NULL (default) for two-sided coding.

Value

A numeric vector with values above or below the threshold replaced (on the original scale).

Examples

if (FALSE) { # \dontrun{
library(lissyrtools)
library(dplyr)

data <- lissyrtools::lissyuse(data = "it16", vars = c("dhi"))
iqr_top_bottom_coding(data$it16h$dhi)
iqr_top_bottom_coding(data$it16h$dhi, data$it16h$hwgt, one_sided = "top")
} # }