Skip to contents

Adjust Weights for Censoring and Sample Design with Progress Reporting

Source: R/margot_adjust_weights.R
margot_adjust_weights.Rd

This function calculates and adjusts weights for censoring, combining them with sample weights if provided. It also offers options for trimming and normalising the resulting weights. Progress is reported using the cli package.

Usage

margot_adjust_weights(
  pscore,
  censoring_indicator,
  sample_weights = NULL,
  trim = TRUE,
  normalize = TRUE,
  lower_percentile = 0.01,
  upper_percentile = 0.99,
  na.rm = TRUE
)

Arguments

pscore

Numeric vector of predicted probabilities from a censoring model. Values must be between 0 and 1.

censoring_indicator

Logical vector or 0/1 numeric vector indicating censoring status (TRUE/1 if censored, FALSE/0 if not).

sample_weights

Optional numeric vector of sample weights.

trim

Logical; whether to trim weights (default is TRUE).

normalize

Logical; whether to normalise weights (default is TRUE).

lower_percentile

Numeric; lower percentile for trimming (default is 0.01).

upper_percentile

Numeric; upper percentile for trimming (default is 0.99).

na.rm

Logical; whether to remove NA values (default is TRUE).

Value

A list containing adjusted weights and summary statistics.