efronlim: Constrained Kaplan-Meier Estimator

In BuyseTest: Generalized Pairwise Comparisons

Constrained Kaplan-Meier Estimator

Description

Kaplan-Meier estimator, possibly stratified, constrained to 0 just after end of follow-up in each strata.

Usage

efronlim(formula, data, ...)

Arguments

formula	formula with on the right-hand side Hist(time,event) or Surv(time,event) and 1 or stratification variable(s) on the right-hand side.
data	A data.frame containing the variables of present in argument formula.
...	additional arguments passed to prodlim::prodlim

Details

If in any strata there is censoring at the observed time, then the dataset is updated by setting one of the censored observations to an infinitesimal later timepoint with an event instead of censoring. Then the possibly stratified Kaplan-Meier estimator is run on this updated dataset.

Value

A prodlim object.

Examples

library(data.table)

set.seed(1)
dt.data <- simBuyseTest(1e2, n.strata = 2)
dt.data$time1 <- pmin(dt.data$eventtime, 1)
dt.data$status1 <- dt.data$status * (dt.data$eventtime<=1)

## KM
if(require(prodlim)){
   e.KM <- prodlim(Hist(time1,status1)~1, data = dt.data)
   plot(e.KM)
}

e.KM0 <- efronlim(Hist(time1,status1)~1, data = dt.data)
plot(e.KM0)

## stratfied KM
if(require(prodlim)){
   e.KMS <- prodlim(Hist(time1,status1)~strata, data = dt.data)
   plot(e.KMS)
}

e.KMS <- efronlim(Hist(time1,status1)~strata, data = dt.data)
plot(e.KMS)

BuyseTest documentation built on June 8, 2025, 11:19 a.m.