rcens Package

This package provides functions to generate censored samples of type I, II and III, from any random sample generator. It also provides the option to create left and right censorship Along with this, the generation of samples with interval censoring is in the testing phase. With two options of fixed length intervals and random lengths.

Installation

You can install rcens from CRAN via:

install.packages("rcens")

Or install the latest development version (on GitHub) via {devtools}:

devtools::install_github("dlsaavedra/rcens")

Getting Started

Create easily a new censored data set establishment the percentage of censoring and the original distribution of random variable.

Right Censored Data, type III (Random)

In this example create a right censored data type III and fit Kaplan Meier (survival library).

#Example Exponential - Uniform

devtools::install_github("dlsaavedra/rcens")
library(rcens)
library(survival)

Data = rcensT3(rdistrX = rexp, pdistrC = punif, rdistrC = runif,
                 param_X = list("rate" = 2),
                 param_C = list("min" = 0, "max" = "lambda"),
                 n = 1e02, theta = .5, right = TRUE)

S = Surv(Data$sample_censored,Data$censored_indicator, type = "right")
s1 = survfit(S ~ 1)

CDF_censored = ecdf(Data$sample_censored)
Survival_CDF = Vectorize(function(x){ 1 - CDF_censored(x)})
CDF_original= ecdf(Data$sample_uncensored)
Survival_CDF_original = Vectorize(function(x){ 1 - CDF_original(x)})

plot(Survival_CDF, col = "blue", xlim = c(0,2))
title("Survival Curve")
plot(Survival_CDF_original, col = "red", add= TRUE, xlim = c(0,2))
lines(s1$time, s1$surv, col = "green", xlim = c(0,2))
legend("topright",c("original","censured", "Survival_KM"),
       col = c("red", "blue", "green"), lty = 1)

Other Examples

In the folder examples_plot you can find example for each function in this packages.

Citation

To cite rcens in publication use:

Saavedra D, Ramos PL (2024). rcens: Generate Sample Censoring. R package version 0.1.0, https://github.com/dlsaavedra/rcens.

A BibTeX entry for LaTeX users is:

  @Manual{,
    title = {rcens: Generate Sample Censoring},
    author = {Daniel Saavedra and Pedro L Ramos},
    year = {2024},
    note = {R package version 0.1.1},
    url = {https://github.com/dlsaavedra/rcens},
  }

Future Work

We are currently working to generate data with interval right censoring, both with fixed and random interval length. Soon we will propose a function to generate data with double interval censoring. All this while maintaining the possibility of using any distribution for the original data.

Complex censoring extension: rcenscomp

Version 0.2.0 adds an independent family of generators whose names begin with rcenscomp. The original functions and their behavior are retained. The new API separates complete latent event times from the mechanism that determines what is observed and returns a consistent object containing observed data, optional latent values, design information, and diagnostics.

Generate latent lifetimes internally and apply Hybrid Type-I censoring:

set.seed(2026)
dat <- rcenscomp(
  scheme = "hybrid_type1",
  rdistrX = rexp,
  param_X = list(rate = 2),
  n = 100,
  T = 1,
  r = 60
)

dat
as.data.frame(dat)

Apply an inspection schedule to an existing latent sample:

x <- c(0.10, 0.35, 0.75, 1.40)
rcenscomp_interval(x, visits = c(0.20, 0.50, 1.00))

The extension covers interval and current-status observations, delayed entry, Hybrid Type-I and Type-II designs, generalized hybrid censoring, progressive designs, covariate-dependent censoring, shared-process frailty, and competing risks. Weibull helpers use shape alpha and rate-like beta, equivalent to stats::rweibull(..., shape = alpha, scale = beta^(-1 / alpha)).