nonzerocure_test: Non-parametric pest for a non-zero cured fraction

In hdcuremodels: Penalized Mixture Cure Models for High-Dimensional Data

Non-parametric pest for a non-zero cured fraction

Description

Tests the null hypothesis that the proportion of observations susceptible to the event = 1 against the alternative that the proportion of observations susceptible to the event is < 1. If the null hypothesis is rejected, there is a significant cured fraction.

Usage

nonzerocure_test(object, Reps = 1000, seed = NULL, plot = FALSE, B = NULL)

Arguments

object	a survfit object.
Reps	number of simulations on which to base the p-value (default = 1000).
seed	optional random seed.
plot	logical. If TRUE a histogram of the estimated susceptible proportions over all simulations is produced.
B	optional. If specified the maximum observed time for the uniform distribution for generating the censoring times. If not specified, an exponential model is used for generating the censoring times (default).

Value

proportion_susceptible	estimated proportion of susceptibles
proportion_cured	estimated proportion of those cured
p.value	p-value testing the null hypothesis that the proportion of susceptibles = 1 (cured fraction = 0) against the alternative that the proportion of susceptibles < 1 (non-zero cured fraction)
time_95_percent_of_events	estimated time at which 95% of events should have occurred

References

Maller, R. A. and Zhou, X. (1996) Survival Analysis with Long-Term Survivors. John Wiley & Sons.

See Also

survfit, cure_estimate, sufficient_fu_test

Examples

library(survival)
set.seed(1234)
temp <- generate_cure_data(N = 100, J = 10, nTrue = 10, A = 1.8)
training <- temp$Training
km.fit <- survfit(Surv(Time, Censor) ~ 1, data = training)
nonzerocure_test(km.fit)

hdcuremodels documentation built on June 22, 2024, 10:17 a.m.