FH.test: Fleming-Harrington weighted log-rank tests

In lilywang1988/IAfrac: Tools for weighted log-rank tests

Description

Calculating the Fleming-Harrington weighted log-rank tests

Usage

FH.test(survival, delta, trt, rho, gamma)

Arguments

survival	Time to event or censoring.
delta	Event indicators.
trt	Treatment assignment indicator with 1 denoting the treated group, and 0 denoting the placebo group.
rho	First power parameter for the Fleming-Harrington weight which weighs on the early departures: S(t^-)^ρ(1-S(t^-))^γ.
gamma	Second power parameter for the Fleming-Harrington weight which weighs on the late departures: S(t^-)^ρ(1-S(t^-))^γ.

Value

A list 3 different components

O1	Observed number of weighted events (with a multiplication of corresponding weights) in the treatment arm.
E1	Expected number of weighted events (with a multiplication of corresponding weights) in the treatment arm.
Z	Weighted log-rank test statistic.

Author(s)

Lili Wang

See Also

WLR.test

Examples

# Examples for FH.test and WLR.test
set.seed(12345)
data_temp<- nphsim(nsim=1,lambdaC=log(2)/6, lambdaE = c(log(2)/6,log(2)/6*0.7), ssC=250, intervals = c(2),ssE=250, gamma=500/14, R=14, eta=1e-5, fixEnrollTime = TRUE)$simd
data_final<-data.trim.d(100,data_temp)[[1]]
rho=1
gamma=0
# compare the 3 different ways below:
#library(survival)
sqrt(survdiff(Surv(survival,delta)~trt, data =data_final,rho=rho)$chisq)
FH.test(survival=data_final$survival,delta=data_final$delta,trt=data_final$trt,rho=rho,gamma=gamma)
WLR.test(survival=data_final$survival,delta=data_final$delta,trt=data_final$trt,w=function(...){survKM_minus(...)^rho*(1-survKM_minus(...))^gamma})

lilywang1988/IAfrac documentation built on March 11, 2021, 11:53 a.m.