wlogr2 {WWR} | R Documentation |
This will calculate the log-rank and Gehan statistics along with their variances
wlogr2(y, d, z, wty = 1)
y |
a vector of observed event times |
d |
a vector of event indicators with 1=event and 0=censored |
z |
a vector of group indicators with 1=treatment and 0=control |
wty |
a vector of weight indicators with 1=Gehan and 2=log-rank |
wty |
Type of statistics, 1=Gehan, 2=log-rank |
stat |
value of the stat |
vstat |
estimated variance |
tstat |
standardized test stat |
pstat |
2-sided p-value of the standardized test stat |
This provides Gehan test that is usually ignored
Xiaodong Luo
Gehan E.A. 1965. A generalized Wilcoxon test for comparing arbitrarily single-censored samples. Biometrika, 53, 203-223.
Peto R. and Peto J. 1972. Asymptotically Efficient Rank Invariant Test Procedures. Journal of the Royal Statistical Society, Series A, 135, 185-207.
n<-300 b<-0.2 bc<-1.0 lambda0<-0.1;lambdac0<-0.09 lam<-rep(0,n);lamc<-rep(0,n) z<-rep(0,n) z[1:(n/2)]<-1 lam<-lambda0*exp(-b*z) lamc<-lambdac0*exp(-bc*z) tem<-matrix(0,ncol=2,nrow=n) tem[,1]<--log(1-runif(n))/lam tem[,2]<--log(1-runif(n))/lamc y<-apply(tem,1,min) d<-as.numeric(tem[,1]<=y) i<-1 ##i=1,2 wtest<-wlogr2(y,d,z,wty=i) wtest