btvarCP2=function(time,cens,t1,t2,nbt,s0=0,alpha=0.05)
{
##############################################
# bootstrap time and cens to obtain          #
# an estimate of the variance                #
# and the confidence intervals for the       #
# probability to fail beetween t1 and t2     #
#    from the event coded as 1 in cens       #
#    knowing that the patient is free        #
#    of any event at time t1                 #
# time=time to first failure                 #
# cens=1 for event of interest               #
#      2 for competing risk                  #
#      0 for no event                        #
# t1 and t2 the time window for which        #
#   the conditional probability is calculated#
# nbt=number of bootstrap repetitions        #
# s0=the seed (for exact repliaction)        #
##############################################
n=length(time)
bt=function()
 {
idx=sample(c(1:n),size=n,replace=T) 
tbt=time[idx]
cbt=cens[idx]
fitbt=cuminc(tbt,cbt)
sumbt=timepoints(fitbt,times=c(t1,t2))
cp2=(sumbt$est[1,2]-sumbt$est[1,1])/(1-sumbt$est[2,1]-sumbt$est[1,1])
return(cp2)
 }
CP2=vector("numeric",nbt)
set.seed(s0)
for (i in 1:nbt)  CP2[i]=bt()
varCP2=var(CP2)
fit=cuminc(time,cens)
sumfit=timepoints(fit,times=c(t1,t2))
cpw=(sumfit$est[1,2]-sumfit$est[1,1])/(1-sumfit$est[2,1]-sumfit$est[1,1])
z=qnorm(1-alpha/2)
A=z*sqrt(varCP2)/(cpw*log(cpw))
loA=exp(-A)
upA=exp(A)
lo=cpw^loA
up=cpw^upA
r=data.frame(CPw=cpw,varCPw=varCP2,"L95%CI"=lo,"U95%CI"=up)
row.names(r)=""
return(r)
}