riskRegression (version 2022.03.22)

subjectWeights: Estimation of censoring probabilities at subject specific times

Description

This function is used internally to contruct pseudo values by inverse of the probability of censoring weights.

Usage

subjectWeights(
  formula,
  data,
  method = c("cox", "marginal", "km", "nonpar", "forest", "none"),
  args,
  lag = 1
)

Value

times

The times at which weights are estimated

weights

Estimated weights at individual time values subject.times

lag

The time lag.

fit

The fitted censoring model

method

The method for modelling the censoring distribution

call

The call

Arguments

formula

A survival formula like, Surv(time,status)~1 or Hist(time,status)~1 where status=0 means censored. The status variable is internally reversed for estimation of censoring rather than survival probabilities. Some of the available models, see argument model, will use predictors on the right hand side of the formula.

data

The data used for fitting the censoring model

method

Censoring model used for estimation of the (conditional) censoring distribution.

args

Arguments passed to the fitter of the method.

lag

If equal to 1 then obtain G(T_i-|X_i), if equal to 0 estimate the conditional censoring distribution at the subject.times, i.e. (G(T_i|X_i)).

Author

Thomas A. Gerds

Details

Inverse of the probability of censoring weights usually refer to the probabilities of not being censored at certain time points. These probabilities are also the values of the conditional survival function of the censoring time given covariates. The function subjectWeights estimates the conditional survival function of the censoring times and derives the weights.

IMPORTANT: the data set should be ordered, order(time,-status) in order to get the weights in the right order for some choices of method.

Examples

library(prodlim)
library(survival)
dat=SimSurv(300)

dat <- dat[order(dat$time,-dat$status),]

# using the marginal Kaplan-Meier for the censoring times

WKM=subjectWeights(Hist(time,status)~X2,data=dat,method="marginal")
plot(WKM$fit)
WKM$fit
WKM$weights

# using the Cox model for the censoring times given X2

WCox=subjectWeights(Surv(time,status)~X2,data=dat,method="cox")
WCox
plot(WCox$weights,WKM$weights)

# using the stratified Kaplan-Meier for the censoring times given X2

WKM2 <- subjectWeights(Surv(time,status)~X2,data=dat,method="nonpar")
plot(WKM2$fit,add=FALSE)