SurvFn: Arm-specific survival functions

In nchenderson/DelayedSurvFit: Constrained Survival Estimation with Delayed Treatment

Arm-specific survival functions

Description

Given a fitted object from either DelayedSurvFit or DelayedHazFit, this will return a step function which represents the estimated survival function for either the control arm or active treatment arm.

Usage

SurvFn(obj, arm = 1)

Arguments

obj	An object of class surv.delay
arm	The treatment arm; arm = 0 for the control arm; arm = 1 for the active treatment arm

Value

A function of class "stepfun"

Author(s)

Nicholas Henderson

Examples

### Generate simulated survival data
## Not run: 
n <- 400
cens.time <- 8
X1 <- rweibull(n, shape=2.3, scale=7)
X2 <- rweibull(n, shape=.8, scale=10)
Y1 <- pmin(X1, cens.time)
Y2 <- pmin(X2, cens.time)
e1 <- as.numeric(X1 < cens.time)
e2 <- as.numeric(X2 < cens.time)

times <- c(Y1, Y2)  ## follow-up times
events <- c(e1, e2) ## event indicators
trt <- rep(c(0,1), each=n)  ## treatment arm assignment

dfs.obj <- DelayedSurvFit(times, events, trt, max.times=50) 

S1 <- SurvFn(dfs.obj, arm=1) # active treatment arm survival function
S0 <- SurvFn(dfs.obj, arm=0) # control treatment arm survival function

## Difference in survival function at time 2
S1(2) - S0(2)

## Difference in survival function at time 5
S1(5) - S0(5)

## End(Not run)

nchenderson/DelayedSurvFit documentation built on Aug. 18, 2024, 11:49 a.m.