predband.scproc: Prediction bands for 'scproc' object

In goftte: Goodness-of-Fit for Time-to-Event Data

Description

Calculates the prediction bands under the null hypothesis.

Usage

## S3 method for class 'scproc'
predband(x, idx=1:length(x$variable), level=0.95, cval=NULL, ...)
       

Arguments

x	An object of class scproc.
idx	The index number of covariates whose the prediction band is calculated.
level	The required level for prediction bands.
cval	Parameter used to calculate the symmetric prediction bands defined by the standard error multiplied by cval.
...	additional arguments.

Details

This function is derived from confint.cumres of gof package.

Value

t	Ordered values of the variable used : unique times for proportionality and unique observations of covariates for functional form.
yu	Upper simultaneous confidence limit.

Author(s)

Patrick Sfumato and Jean-Marie Boher.

References

Holst KK (2014). gof: Model-diagnostics based on cumulative residuals. R package version 0.9.1, URL https://CRAN.R-project.org/package=gof.

Examples

require(cmprsk)

#Simulating survival data with competing events
set.seed(10)
ftime <- rexp(200)
fstatus <- sample(0:2,200,replace=TRUE)
cov <- matrix(runif(200),nrow=200)

# Fine & Gray regression
fit.crr <- crr(ftime,fstatus,cov)

#Checking the covariates functional form assumption
k<-fcov(model=fit.crr, ftime=ftime,fstatus=fstatus,cov1=cov)

#Prediction bands
predband(k)

goftte documentation built on May 2, 2019, 5:12 a.m.