stepp.test {stepp}

The constructor to generate a complete steppes object with effect estimates and test statistics

Description

This is a constructor function for the steppes object. In addition, it estimates all the effects of each subpopulations, performs permutation tests and generates all covariance matrices and statistics.

Usage

  stepp.test(subpop, model, nperm, showstatus = TRUE)

Arguments

subpop

the filled stepp subpopulation object

model

the model of the data to be used for stepp analysis

nperm

number of permutation used in the permutation test

showstatus

display the progress bar for the permutation test; default is TRUE

Details

Permutation tests for all the statistics are done (see ref below). For best results, considering using 2500 permutations to obtain a rich distribution from which to draw inference.

Value

It returns a steppes object with all effects estimates, covariance matrices and statistics.

Author(s)

Wai-ki Yip

References

Bonetti M, Zahrieh D, Cole BF, Gelber RD. A small sample study of the STEPP approach to assessing treatment-covariate interactions in survival data. Statistics in Medicine 2009; 28(8):1255-68.

See Also

stwin, stsubpop, stmodelKM, stmodelCI, stmodelGLM, steppes, stmodel, stepp.win, stepp.subpop, stepp.KM, stepp.CI, stepp.GLM, estimate, generate

Examples

data(bigCI)

rxgroup <- bigCI$trt
time    <- bigCI$time
evt     <- bigCI$event
cov     <- bigCI$ki67

#
# using constructor functions
swin    <- stepp.win(type="sliding", r1=50, r2=150)
subp    <- stepp.subpop(swin=swin, cov=cov)
summary(subp)

smodel  <- stepp.CI(coltrt=rxgroup, trts=c(1,2), coltime=time, coltype=evt, timePoint=4)
# Warning: In this example, the permutations have been set to 0 to allow the function
# to finish in a short amount of time.  IT IS RECOMMEND TO USE AT LEAST 2500 PERMUTATIONS TO 
# PROVIDE STABLE RESULTS.
statCI  <- stepp.test(subpop=subp, model=smodel, nperm=0)