stm: Fit a semiparametric transformation model

In sptm: SemiParametric Transformation Model Methods

Description

Fit a semiparametric transformation model

Usage

stm (formula, dat, strata.formula, phase2.ind=NULL, imputation.formula=NULL, 
    family=c("PH","PO","P2"), ee=c("fine2","fine1","kong"), var.est.type=c("1","2"),
    t0, init=NULL, maxit=1000,
    intermediate=FALSE, verbose=FALSE, show.time.elapsed=TRUE)


## S3 method for class 'stm'
 getFixedEf(object, ...)

Arguments

formula	formula. Regression model of interest
dat	data frame.
strata.formula	formula.
phase2.ind	Boolean vector. If TRUE, phase II samples; if FALSE, phase I samples. If NULL, will try to infer from which subjects have phase II variables. Should not be 0/1
imputation.formula	formula. If not NULL, calibration weighting is done
family	string.
ee	string. Type of design matrix used in estimating equation
var.est.type	string. 1: one-stage estimator, 2: two-stage estimator
t0	numeric. Should be close to the end of study time
init	numerical vector.
maxit	integer. Maximum number of iterations in the optimization process
intermediate	Boolean.
verbose	Boolean.
show.time.elapsed	Boolean.
object	an object of type stm
...	additional arguments

Details

Fit stm both with and without calibration. Calls stm.internal.

Value

An object of type stm

Examples

n=100
beta= c(log(.5), log(.7), log(1.2)) 
t0=2.9999
init = c(log(0.0373*t0),beta)        
dat = sim.fong(n, family="PH", beta, random.censoring="0", design="CC", auxiliary="weak", seed=1)

est = stm(formula=Surv(X,d) ~ z + s + z:s, dat, strata.formula=~d, family="PH", t0=t0, init=init,
    var.est.type="1", verbose=3)

sptm documentation built on Dec. 1, 2019, 1:08 a.m.