get.pseudodata.cox {icmm}

Obtain pseudodata based on the Cox's regression model.

Description

For Cox's regression model, given the current estimates of regression coefficients, working responses and their corresponding weights are obtained.

Usage

get.pseudodata.cox(Y, X, event, beta, time, ntime, sumevent)

Arguments

Y

an (n*1) numeric matrix of time response.

X

an (n*p) numeric design matrix.

event

an (n*1) numeric matrix of status: of status indicator: 0=right censored, 1=event at time.

beta

a (p*1) matrix of regression coefficients.

time

a vector or sorted value of Y.

ntime

length of the vector time.

sumevent

a vector of size ntime where each element is the sum of event where Y is equal to each value in time.

Value

Return a list including elements

z

an (n*1) matrix of working responses

sigma2

an (n*1) matrix of inverse of weights.

Author(s)

Vitara Pungpapong, Min Zhang, Dabao Zhang

Examples

data(simCox)
Y<-as.matrix(simCox[,1])
event<-as.matrix(simCox[,2])
X<-as.matrix(simCox[,-(1:2)])
time<-sort(unique(Y))
ntime<-length(time)
# sum of event_i where y_i =time_k
sumevent<-rep(0, ntime)
for(j in 1:ntime)
{
  sumevent[j]<-sum(event[Y[,1]==time[j]])
}
# Obtain initial values from lasso
data(initbetaCox)
initbeta<-as.matrix(initbetaCox)
# Get Pseudodata
pseudodata<-get.pseudodata.cox(Y, X, event, initbeta, time, ntime, sumevent)
z<-pseudodata$z
sigma<-sqrt(pseudodata$sigma2)