coxseisim {coxsei}

A function to simulate a CoxSEI process conditional on specified covariate values

Description

simulate the sample path of the CoxSEI model with given covariate process values, and excitation function and order of autodependence in the excitation term.

Usage

coxseisim(parreg, parg, lmd0 = function(tt) (1 + 0.5 * cos(2 * pi *
tt)),
          g = function(x, parg) {
                 ifelse(x <= 0, 0, parg[1] * parg[2] * exp(-parg[2] * x))
              },
          censor = 1, m = 2, trace=TRUE,
          Z = function(x) matrix(0, length(x), length(parreg))
         )

Arguments

parreg

the regression parameter

parg

parameters of the excitation function

lmd0

the baseline intensity function

g

the excitation function

censor

the censoring time

m

order of autoregression in the excitation component of the intensity process

trace

whether to trace the data generation process; defaults to TRUE

Z

a function to calculate the covariate values at a specified event time

Value

A data frame with provided covariate values and the censoring time, and the generated event times.

Author(s)

Feng Chen <[e-mail]>

Examples

    n.smp <- 100;
    z <- matrix(NA,n.smp,3)
    for(i in 1:n.smp)
    z[i,] <- round(c(runif(1,0.5,1.5),runif(1,1.5,2.5),rbinom(1,1,0.5)),2)
    dat <- coxseisim(1:3*0.2,c(0.07,10),censor=rlnorm(1,0,0.1),m=2,
    Z=function(x)matrix(z[1,],length(x),3,byrow=TRUE))
    dat$id <- 1;
    for(i in 2:n.smp){
      dattmp <- coxseisim(1:3*0.2,c(0.07,10),censor=rlnorm(1,0,0.1),m=2,
      Z=function(x)matrix(z[i,],length(x),3,byrow=TRUE))
      dattmp$id <- i;
      dat <- rbind(dat,dattmp)
    }