rARCens {ARCensReg} R Documentation

## Generating censored autoregressive data

### Description

It simulates a censored response variable with autoregressive errors of order p following normal or Student-t innovations, with an established censoring rate.

### Usage

rARCens(n, beta, phi, sig2 = 1, x = rep(1, n), cens = "left",
pcens = 0.1, innov = "norm", nu = NULL)

### Arguments

 n Length of the desired time serie. beta Vector of theoretical regression parameters of length l. phi Vector of theoretical autoregressive coefficients of length p. sig2 Theoretical variance of the error. x Matrix of covariates of dimension nxl (in models that include an intercept x should contain a column of ones). cens 'left' for left censoring, 'right' for right censoring. pcens Desired censoring rate. innov Distribution of the innovation variable. The values are 'norm' and 't' for normal and Student-t distribution, respectively. nu Degrees of freedom for Student-t innovations.

### Value

 data Generated response (y), censoring indicator (cc), and lower (lcl) and upper (ucl) bounds of the interval, which contains the true value of the censored observation. param Theoretical parameters (beta, sig2, phi).

### Note

For data generation with Student-t innovations, the first p observations are not censored.

### Author(s)

Fernanda L. Schumacher, Katherine L. Valeriano, Victor H. Lachos, Christian E. Galarza, and Larissa A. Matos

### Examples

library(ggplot2)

## Example 1: Generating a sample with normal innovations
set.seed(1234)
dat = rARCens(n=100, beta=c(1,-1), phi=c(.48,-.2), sig2=.5,
x=cbind(1,runif(100)), cens='left', pcens=.10)

# Plotting the time serie
ggplot(data.frame(dat$data$y), aes(x=1:100, y=dat$data$y)) + geom_line() +
geom_line(aes(x=1:100, y=dat$data$ucl), color="red", linetype="twodash") +
labs(x="Time", y=bquote(y["obs"])) + theme_bw()

table(dat$data$cc)

dat$param #[1] 1.00 -1.00 0.50 0.48 -0.20 ## Example 2: Generating a sample with Student-t innovations set.seed(8278) dat1 = rARCens(n=100, beta=c(1,-1), phi=c(.48,-.2), sig2=.5, x=cbind(1,rnorm(100)), cens='right', pcens=.10, innov='t', nu=3) # Plotting the time serie ggplot(data.frame(dat1$data$y), aes(x=1:100, y=dat1$data$y)) + geom_line() + geom_line(aes(x=1:100, y=dat1$data$lcl), color="red", linetype="twodash") + labs(x="Time", y=bquote(y["obs"])) + theme_bw() dat1$param
#[1]  1.00 -1.00  0.50  0.48 -0.20  3.00

[Package ARCensReg version 3.0.1 Index]