cpcens (version 0.1.0)

rcma: Simulate censored MA time series

Description

Randomly Generate Censored MA

Usage

rcma(n = 100, ma = 0.6, mu = 0, siga = 1, rates = c(NA, 0.2),
  Mrate = 0)

Arguments

n

length of series, default 100.

ma

ma coefficients

mu

mean

siga

standard deviation of innovations

rates

either a vector of length 2 or a matrix with n rows and 2 columns. In the vector case, the first element indicates the left-censor rate and the second element indicates the right-censor rate. Set to NA if there is no censoring. Interval censored data corresponds to setting both a left-censor rate and right-censor rate. The default setting indicates a right-censor rate 0.2 with no left censoring. The vector case handles single censoring and the matrix case is for multiple censor points. In this case each column indicates the corresponding censoring for each observation.

Mrate

fraction of missing values. Default is 0

Value

an object of class 'censored' which is a list with three elements. First element, 'y', is the censored time series. Second element, 'iy', indicates for each observed valued "o", "L", "R", NA according to whether the value is fully observed, left-censored, right-censored, or missing. Third element, 'censorPts', is a matrix with 2 columns indicating the censor point or NA if no censoring is applicable. Note that censorPts does not indicate if the observation was actually censored since this depends on the unknown latent variable. An observation is censored if and only if the corresponding entry in iy is either "L" or "R".

Examples

# NOT RUN {
#Default example
library(cpcens)
ans<-rcma()
#example (right censoring)
ans = rcma (n=100 , ma=0.2, mu=0 , siga=1, rates=c(NA,0.7), Mrate=0 )
#example (left censoring)
ans = rcma (n=100 , ma=0.7, mu=0 , siga=1, rates=c(0.3,NA), Mrate=0 )
#example (interval censoring)
ans = rcma (n=100 , ma=0.7, mu=0 , siga=1, rates=c(0.1,0.3), Mrate=0 )
# }