CloudCeiling {ARCensReg}

Cloud ceiling height

Description

The cloud ceiling heights, collected by the National Center for Atmospheric Research (NCAR), were observed hourly in San Francisco during March 1989, consisting of n=716 observations (Park et al. 2007).

Usage

data(CloudCeiling)

Format

This data frame contains the following columns:

y

Logarithm of the cloud ceiling heights.

cc

Right censoring indicator (1 if the observation is right-censored and 0 otherwise).

Source

Park JW, Genton MG, Ghosh SK (2007). “Censored time series analysis with autoregressive moving average models.” Canadian Journal of Statistics, 35(1), 151–168.

See Also

ARCensReg, ARtCensReg

Examples

library(ggplot2)

data(CloudCeiling)
ggplot(CloudCeiling) + geom_line(aes(x=1:length(y), y=y)) + 
  labs(x="Time") + theme_bw()

# Proportion of censoring
prop.table(table(CloudCeiling$cc))

## Not run: 
# A censored regression model
## This may take a long time due to the number of censored observations.
## For other examples see help(ARCensReg).

x   = as.matrix(rep(1, length(CloudCeiling$y)))
cc  = CloudCeiling$cc
lcl = CloudCeiling$y
ucl = rep(Inf, length(CloudCeiling$y))
miss =  which(is.na(CloudCeiling$y))
cc[miss]  = 1
lcl[miss] = -Inf
AR_reg = ARCensReg(cc, lcl, ucl, CloudCeiling$y, x, p=1, tol=.001)
## End(Not run)