Creates a multivariate time series which is suitable for using
as a regression variable for monthly or quarterly data.
The regression variable models the effects of the
number of trading days in a month.
a univariate or multivariate time series.
The actual data in the series are not used.
start
either a number or an integer vector of length
2 giving the
starting date for the series in years (see the function
rts).
end
either a number or an integer vector of length
2 giving the
ending date for the series in years (see the function
rts).
frequency
observations frequency.
Only monthly data (
frequency equals
12) and quarterly data (
frequency
equals
4) are permitted.
VALUE:
a multivariate time series is returned with seven columns.
The first column gives the number of days in the month (quarter),
and the following six columns give the difference between
the number of Saturdays, Sundays,
Mondays, Tuesdays, Wednesdays and Thursdays,
and the number of Fridays in the month (quarter).
Thus the last six columns consist of numbers from
{-1, 0, 1} for monthly
data.
REFERENCES:
Hillmer, S. C., Bell, W. R., and Tiao G. C. (1983).
Modeling considerations in the seasonal adjustment of economic time series.
In
Applied Time Series Analysis of Economic Data.
A. Zellner, ed. U.S. Bureau of the Census, Washington, DC. pp. 74-100.
SEE ALSO:
,
.
EXAMPLES:
# Fit an ARIMA (0,1,1) model to the ship data using trading day
# variables as covariates.
td.ship <- arima.td(ship)
mle.td <- arima.mle(ship,model=list(order=c(0,1,1)),xreg=td.ship)