DutchAdvert | R Documentation |
TV and Radio Advertising Expenditures Data
Description
Time series of television and radio advertising expenditures (in real terms) in The Netherlands.
Usage
data("DutchAdvert")
Format
A four-weekly multiple time series from 1978(1) to 1994(13) with 2 variables.
- tv
Television advertising expenditures.
- radio
Radio advertising expenditures.
Source
Originally available as an online supplement to Franses (1998). Now available via online complements to Franses, van Dijk and Opschoor (2014).
References
Franses, P.H. (1998). Time Series Models for Business and Economic Forecasting. Cambridge, UK: Cambridge University Press.
Franses, P.H., van Dijk, D. and Opschoor, A. (2014). Time Series Models for Business and Economic Forecasting, 2nd ed. Cambridge, UK: Cambridge University Press.
See Also
Franses1998
Examples
data("DutchAdvert")
plot(DutchAdvert)
## EACF tables (Franses 1998, Sec. 5.1, p. 99)
ctrafo <- function(x) residuals(lm(x ~ factor(cycle(x))))
ddiff <- function(x) diff(diff(x, frequency(x)), 1)
eacf <- function(y, lag = 12) {
stopifnot(all(lag > 0))
if(length(lag) < 2) lag <- 1:lag
rval <- sapply(
list(y = y, dy = diff(y), cdy = ctrafo(diff(y)),
Dy = diff(y, frequency(y)), dDy = ddiff(y)),
function(x) acf(x, plot = FALSE, lag.max = max(lag))$acf[lag + 1])
rownames(rval) <- lag
return(rval)
}
## Franses (1998, p. 103), Table 5.4
round(eacf(log(DutchAdvert[,"tv"]), lag = c(1:19, 26, 39)), digits = 3)