seasonal.dummy: Seasonal dummy variables

In ttnsdcn/forecast-package: Forecasting functions for time series

Description

seasonaldummy and seasonaldummyf return matrices of dummy variables suitable for use in arima, lm or tslm. The last season is omitted and used as the control.

fourier and fourierf return matrices containing terms from a Fourier series, up to order K, suitable for use in arima, lm or tslm.

Usage

seasonaldummy(x)
seasonaldummyf(x,h)
fourier(x,K)
fourierf(x,K,h)

Arguments

x	Seasonal time series
h	Number of periods ahead to forecast
K	Maximum order of Fourier terms

Value

Numerical matrix with number of rows equal to the length(x) and number of columns equal to frequency(x)-1 (for seasonaldummy and seasonaldummyf or 2*K (for fourier or fourierf).

Author(s)

Rob J Hyndman

Examples

plot(ldeaths)

# Using seasonal dummy variables
month <- seasonaldummy(ldeaths)
deaths.lm  <- tslm(ldeaths ~ month)
tsdisplay(residuals(deaths.lm))
ldeaths.fcast <- forecast(deaths.lm, 
  data.frame(month=I(seasonaldummyf(ldeaths,36))))
plot(ldeaths.fcast)

# A simpler approach to seasonal dummy variables
deaths.lm  <- tslm(ldeaths ~ season)
ldeaths.fcast <- forecast(deaths.lm, h=36)
plot(ldeaths.fcast)

# Using Fourier series
X <- fourier(ldeaths,3)
deaths.lm  <- tslm(ldeaths ~ X)
ldeaths.fcast <- forecast(deaths.lm, 
  data.frame(X=I(fourierf(ldeaths,3,36))))
plot(ldeaths.fcast)

ttnsdcn/forecast-package documentation built on June 1, 2019, 2:49 a.m.