tfarima (version 0.4.1)

CalendarVar: Calendar variables

Description

`CalendarVar()` creates a set of deterministic regressors to capture calendar effects (trading/working days, length-of-month, leap-year and Easter).

Usage

CalendarVar(
  x,
  form = c("dif", "td", "td7", "td6", "wd", "null"),
  ref = 0,
  lom = TRUE,
  lpyear = TRUE,
  easter = FALSE,
  len = 4,
  easter.mon = FALSE,
  n.ahead = 0
)

Value

An object of class `ts` or `mts` with the requested regressors.

Arguments

x

A `ts` object used to determine start, length and frequency.

form

Character selecting the set of calendar variables: `"dif"` (differences wrt reference day), `"td"` (6 dummies + lom; omits reference), `"td7"` (7 dummies), `"td6"` (6 dummies; omits reference), `"wd"` (weekdays vs weekend), or `"null"` (no trading-day regressors).

ref

Non-negative integer (0–6) indicating the reference day (0 = Sunday, 1 = Monday, …, 6 = Saturday). Ignored unless `form` needs it.

lom

Logical. If `TRUE` include a length-of-month regressor.

lpyear

Logical. If `TRUE` include a leap-year regressor.

easter

Logical. If `TRUE` include an Easter regressor.

len

Integer duration for Easter effect (days). Typical values: 4–8.

easter.mon

Logical. `TRUE` if Holy Monday is a public holiday.

n.ahead

Integer. Extra observations to extend the sample (forecast horizon).

References

Bell, W.R. and Hillmer, S.C. (1983) “Modeling time series with calendar variation”, *Journal of the American Statistical Association*, 78, 526–534.

Examples

X <- CalendarVar(AirPassengers, form = "wd", easter = TRUE, len = 5)