| Specials {sspir} | R Documentation |
Auxiliary function for modelling the time-dependent trend patterns
when formulating ssm models.
polytrig(time, period = 365.25, degree = 1) polytime(time, degree = 1) sumseason(time, period = 12) season(time, period = 12)
time |
a vector giving the observation times, eg. 1:n. The
variable 'time' is by default set to time(y), where y is
the response in the formula in the ssm call. |
period |
the period of the seasonal pattern. |
degree |
the degree of the trigonometric polynomial or the degree of the polynomial time trend. |
polytrig defines a trigonometric polynomial in time,
polytime defines a polynomial time trend, sumseason
produces a seasonal pattern in which the factors sum to zero over the
period, and season defines a factor. The functions polytrig and
season can be used outside the ssm call. However,
polytime and sumseason are treated mainly inside the
ssm call.
Claus Dethlefsen and Søren Lundbye-Christensen.
ssm, kfilter,
smoother, getFamily.
polytrig(1:10,degree=2) season(1:12,period=4)