R: Describing Intervention Effects for Time Series with...
interv_covariate
R Documentation
Describing Intervention Effects for Time Series with Deterministic Covariates
Description
Generates covariates describing certain types of intervention effects according to the definition by Fokianos and Fried (2010).
Usage
interv_covariate(n, tau, delta)
Arguments
n
integer value giving the number of observations the covariates should have.
tau
integer vector giving the times where intervention effects occur.
delta
numeric vector with constants specifying the type of intervention (see Details). Must be of the same length as tau.
Details
The intervention effect occuring at time τ is described by the covariate
X_t = δ^(t-τ) I(t>=τ),
where I(t>=τ) is the indicator function which is 0 for t < τ and 1 for t >= τ. The constant δ with 0 <= δ <= 1 specifies the type of intervention. For δ = 0 the intervention has an effect only at the time of its occurence, for 0 < δ < 1 the effect decays exponentially and for δ = 1 there is a persistent effect of the intervention after its occurence.
If tau and delta are vectors, one covariate is generated with tau[1] as τ and delta[1] as δ, another covariate for the second elements and so on.
Value
A matrix with n rows and length(tau) columns. The generated covariates describing the interventions are the columns of the matrix.
Liboschik, T., Kerschke, P., Fokianos, K. and Fried, R. (2014) Modelling interventions in INGARCH processes. International Journal of Computer Mathematics (published online), http://dx.doi.org/10.1080/00207160.2014.949250.
See Also
tsglm for fitting a GLM for time series of counts.
interv_test, interv_detect and interv_multiple for tests and detection procedures for intervention effects.