weakARMA (version 1.0.3)

estimation: Parameters estimation of a time series.

Description

Estimates the parameters of a time series for given orders p and q

Usage

estimation(p = NULL, q = NULL, y, meanparam = FALSE)

Arguments

p

Order of AR, if NULL, MA is computed.

q

Order of MA, if NULL, AR is computed.

y

Univariate time series.

meanparam

Logical argument if the mean parameter has to be computed or not. If FALSE \(\mu\) is not computed.

Value

List of estimate coefficients:

mu

Mean parameter

.

ar

Vector of AR coefficients with length is equal to p.

ma

Vector of MA coefficients with length is equal to q.

sigma.carre

Mean square residuals.

Details

This function uses the algorithm BFGS in the function optim to minimize our objective function meansq.

References

Francq, C. and Zako<U+00EF>an, J. 1998, Estimating linear representations of nonlinear processes Journal of Statistical Planning and Inference, vol. 68, no. 1, pp. 145-165.

Examples

# NOT RUN {
y<-sim.ARMA(1000,ar = c(0.9,-0.3), ma = 0.2, method = "product")
estimation(p = 2, q = 1, y = y)

estimation(p = 1, q = 1, y = CAC40return.sq, meanparam = TRUE)

# }