freqdom (version 1.0.4)

lagreg.est: Estimate the optimal dimension in linear regression problem

Description

Estimate regression operator \(P_k\) (matrices \(d \times d\)) in model $$Y_t = \sum_{l \in L} P_k X_{t-l} + \varepsilon_t,$$ where \(X_t\) is a \(d\)-dimensional stationary process and \(\varepsilon_t\) forms a white noise.

Usage

lagreg.est(X, Y, lags = -5:5, K = NULL, Kconst = 1)

Arguments

X

regressors process

Y

response process

lags

lags which should be estimated

K

how many directions should be inverted (as in pseudoinverse)

Kconst

constant for heuristic (as in reg.dim.est)

Value

Estimated regression operator

References

Siegfried Hormann and Lukasz Kidzinski A note on estimation in Hilbertian linear models Research report, 2012

See Also

reg.dim.est, speclagreg

Examples

# NOT RUN {
X = rar(100)
e = rar(100)
Y = X + 0.3 * e
Psi = lagreg.est(X,Y,lags=0:2)
# }