lagged.cov: Compute cross covariance with a given lag

In freqdom: Frequency Domain Analysis for Multivariate Time Series

Description

For a given multivariate stationary time series estimates a covarianve matrix C_{XY}^k = Cov(X_k,Y_0) using the formula

\hat C_{XY}^k = \frac{1}{n} ∑_{i=1}^{n-k} X_{k+i} Y_k'.

Usage

lagged.cov(X, Y = NULL, lag = 0)

Arguments

X	first process
Y	second process, if null then autocovariance of X is computed
lag	the lag that we are interested in

Value

Covariance matrix

Examples

X = rar(100)
Y = rar(100)
lagged.cov(X,Y)

freqdom documentation built on May 2, 2019, 5:55 p.m.