cross_corr_ar¶
-
hyppo.tools.
cross_corr_ar
(n, lag=1, phi=0.5, sigma=1)¶ 2 linearly dependent time series simulation.
Xt and Yt are together a bivariate univarite AR(
1ag
) with ϕ=[00.50.50] for both series. Noise follows N(0,σ). With lag (1), this is[XtYt]=[0ϕϕ0][Xt−1Yt−1]+[ϵtηt]- Parameters
- Returns
x,y (
ndarray
) -- Simulated data matrices.x
andy
have shape(n,)
where n is the number of samples.