time series - Simulate cross-lag panel model data in R

Question

I'm looking for a way to concisely and flexibly simulate cross-lag panel data in R.

Considering a 2-variable use case for X and Y over some arbitrary number of time units n, I'd like to be able to specify:

1. Autoregressive relationships for each variable independently (so, AR(1) for each of X & Y)
2. Overall mean for each variable
3. Variance for each variable
4. Cross-lagged relationships between X and Y (so, the b in X_n = b * Y_(n-1), and vice versa)
5. [less important, but useful] Change over time (slope from 1 to n for each variable)

The single-variable version of this seems possible with something like arima.sim(), but I'm not sure how to best to approach this in the multivariate case.

Custom approaches are fine, but any packages implementing a solution would be even better.

Answer 1

等待大神答复