machine learning - Granger causality for muti-dimensional time series for single variable -

i trying use granger causality testing causal dependency between 2 variables. granger causal models (and vector autoregression models) use single time series (one sample per variable) each of variables (or if there multiple variables, 1 time series each variable). python statsmodels granger causality implementation gives example of this.

however in case have 2 variables, multiple samples of time series each variable , want fuse them granger causal model together. instead of considering each time series pair separately variables output of granger causal model, possible use of them form kind of model analogous regression model.