Defining Vectorizable Helper Function given Posterior Draws (Forecasting with BVAR)

Huh, that actually allows me to start drawing samples. Interesting! TY!

however, i am getting massive performance issue drops. with four equations and 2 lags (4 chains, 500 draws) , my approximate time to complete is 30 mins…it’s probably 5 when i define my model in this manner: Bayesian Vector Autoregressive Models — PyMC example gallery