so if I look at model.test_point these are all zero
Issue is for this value my model will be dividing by zero in some cases. But for my prior, there is zero probability of these parameters being zero.
Even if I change the lower bound on some parameters, test_point still gives zero values!
Am I being stupid . . 