Thank you. I get Frozen({'chain': 4, 'draw': 2000, 'obs_dim_0': 48}). But the new data points I want to draw samples for are only length 12, which are the lengths of the data I’m putting in for pm.set_data argument shown below.
What is the correct way to be able to get the posterior predictive mean and hdi for the 12 new data points?
pm.set_data(new_data = {'promo':promo_test, 't':t_test, 's':s_test, 'a':a_test}, model = model) test_ppc = pm.sample_posterior_predictive(trace, model=model)