Creating hierarchical models using 3 dimensional xarray data

Sorry, haven’t had the time to work on this further. What I did discover was that I need to impute missing values inside the model.

Something like

x_mu = pm.Normal('x_mu', mu = 0, sd = 1, dims = dims)
x_imputed = pm.Normal('x_imputed', mu = x_mu, sd = 1, observed = x, dims = dims)

Once I have more time I’ll post my findings here.
It’s a bit hectic here with a new child on the way and a new job :relaxed:

1 Like