Impute results in mismatch dimensions in dims and data

Weird, seems like something unexpected is happening when you directly pass xarrays to the observed arguments in velocity and load. @OriolAbril might know more?

I was able to clear the error by wrapping the data in pm.Data before passing them as observed:

    velocity_data = pm.Data('velocity_data', dataset['velocity_std_masked'], mutable=True)
    load_data = pm.Data('load_data', dataset['load_std_masked'], mutable=True)

then observed=velocity_data and observed=load_data on the two respective likelihoods.