I have the following code: #std_range = df.std_range.values[:, np.newaxis] #shared(std_range, broadcastable=(False, True)) #x1_shift = ppc['obs'][:K] x1_shift = shared(df['x1'].values[:, np.newaxis], broadcastable=(False, True)) x2obs = df['x2'].values[:, np.newaxis] #Model for x2 with pm.Model() a…

Pm.sample_ppc breaks with AttributeError

mullachv February 5, 2018, 12:56pm 17

Yes, unbounded works. If you look at my notebook, that is how I get ‘x1’.
With 50 samples for x2:

Topic		Replies	Views
Shape mismatch error in 'sample_ppc' Questions	8	905	March 6, 2019
Using a bounded variable within a Mixture Questions	8	1862	January 24, 2018
Trouble understanding sample_ppc Questions	4	651	June 7, 2018
Prediction using sample_ppc in Hierarchical model Questions from_github	6	5004	December 14, 2017
KeyError Setting Mixture Proportions for Mixture Model Questions	26	1503	May 13, 2019