Dimension error for Hierarchical model

Questions version agnostic
6

@cluhmann Thank you for clarifying it. so how can I fix it to give me the shape 120 but also the dimensions of 5 per region?
do I use

control_beta = pm.Normal(f"{control_var}_control_coef",
                                    mu=cont_beta_mu,
                                    sigma=cont_beta_sigma,
                                    dims= 'Region', 
                                    shape= (5,120)
)

instead of

control_beta = pm.Normal(f"{control_var}_control_coef",
                                    mu=cont_beta_mu,
                                    sigma=cont_beta_sigma,
                                    dims= 'Region')

And I am wondering why that error didn’t occur in the channel_b:

channel_b = pm.Exponential(f"{channel}_coef",
                                    lam=channel_beta_lam,
                                    dims= 'Region')

which is also multiplied by the saturation tensor

channel_contribution= pm.Deterministic(
                f'contribution_{channel}',
                channel_b * saturation_tensor)

appreciate your support to help me understand this.

thanks :slight_smile: