Make two random variables equal

Questions v5
7

Yeah I think so. Also could be more complex.

    ### mu ARR
    mu_ARR = pm.Deterministic('mu_ARR', beta_IAR[year_index_obs] + gamma_IAR[cell_index_obs] + \
                                          phi_IAR[cell_index_obs] * sigma_phi_IAR)
    sigma_ARR = pm.HalfNormal('sigma_ARR', sigma=100)
    
    ### ARR
    ARR = pm.Normal('ARR', mu=mu_ARR, sigma=sigma_ARR) ## Actually ARRij

The ARR is a combination of year effect and cell effect (spatial effect). I think I can remove the beta_IAR to get a mean_ARR across year, and then expand it into the full size by adding the year effect.

1 Like