Attempting multiplication of list and float. about mmm_stan HOT 4 CLOSED

The simple solution is to just remove the .tolist() method from data['X'] and data['y]

def train_hill_model(df, mc_df, adstock_params, media, sm):
    data, sc = create_hill_model_data(df, mc_df, adstock_params, media)
    fit = sm.sampling(data=data, iter=2000, chains=4)
    fit_result = fit.extract()
    hill_model = {
        'beta_hill_list': fit_result['beta_hill'].tolist(),
        'ec_list': fit_result['ec'].tolist(),
        'slope_list': fit_result['slope'].tolist(),
        'sc': sc,
        'data': {
            'X': data['X'],
            'y': data['y'],
        }
    }
    return hill_model

from mmm_stan.

Thanks for pointing out this! Your solution is correct.
tolist() is used to save the model as json file, because arrays cannot be saved as json. I will revise those model saving and loading functions later this weekend.

from mmm_stan.

No problem! Thanks for putting this together; I've found it quite helpful.

from mmm_stan.

For the sake of model storage, I still prefer to store data['X'] and data['y'] as lists, so that the hill model can be saved as json using save_json(). I modified evaluate_hill_model(), converting them to arrays.

from mmm_stan.