Shape Parameter for Multiple Treatment Groups

Questions
4

Thanks for the response! Almost there but not quite, the final line with the likelihood is now providing the error. I’m generating the fake data like so:

count_data = pd.DataFrame({
 "n": np.random.poisson(lam = 15, size = 74*3),
 "person_id": np.repeat([0,1,2], 74)
})

and get a ValueError: Input dimension mis-match. (input[0].shape[1] = 222, input[1].shape[1] = 3) As a reminder I’m using 3 people, each with 74 items of data.

and as above, the likelihood function is

 observation = pm.Poisson("obs", mu=lambda_[count_data['person_id']], observed=count_data['n'])