Creating custom Joint Distribution in pymc3

Questions
3

Hi,

Here’s the code. I’m currently working with simulated data from stats.binomial.rvs()

alpha_a = 1
beta_a = 1
alpha_b = 1
beta_b = 1
n = 10000

variants = ["Control", "Optimisation"]

with pm.Model() as ab_model:
    
    theta_a = pm.Beta(variants[0], alpha = alpha_a, beta = beta_a)
    theta_b = pm.Beta(variants[1], alpha = alpha_b, beta = beta_b)
    
    data_a = pm.Binomial("A Obs", n = n, p = theta_a, observed = data["Control"])
    data_b = pm.Binomial("B Obs", n = n, p = theta_b, observed = data["Optimisation"])
    
    step = pm.NUTS()
    trace = pm.sample(10000, step = step, return_inferencedata=True)
1 Like