Bayesian regression predictions

Tutorials
9

Hi, I have a question related to this. I have a simple model:

def model(developer, time_since_joined):
    b_sigma = abs(pyro.sample('b_sigma', dist.Normal(0, 300)))
    c_sigma = abs(pyro.sample('c_sigma', dist.Normal(0, 6)))
    b0 = pyro.sample("b0", dist.Normal(0, 200))
    b1 = pyro.sample("b1", dist.Normal(0, 200))
    c0 = pyro.sample("c0", dist.Normal(0, 10))
    c1 = pyro.sample("c1", dist.Normal(0, 10))

    with pyro.plate('developer', developer):
        slack = pyro.sample("slack_comments", dist.Normal(b0 + b1 * time_since_joined, b_sigma), obs=slack_comments)
        github = pyro.sample("github_commits", dist.Normal(c0 + c1 * time_since_joined, c_sigma), obs=github_commits)
        return slack, github

And the following data:

comments commits time
Alice 7500 25 4.5
Bob 10100 32 6.0
Cole 18600 49 7.0
Danielle 25200 66 12.0
Erika 27500 96 18.0 
slack_comments = torch.tensor(data.comments.values)
github_commits = torch.tensor(data.commits.values)
time = torch.tensor(data.time.values)

dims={
    "slack_comments": ["developer"],
    "github_commits": ["developer"],
    "time": ["developer"],
}

data_dict = {
    "developer": N,
    "time_since_joined": time
}

The developer variable is used to set the dimensions of the data and the lengths of developer and time_since_joined are equal. I have the following questions:

  1. When I compute posterior_predictive, I only get the observed data resampled for each iteration. Why is this happening?
nuts_kernel = NUTS(model, jit_compile=True, ignore_jit_warnings=True)
mcmc = MCMC(nuts_kernel, num_samples=400, warmup_steps=400,
            num_chains=4, disable_progbar=True)
mcmc.run(**data_dict)
posterior_samples = mcmc.get_samples()
posterior_predictive = Predictive(model, posterior_samples).forward(**data_dict)
prior = Predictive(model, num_samples=150).forward(**data_dict)

posterior_predictive

{'slack_comments': tensor([[ 7500, 10100, 18600, 25200, 27500],
         [ 7500, 10100, 18600, 25200, 27500],
         [ 7500, 10100, 18600, 25200, 27500],
         ...,
         [ 7500, 10100, 18600, 25200, 27500],
         [ 7500, 10100, 18600, 25200, 27500],
         [ 7500, 10100, 18600, 25200, 27500]]),
 'github_commits': tensor([[25, 32, 49, 66, 96],
         [25, 32, 49, 66, 96],
         [25, 32, 49, 66, 96],
         ...,
         [25, 32, 49, 66, 96],
         [25, 32, 49, 66, 96],
         [25, 32, 49, 66, 96]])}
  1. When I need predictions on new data (with different length than the data passed initially) I still get predictions that are of same length as the initial data and these are also observed values being resampled.
predictions_dict = {
    "developer": 2,
    "time_since_joined": candidate_devs_time
}
predictions = Predictive(model, posterior_samples).forward(**predictions_dict)
predictions

{'slack_comments': tensor([[ 7500, 10100, 18600, 25200, 27500],
         [ 7500, 10100, 18600, 25200, 27500],
         [ 7500, 10100, 18600, 25200, 27500],
         ...,
         [ 7500, 10100, 18600, 25200, 27500],
         [ 7500, 10100, 18600, 25200, 27500],
         [ 7500, 10100, 18600, 25200, 27500]]),
 'github_commits': tensor([[25, 32, 49, 66, 96],
         [25, 32, 49, 66, 96],
         [25, 32, 49, 66, 96],
         ...,
         [25, 32, 49, 66, 96],
         [25, 32, 49, 66, 96],
         [25, 32, 49, 66, 96]])}

Kindly suggest me steps to solve this.