Bayesian workflow example

[fonnesbeck]

A notebook adapted from course material on how to implement Gelman et al.'s Bayesian workflow, using a COVID-19 dataset.

• Notebook follows style guide https://docs.pymc.io/en/latest/contributing/jupyter_style.html
• PR description contains a link to the relevant issue:
  • a tracker one for existing notebooks (tracker issues have the "tracker id" label)
  • or a proposal one for new notebooks
• Check the notebook is not excluded from any pre-commit check: https://github.com/pymc-devs/pymc-examples/blob/main/.pre-commit-config.yaml

Helpful links

• https://github.com/pymc-devs/pymc-examples/blob/main/CONTRIBUTING.md

---

📚 Documentation preview 📚: https://pymc-examples--797.org.readthedocs.build/en/797/

[review-notebook-app]

Check out this pull request on 

See visual diffs & provide feedback on Jupyter Notebooks.

---

Powered by ReviewNB

[fonnesbeck]

Not sure if this belongs in case studies or elsewhere ...

[review-notebook-app]

View / edit / reply to this conversation on ReviewNB

twiecki commented on 2025-06-19T14:52:53Z
----------------------------------------------------------------

A lot of these have not plot?

fonnesbeck commented on 2025-06-19T19:58:54Z
----------------------------------------------------------------

I don't understand the question.

aloctavodia commented on 2025-06-19T20:06:11Z
----------------------------------------------------------------

I think he means that a lot of the plots are not visible...

fonnesbeck commented on 2025-06-19T20:11:30Z
----------------------------------------------------------------

That's odd. Will ask Oriol about this. Would be a shame if we can't use plotly in the docs.

fonnesbeck commented on 2025-06-19T20:48:38Z
----------------------------------------------------------------

OK, I've fixed this. Thanks.

[review-notebook-app]

View / edit / reply to this conversation on ReviewNB

twiecki commented on 2025-06-19T14:55:49Z
----------------------------------------------------------------

Maybe add COVID-19 or infection modeling to the title.

[twiecki]

Not sure if this belongs in case studies or elsewhere ...

Either case-study or howto, leaning towards the latter.

[review-notebook-app]

View / edit / reply to this conversation on ReviewNB

aloctavodia commented on 2025-06-19T18:40:00Z
----------------------------------------------------------------

The first two sentences could be "Bayesian modeling is a robust approach to draw conclusions from data. Successful modeling involves an interplay..."

[review-notebook-app]

View / edit / reply to this conversation on ReviewNB

aloctavodia commented on 2025-06-19T18:40:01Z
----------------------------------------------------------------

A slightly different way to write this would be

dist = pz.Gamma()

pz.maxent(dist, lower=0.1, upper=20, mass=0.95);

Then, in the next cell

fig = px.histogram(x=dist.rvs(1000), nbins=20, title="Gamma Distribution Samples")

fig.show()

and in the model

alpha = dist.to_pymc("alpha")

[review-notebook-app]

View / edit / reply to this conversation on ReviewNB

aloctavodia commented on 2025-06-19T18:40:01Z
----------------------------------------------------------------

we can use

alpha = pz.maxent(pz.Gamma(), lower=0.1, upper=20, mass=0.95, plot=False).to_pymc("alpha")

instead of

  gamma_params = pm.find_constrained_prior(

    pm.Gamma, lower=0.1, upper=20, init_guess={"alpha": 6, "beta": 1}, mass=0.95

  )

  alpha = pm.Gamma("alpha", **gamma_params)

[review-notebook-app]

View / edit / reply to this conversation on ReviewNB

aloctavodia commented on 2025-06-19T18:40:02Z
----------------------------------------------------------------

plt.tight_layout() not neccesary given that we are using "arviz-doc" style

[review-notebook-app]

View / edit / reply to this conversation on ReviewNB

aloctavodia commented on 2025-06-19T18:40:03Z
----------------------------------------------------------------

plt.show() not neccesary?

[review-notebook-app]

View / edit / reply to this conversation on ReviewNB

aloctavodia commented on 2025-06-19T18:40:03Z
----------------------------------------------------------------

Just for your information "New ArviZ", has functionality for prior/likelihood sensitivity checks, see https://arviz-devs.github.io/EABM/Chapters/Sensitivity_checks.html

No needed to be done now, but maybe in a next round of revision we could use "new arviz" its new prior checks, or new plot like rank plots...

[review-notebook-app]

View / edit / reply to this conversation on ReviewNB

aloctavodia commented on 2025-06-19T18:40:04Z
----------------------------------------------------------------

"sampels"

[fonnesbeck]

I don't understand the question.

---

View entire conversation on ReviewNB

[aloctavodia]

I think he means that a lot of the plots are not visible...

---

View entire conversation on ReviewNB

[fonnesbeck]

That's odd. Will ask Oriol about this. Would be a shame if we can't use plotly in the docs.

---

View entire conversation on ReviewNB

[fonnesbeck]

OK, I've fixed this. Thanks.

---

View entire conversation on ReviewNB

[review-notebook-app]

View / edit / reply to this conversation on ReviewNB

aloctavodia commented on 2025-06-19T21:03:15Z
----------------------------------------------------------------

This can be removed

[review-notebook-app]

View / edit / reply to this conversation on ReviewNB

aloctavodia commented on 2025-06-19T21:03:16Z
----------------------------------------------------------------

Add ;

[review-notebook-app]

View / edit / reply to this conversation on ReviewNB

aloctavodia commented on 2025-06-19T21:03:16Z
----------------------------------------------------------------

Also add ;

[review-notebook-app]

View / edit / reply to this conversation on ReviewNB

aloctavodia commented on 2025-06-19T21:03:17Z
----------------------------------------------------------------

Add ;

[review-notebook-app]

View / edit / reply to this conversation on ReviewNB

aloctavodia commented on 2025-06-19T21:03:18Z
----------------------------------------------------------------

Add ;

[review-notebook-app]

View / edit / reply to this conversation on ReviewNB

aloctavodia commented on 2025-06-19T21:03:18Z
----------------------------------------------------------------

Add ;

[fonnesbeck]

Everyone cool with these changes?