Deprecate and disable budget optimization method & Updating notebook #1832

cetagostini · 2025-07-16T09:09:27Z

Description

Marked the budget optimization method as deprecated and replaced its implementation with a NotImplementedError. Users are advised to migrate to the Multidimensal.MMM class.

Related Issue

Closes #
Related to #

Checklist

Checked that the pre-commit linting/style checks pass. Feel free to comment pre-commit.ci autofix to auto-fix.
Included tests that prove the fix is effective or that the new feature works
Added necessary documentation (docstrings and/or example notebooks) using numpydoc format.
If you are a pro: each commit corresponds to a relevant logical change

📚 Documentation preview 📚: https://pymc-marketing--1832.org.readthedocs.build/en/1832/

Marked the budget optimization method as deprecated and replaced its implementation with a NotImplementedError. Users are advised to migrate to the `Multidimensal.MMM` class.

codecov · 2025-07-16T09:12:15Z

Codecov Report

✅ All modified and coverable lines are covered by tests.
⚠️ Please upload report for BASE (main@492930a). Learn more about missing BASE report.

Additional details and impacted files

@@           Coverage Diff           @@
##             main    #1832   +/-   ##
=======================================
  Coverage        ?   92.28%           
=======================================
  Files           ?       64           
  Lines           ?     7430           
  Branches        ?        0           
=======================================
  Hits            ?     6857           
  Misses          ?      573           
  Partials        ?        0

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

williambdean · 2025-07-16T10:21:04Z

pymc_marketing/mmm/mmm.py

-            budget_bounds=budget_bounds,
-            callback=callback,
-            **minimize_kwargs,
+        raise NotImplementedError(


can we warn instead of raise?

I think better raise, only because the current method even if it works will return incorrect results. No sense to keep it, but what do you think?

williambdean · 2025-07-16T10:22:31Z

pymc_marketing/mmm/mmm.py

+        """Optimize the given budget based on the specified utility function over a specified time period (DEPRECATED).

-        This function optimizes the allocation of a given budget across different channels
+        DEPRECATED: This function optimizes the allocation of a given budget across different channels


There is a specific way to call out in numpydocs

https://numpydoc.readthedocs.io/en/latest/format.html#deprecation-warning

I'll do this!

cetagostini · 2025-07-25T00:03:10Z

This complements #1849 deprecating method, meanwhile 1849 remove and update notebooks and dependencies.

review-notebook-app · 2025-07-26T08:54:25Z

Check out this pull request on

See visual diffs & provide feedback on Jupyter Notebooks.

Powered by ReviewNB

…ithub.com/pymc-labs/pymc-marketing into cetagostini/deprecating_optimizer_for_mmm

juanitorduz · 2025-07-27T05:00:58Z

@cetagostini i think its better to warn on the deprecation and remove in in 0.17.0

People are running these models in production and this is a key method. We will remove it for sure, but I think we need to do it in two steps.

Thoughts?

juanitorduz · 2025-07-27T05:07:39Z

In the meantime… while warning. Can we somehow use the optimizer from the mmm multi clas I the old mmm class?

PabloRoque · 2025-07-28T10:40:58Z

@cetagostini i think its better to warn on the deprecation and remove in in 0.17.0

People are running these models in production and this is a key method. We will remove it for sure, but I think we need to do it in two steps.

Thoughts?

Users can always pin to 0.15.1. Maybe raise the error and tell them to use 0.15.1?

cetagostini · 2025-07-29T17:34:05Z

In the meantime… while warning. Can we somehow use the optimizer from the mmm multi clas I the old mmm class?

So, I could work on a PR to modify MMM base model to work as multidimensional, and be okay with the optimizer again. Nevertheless, probably I agree with Pablo that users who wants to continue with MMM base can always pin 15.1. But I follow your lead here @juanitorduz

You dictate the course and I'll follow capitan!

juanitorduz · 2025-07-29T17:35:24Z

So, I could work on a PR to modify MMM base model to work as multidimensional, and be okay with the optimizer again.

How much work would this be for you ? :)

cetagostini · 2025-07-29T17:37:44Z

So, I could work on a PR to modify MMM base model to work as multidimensional, and be okay with the optimizer again.

How much work would this be for you ? :)

Maybe another day or two? (Meaning maybe like 5 days in OSS days 😅) Should I open the PR only adjusting scaling method on base MMM?

juanitorduz · 2025-07-29T17:42:11Z

I mean, I think we can wait a week (or two), there is really no rush. It would be great to modify MMM base model to work as multidimensional, if possible, in view that I know many people looking into the end-to-end notebook: https://www.pymc-marketing.io/en/latest/notebooks/mmm/mmm_case_study.html. Also, the MMM multidim still does not fully support the HSGP (see #1432). Hence, if we remove the optimizer, users won't be able to have a single model to fit and optimize.

Hence, I suggest we modify MMM base model to work as multidimensional. No rush. Let me know if I can help.

Does it make sense?

cetagostini · 2025-07-29T17:44:52Z

Okay, I'll move this to draft, and will try to crack a PR for that during the next few days!

juanitorduz · 2025-07-29T17:46:53Z

Okay, I'll move this to draft, and will try to crack a PR for that during the next few days!

Thank you @cetagostini ❤️ . I will buy you a beer in September (PyData Berlin) as a reward for this fantastic work with the optimizer! #hero

cetagostini · 2025-08-07T15:53:34Z

I need to get #1862 merge before continue here.

juanitorduz · 2025-08-08T07:16:13Z

I need to get #1862 merge before continue here.

I just merged #1862 ! Thanks 🚀

…p branch version of docs/source/notebooks/mmm/mmm_allocation_assessment.ipynb

… from main

cetagostini · 2025-08-10T19:36:38Z

This looks ready, but the notebook get issue. I feel it's related to the nc loaded from my env, can you give me a hand here? and push the one you have @juanitorduz ? That should be all.

Thank you!

juanitorduz · 2025-08-10T19:42:47Z

@carlosagostini can you please comment out the line mmm.save("multidimensional_model.nc") in the https://www.pymc-marketing.io/en/latest/notebooks/mmm/mmm_multidimensional_example.html#save-model notebook? I think this might be causing the problem as the model is saved every time the notebook tests are run (we had a similar issue in the past with the normal mmm notebook)

juanitorduz · 2025-08-10T20:18:41Z

Thank you @carlosagostini !

rdevvvv · 2025-11-21T16:32:23Z

Hello! Thanks for the good work. I am currently following: https://www.pymc-marketing.io/en/0.11.0/notebooks/mmm/mmm_case_study.html#media-deep-dive and I was wondering what the next course of action was to optimize budget. My MMM is not multidimensionnal so I don't understand. Do I rewrite my code to use the multidimensional class but with just one dimension? It seems inefficient and I don't know if I can even do that. Thanks,

Deprecate and disable budget optimization method

11535c8

Marked the budget optimization method as deprecated and replaced its implementation with a NotImplementedError. Users are advised to migrate to the `Multidimensal.MMM` class.

github-actions bot added the MMM label Jul 16, 2025

williambdean reviewed Jul 16, 2025

View reviewed changes

cetagostini added 2 commits July 25, 2025 02:58

Merge branch 'main' into cetagostini/deprecating_optimizer_for_mmm

60e7f52

Using deprecation numpy style

60aadbb

cetagostini added 2 commits July 26, 2025 03:22

Merge branch 'main' into cetagostini/deprecating_optimizer_for_mmm

a602b23

Adjusting notebook

0d7ac5d

github-actions bot added the docs Improvements or additions to documentation label Jul 26, 2025

cetagostini added 3 commits July 26, 2025 11:54

Merge branch 'main' into cetagostini/deprecating_optimizer_for_mmm

e6da8aa

Solving test issues

9699623

Merge branch 'cetagostini/deprecating_optimizer_for_mmm' of https://g…

5017bb5

…ithub.com/pymc-labs/pymc-marketing into cetagostini/deprecating_optimizer_for_mmm

github-actions bot added the tests label Jul 26, 2025

cetagostini added bug Something isn't working maintenance labels Jul 26, 2025

cetagostini changed the title ~~Deprecate and disable budget optimization method~~ Deprecate and disable budget optimization method & Updating notebook Jul 26, 2025

cetagostini added 6 commits July 26, 2025 12:25

Update multidimensional_model.nc size optimization

7d3f3c9

Merge main branch and resolve conflicts

347c141

Updating notebooks

4ef6169

Merge branch 'main' into cetagostini/deprecating_optimizer_for_mmm

9f7eb7d

Notebook update

d6f6c42

Update model nc

7d911ea

cetagostini self-assigned this Jul 26, 2025

cetagostini requested review from PabloRoque, juanitorduz and williambdean July 26, 2025 22:46

Merge branch 'main' into cetagostini/deprecating_optimizer_for_mmm

d14f525

cetagostini marked this pull request as draft July 29, 2025 17:46

juanitorduz added this to the 0.16.0 milestone Jul 30, 2025

Merge branch 'main' into cetagostini/deprecating_optimizer_for_mmm

eda9f5c

Changing to raise warning

a97670f

cetagostini added 2 commits August 10, 2025 22:02

Merge origin/main into cetagostini/deprecating_optimizer_for_mmm; kee…

8970c81

…p branch version of docs/source/notebooks/mmm/mmm_allocation_assessment.ipynb

Revert docs/source/notebooks/mmm/multidimensional_model.nc to version…

948ccde

… from main

cetagostini marked this pull request as ready for review August 10, 2025 19:35

Following Juan idea

64e60c0

juanitorduz approved these changes Aug 10, 2025

View reviewed changes

juanitorduz mentioned this pull request Aug 10, 2025

Improvements on multidimensional notebook #1876

Merged

cetagostini merged commit 9602e91 into main Aug 10, 2025
31 of 32 checks passed

cetagostini deleted the cetagostini/deprecating_optimizer_for_mmm branch August 10, 2025 20:21

Deprecate and disable budget optimization method & Updating notebook #1832

Deprecate and disable budget optimization method & Updating notebook #1832

Conversation

cetagostini commented Jul 16, 2025 • edited by github-actions bot Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Description

Related Issue

Checklist

Uh oh!

codecov bot commented Jul 16, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Codecov Report

Uh oh!

williambdean Jul 16, 2025

Choose a reason for hiding this comment

Uh oh!

cetagostini Jul 21, 2025

Choose a reason for hiding this comment

Uh oh!

williambdean Jul 16, 2025

Choose a reason for hiding this comment

Uh oh!

cetagostini Jul 21, 2025

Choose a reason for hiding this comment

Uh oh!

cetagostini commented Jul 25, 2025

Uh oh!

review-notebook-app bot commented Jul 26, 2025

Uh oh!

juanitorduz commented Jul 27, 2025

Uh oh!

juanitorduz commented Jul 27, 2025

Uh oh!

PabloRoque commented Jul 28, 2025

Uh oh!

cetagostini commented Jul 29, 2025

Uh oh!

juanitorduz commented Jul 29, 2025

Uh oh!

cetagostini commented Jul 29, 2025

Uh oh!

juanitorduz commented Jul 29, 2025

Uh oh!

cetagostini commented Jul 29, 2025

Uh oh!

juanitorduz commented Jul 29, 2025

Uh oh!

cetagostini commented Aug 7, 2025

Uh oh!

juanitorduz commented Aug 8, 2025

Uh oh!

cetagostini commented Aug 10, 2025

Uh oh!

juanitorduz commented Aug 10, 2025

Uh oh!

juanitorduz commented Aug 10, 2025

Uh oh!

Uh oh!

rdevvvv commented Nov 21, 2025

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

6 participants

cetagostini commented Jul 16, 2025 •

edited by github-actions bot

Loading

codecov bot commented Jul 16, 2025 •

edited

Loading