-
Notifications
You must be signed in to change notification settings - Fork 185
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
MMM Case Study from PyData Global #1044
base: main
Are you sure you want to change the base?
Conversation
Check out this pull request on See visual diffs & provide feedback on Jupyter Notebooks. Powered by ReviewNB |
36197ff
to
148674f
Compare
6647623
to
ebb49eb
Compare
Codecov ReportAll modified and coverable lines are covered by tests ✅
Additional details and impacted files@@ Coverage Diff @@
## main #1044 +/- ##
=======================================
Coverage 95.56% 95.56%
=======================================
Files 39 39
Lines 4011 4011
=======================================
Hits 3833 3833
Misses 178 178 ☔ View full report in Codecov by Sentry. |
83cac99
to
481efc9
Compare
Co-authored-by: Will Dean <[email protected]>
View / edit / reply to this conversation on ReviewNB cetagostini commented on 2024-10-07T14:30:21Z Nice! I would add a note just to mention that, although good as a business practice or for stakeholder management, you must be careful when adding these constraints because basically, by decreasing flexibility, you risk to not finding the most optimal solution.
For example, the true global minimum in our function might be outside these constraints, especially in cases where we add too many of them. |
View / edit / reply to this conversation on ReviewNB cetagostini commented on 2024-10-07T14:30:22Z Could we mention how uncertainty increases given the new level of spending? Historically spending around display, for example, was 200K to 250K and we are recommending 350K or so. This is reflected in the posterior uncertainty when estimating the OOS.
ps: We can preface this by mentioning that this risk/best option dynamic will be something that will soon be added to the optimizer. |
View / edit / reply to this conversation on ReviewNB cetagostini commented on 2024-10-07T14:30:23Z I think it would be interesting to make a comparison with total sales with and without optimization, with their respective intervals. What do you think? It should be relatively simple. |
Closes #1014
📚 Documentation preview 📚: https://pymc-marketing--1044.org.readthedocs.build/en/1044/