Hierarchical Market Mix Modeling

[shuvayan]

Hello Rockstars,

I am trying to develop a hierarchical bayesian regression model across separate cities by following the below article :

[(https://towardsdatascience.com/bayesian-hierarchical-marketing-mix-modeling-in-pymc-684f6024e57a#:~:text=Bayesian%20Marketing%20Mix%20Modeling%20in%20Python%20via%20PyMC3&text=The%20Bayesian%20approach%20works%20exceptionally,example%2C%20spendings%20across%20several%20countries%3F)]

This doesn't work though due to some data structure issue with aesara tensor and hence I was looking for something within the DelayedSaturatedMMM framework. /
Can someone please let me know if there is way to implement Hierarchical models using DelayedSaturatedMMM?

[cetagostini]

Hey @shuvayan,

Just wanted to let you know that we're working on an update for the delayed saturated MMM. You can check out the progress here #460 . It's not quite ready yet, but you can use it as inspiration to create your own custom model in the meantime. We're planning to merge it in the next few weeks probably.

Best,

[shuvayan]

Hello Pymc-Labs/Pymc-Marketing, Thank you so much for letting me know. I am eagerly waiting for the update and raring to apply it on our use case. We know that Robyn does not have this feature and that excites us more as we are a Py team. All the best!!

…

On Wed, 24 Jan 2024, 21:45 Carlos Trujillo, ***@***.***> wrote: Hey @shuvayan <https://github.com/shuvayan>, Just wanted to let you know that we're working on an update for the delayed saturated MMM. You can check out the progress here #460 <#460> . It's not quite ready yet, but you can use it as inspiration to create your own custom model in the meantime. We're planning to merge it in the next few weeks probably. Best, — Reply to this email directly, view it on GitHub <#497 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AC2NL5O2KO54DC6ZLK64EPLYQEXTRAVCNFSM6AAAAABCGZBF42VHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMYTSMBYGQ2TQMBRHA> . You are receiving this because you were mentioned.Message ID: ***@***.***>

[luiztauffer]

hi @shuvayan did you have any progress on this? I'm interested in a similar solution for one of my projects, would be great to hear form someone who has already tried it

[cetagostini]

hi @shuvayan did you have any progress on this? I'm interested in a similar solution for one of my projects, would be great to hear form someone who has already tried it

Hey current PR is #1036 and its in development