Colt Allen

Hey @HCM00, I'm one of the contributors to [pymc-marketing](https://github.com/pymc-labs/pymc-marketing). We're still working on adding validation capabilities at this time, but for modeling concept drift on out-of-sample data in practice, I...

> I'm curious about the reasoning behind weighting this metric by days since the last visit (T - recency). Could you please provide more context or details on how this...

@williambdean I definitely think this is worth implementing for `clv.rfm_summary`. We'll need to keep in mind that function is also called under the hood for `rfm_train_test_split` and `rfm_segmentation`. Do you...

> pm.sample accepts `mp_ctx` keyword argument to define the multiprocessing strategy that is used. The Prefect issue arose in both MacOS and Linux. For the latter maybe the fix is...

Hey @wd60622, Thanks for starting on this. Looks like we should make some progress on https://github.com/pymc-labs/pymc-marketing/issues/527 first and get these methods renamed for consistency before continuing.

> @ColtAllen, is there a concrete issue with this? I've updated https://github.com/pymc-labs/pymc-marketing/issues/527 with the list of steps/PRs that need to be taken.

The `BetaGeoModel.expected_purchases` method has been added, so this should be unblocked now and able to proceed.

Hey @asifzubair, resharing the link @ricardoV94 shared on [improper distribution sampling](https://stats.stackexchange.com/questions/211917/sampling-from-an-improper-distribution-using-mcmc-and-otherwise). Are you able to add a test to meet @jessegrabowski's criteria? > To merge the PR, I would want...

The research paper listed several non-marketing applications for this model, so it seemed general enough to add here. Otherwise I'm indifferent.

Closing because although this distribution is generalized in application, it has several post-inference derivations that make more sense to bundle into a dedicated model in `pymc-marketing`.