Colt Allen

The [ArviZ library](https://python.arviz.org/en/stable/examples/index.html) also has a lot of useful plotting options to compliment whatever we add in `pymc-marketing`, and examples would be a great addition to the tutorial notebooks. I've...

Hey @jcfisher, There is already a PR open for the method you are proposing: https://github.com/pymc-labs/pymc-marketing/pull/216 It was abandoned by its original author quite some time ago though, so it may...

@wd60622 this is the utility function which will need to be adapted from `lifetimes` to plot cumulative transactions over time: https://github.com/CamDavidsonPilon/lifetimes/blob/master/lifetimes/utils.py#L506

I could also see [RFM Segmentation](https://github.com/pymc-labs/pymc-marketing/issues/523) being valuable for this, because customers in different segments would benefit from different strategies, which can be tied back directly to marketing budgets and...

I added the likelihood expression to the docstrings, along with a dev notebook. Not sure if the LaTeX rendered properly in the docstrings though, because the docs build is failing...

> I pushed a version of the logp that works with Scan. I think we can make it work without a Scan, by evaluating on the maximum T-tx range. I...

Tests are still failing with `vectorize_graph`. Do the library dependencies require updating?

I had to increase `rtol` because we're comparing against the data simulation function in `lifetimes`, and it lacks a random seed. If the `sim_data` function looks good and I'm indexing...

Hey @sarim-zafar, Thanks for sharing. The Jenks Breaks methodology has some parallels with [Dirichlet processes](https://en.wikipedia.org/wiki/Dirichlet_process), which can be used to automatically infer the optimal number of clusters.

I've decided to merge efforts with the PyMC Labs team and work on the [pymc-marketing](https://github.com/pymc-labs/pymc-marketing) project, which will become the premiere solution for CLV modeling going forward. BTYD has been...