Add new customer_choice module (#803)

* Update version.txt (#640)

* [pre-commit.ci] pre-commit autoupdate (#642)

* Fix build badge (#645)

* Add downloads stats to README

* Pareto/NBD Example Notebook (#646)

* notebook opening and imports

* model definition markdown

* Data Load Notebook section

* WIP model fitting section

* added notebook to docs directory

* notebook edits and graph code

* ppc section and nb cleanup

* demz sampling and WIP plotting

* WIP predictive plots

* WIP heatmap plots

* predictive plots

* WIP covariates and nbqa-ruff edits

* covariate section

* plot additions

* fig sizes

* remove model file

* add spaces, increase indentation, and fix number order to Pareto notebook (#651)

* add spaces, increase indentation, and fix number order

* explicit with 6

* Add link to new Pareto notebook (#649)

* Plot Waterfall Components Decomposition (#631)

* Creating plot waterfall

Co-Authored-By: Carlos Trujillo <[email protected]>

* requested changes

* pre-commit

---------

Co-authored-by: Carlos Trujillo <[email protected]>

* Update resources.md (#652)

Databricks should have a lower-case b.

* fix ylabel (#654)

* [pre-commit.ci] pre-commit autoupdate (#655)

updates:
- [github.com/astral-sh/ruff-pre-commit: v0.4.1 → v0.4.2](astral-sh/ruff-pre-commit@v0.4.1...v0.4.2)
- [github.com/pre-commit/mirrors-mypy: v1.9.0 → v1.10.0](pre-commit/mirrors-mypy@v1.9.0...v1.10.0)

Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>

* Lift test fixes (#656)

* support for negative values and dates (if used)

* fix terrible spelling

* test dates in coords

* cover numpy objects

* consolidate the tests

* add to docstrings tanh saturation (#657)

Co-authored-by: Juan Orduz <[email protected]>

* Add quickstart to readme (#653)

* add quickstartt to readme

* add pymc labs logo

* typos

* add community links

* fixes

* Add contributors to README (#659)

* Add contributors to README

* small code style improvements

* Time varying intercept (#628)

* Add time-varying prior functionality to DelayedSaturatedMMM

* resolve wd's comments

* resolve failing pre-commits

* add tvp_kwargs to model_config

* fix typo

* replace softplus

* resolve minor review comments

* Add option to supply `ax` to `plot_posterior_predictive`

* bugfix: time_index was not set correctly for OOS

If provided `X` to `_data_setter` was *not* the immediate sequence following the training set, the time_index would we wrong. With this fix, the `date["time_index"]` which gets set upon providing a new `X`, is inferred from the `self.date_column` column of provided `X` (by comparing it to same date column in the training data `self.X`).

* Clean up example notebook

* Make utility function `transform_1d_array`

* 'tvp_kwargs' -> 'intercept_tvp_kwargs'

* move `infer_time_index` into utils

* add tests for new utils

* small fixes (found in tests)

* add tests to cover all added cases

* fix ruff check

* update typehints

* resolve review comments

* refactor model logic for tv intercept

* address review comment for util test

* .

* fix documentation link

* change variable name

* fix hsgp_dims

* update time_varying_prior to be centered on 1

* review fixes

* fix broken test

* add final tests

* fix coverage issues

* Update tests/mmm/test_tvp.py

Co-authored-by: Will Dean <[email protected]>

* Update pymc_marketing/mmm/tvp.py

Co-authored-by: Will Dean <[email protected]>

* Update tests/mmm/test_tvp.py

Co-authored-by: Will Dean <[email protected]>

* Update tests/mmm/test_tvp.py

Co-authored-by: Will Dean <[email protected]>

* significant improvements to notebook

* fix heading

* update notebook to make it EVEN better

* update legend, add watermark

* fix intro

* fix broken test

* copy sweep with grammarly

---------

Co-authored-by: Will Dean <[email protected]>

* Update README.md (#660)

* Add tv intrecept to readme (#661)

* add tv intrecept to readme

* add to comparison table

* fix title level (#663)

* Remove unnecessary NonImplemented errors from abstract methods (#662)

* ignore non-implemented

* remove not implemented error from abstract classes

* simplify docstrings

* Pass conv mode to adstock functions (#665)

* [Try] Fix compressed images in docs. (#667)

* Update pyproject.toml (#671)

* add license (#673)

* use grep and sed in the env line (#675)

* add sample_kwargs (#676)

* MMM NB Improvements (waterfall & error plots) (#664)

* Update version.txt (#677)

* [pre-commit.ci] pre-commit autoupdate (#683)

updates:
- [github.com/astral-sh/ruff-pre-commit: v0.4.3 → v0.4.4](astral-sh/ruff-pre-commit@v0.4.3...v0.4.4)

Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>

* v0 Streamlit MMM Explainer App (#614)

* feat(streamlit_explainer): Pushing files for Streamlit explainer app, to illustrate saturation, adstock and prior concepts in an intuitive, visual way to stakeholders and new MMMers

* chore(readme): Adding a readme for the app

* fix(env): Updating dependencies to include those needed for the Streamlit app

* Drop python 3.9 support (#615)

* drop python 3.9

* try python 3.12

* undo try python 3.12

* add lift tests check

* Add more content to the Gamma-Gamma Notebook  (#573)

* improve nb

* rm warnings and add link to lifetimes quickstart

* address comments

* feedback part 3

* remove warnings manually

* Add more content to the BG/NBD Notebook (#571)

* add more info to the notebook

* hide plots code

* fix plot y labels

* fix plot outputs and remove model build

* improve final note probability plots

* address comments

* use quickstart dataset

* feedback part 3

* remowe warnings manually

* feedback part 4

* Improve MMM Docs (#612)

* improve mmm docs init

* add more code examples to docstrings

* minor improvemeents

* typo

* better phrasing

* add thomas suggestion

* Fix `clv` plotting bugs and edits to Quickstart (#601)

* move fixtures to conftest

* docstrings and moved set_model_fit to conftest

* fixed pandas quickstart warnings

* revert to MockModel and add ParetoNBD support

* quickstart edit for issue 609

* notebook edit

* [pre-commit.ci] pre-commit autoupdate (#616)

* improve coords matching (#623)

* python 3.12 attempt (#618)

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci

* refactor(saturation): Using pymc-marketing saturation functions rather than coding my own: Removing tanh, logistic and michaelis menten

* refactor(saturation): Remove Hill and Root saturations, as they aren't supported by pymc-marketing currently

* refactor(geometric_adstock): Removing custom adstock and using pymc-marketing adstock function to demo decay. Also updating latex to align with pymc-marketing, where decay factor is represented by alpha rather than beta

* refactor(delayed_adstock): Using pymc-marketing delayed geometric function rather than custom one

* fix(requirements): Adding pymc-marketing to Streamlit requirements for deployment

* Added Dev Container Folder

* refactor(weibull_cdf): Using pymc-marketing function for Weibull CDF

* fix(weibull_cdf): Fixing incorrect dataframe var name for CDF plotting df

* refactor(weibull_pdf): Using pymc-marketing function for WeibullPDF

* refactor(custom_functions): Removing adstock_saturation_functions.py file now that it is no longer required

* chore: Removing devcontainer created by Streamlit

* fix(requirements): Adding preliz to requirements

* refactor(prior_viz): Reworking the prior visualisation to use Preliz instead of custom function, as well as remove the tab-design. Prior distributions can now be specified programmatically.

* refactor(prior_functions.py): Deleting the draw_samples function and replacing it with a programmatic PreliZ function, such that the distribution object is returned when the user passes in the name of a distribution

* fix(requirements): Delete obsolete pymc requirement, which should fix deployment dependency conflicts

* chore(readme): Updating with guidelines on how to add additional distributions or transformation functions to the app

* refactor(plot_config): Moving height and width specifications into constants at top of Adstock and Saturation files, so the plot sizes are set programmatically

---------

Co-authored-by: Juan Orduz <[email protected]>
Co-authored-by: Colt Allen <[email protected]>
Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
Co-authored-by: Carlos Trujillo <[email protected]>

* Correct BetaGeo docstring example (#693)

* fix example in docstring

* change docstring

* Add `BetaGeoBetaBinom` Distribution Block (#431)

* init commit

* removed scan

* Fix logp

* Remove print statement

* Add test for logp notimplemented errors

* docstrings

* dev notebook added

* updated to vectorize_graph

* import order

* update oldest pymc_version

* Update ci.yml pymc version

* Update pyproject.toml pymc version

* WIP sample prior testing

* sample prior compared against lifetimes

* increase rtol

* remove commented code, add logp reference

* fix latex docstring

* notebook testing and misc edits

* revert latex in docstring

* add ruff ignore comment

---------

Co-authored-by: Ricardo Vieira <[email protected]>

* [pre-commit.ci] pre-commit autoupdate (#705)

* Fix related to column renaming after aggregating test frequency (#698)

Co-authored-by: Colt Allen <[email protected]>

* RFM Segmentation (#680)

* init rfm_segments func

* TODOs

* docstrings and for loop

* docstrings and for loop

* WIP dev notebook debugging

* checkpoint commit for remote pull

* code testing in dev notebook

* unit tests added

* dev notebook cleanup

* clean up type hints

* comments and code cleanup

* docstrings

* move formatting to rfm_summary and quickstart edits

* fix rfm_train_test_split bug

* added test for rfm_quartile_labels

* added rfm score warning

* create bgbb_donations.csv (#710)

* closes #678 (#716)

* use URL for README image (#715)

* use URL for image

* additional links that are relative

* remove forward slashes in URL

* closes #264 (#714)

* [pre-commit.ci] pre-commit autoupdate (#719)

updates:
- [github.com/astral-sh/ruff-pre-commit: v0.4.5 → v0.4.7](astral-sh/ruff-pre-commit@v0.4.5...v0.4.7)

Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>

* Update `BetaGeoModel` API (#709)

* _extract_predictive_variables util

* deprecation warnings

* expected_purchases_new_customer

* TestBetaGeoModel.setup_class

* test cleanup and test_expected_purchases_new_customer

* expected_probability_alive

* expected_purchases

* TODOs and docstrings

* update runslow tests rtol for new test dataset

* prob_alive_matrix plot fix and notebook testing

* alive loop in bgnbd nb

* quickstart nb fix

* docstring and TODO revisions

* docstring syntax

* docstring edits

* more docstring fixes

* docstrings indent

* docstring indent, clear codecov bug

---------

Co-authored-by: Juan Orduz <[email protected]>

* User-defined media transformations and custom ordering (#632)

* Allowing Custom Saturation & Lagging functions

Co-Authored-By: Carlos Trujillo <[email protected]>

* Small adding

Co-Authored-By: Carlos Trujillo <[email protected]>

* test commit to carlos branch

* hook up saturation class to lift method

* pull out last saturation function into method

* migrate the logic for tests

* reduce what is needed for a new class

* remove the noqa

* model docstring for inheritance

* rewrite of the components

* rewrite of the components

* change name to logistic

* map names for model_config

* switch out order

* Changes

* solving issues

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci

* solving errors

* New class name

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci

* Update pymc_marketing/mmm/transformers.py

Co-authored-by: Will Dean <[email protected]>

* pre-commit change

* Update pymc_marketing/mmm/models/components/base.py

Co-authored-by: Will Dean <[email protected]>

* Update pymc_marketing/mmm/models/components/saturation.py

Co-authored-by: Will Dean <[email protected]>

* Update pymc_marketing/mmm/models/components/lagging.py

Co-authored-by: Will Dean <[email protected]>

* Requested changes

* pre-commit changes

* Requested changes

* more changes

* use reserved parameter name

* Returning name and adding deprecation warning

* Re-structuring the budget optimizer

* Update pymc_marketing/mmm/components/adstock.py

Co-authored-by: Will Dean <[email protected]>

* Will recommendations

* Temporal comment

* test more assumed behavior

* more budget optimizer changes

* docstrings for classes

* small changes about l_max on adstock

* small addition

* Update pymc_marketing/mmm/delayed_saturated_mmm.py

Co-authored-by: Will Dean <[email protected]>

* temporal notebook copy

* update

* docstring

* adding missing

* Update pymc_marketing/mmm/components/adstock.py

Co-authored-by: Will Dean <[email protected]>

* Update pymc_marketing/mmm/components/adstock.py

Co-authored-by: Will Dean <[email protected]>

* Making mypy happy part 1

* Small example

* finally?

* small change

* docstrings

* removing mokeytype

* more docstrings

* Update pymc_marketing/mmm/transformers.py

Co-authored-by: Will Dean <[email protected]>

* tests for out of box saturation and adstocks

* test the passthrough

* Update pymc_marketing/mmm/components/adstock.py

Co-authored-by: Will Dean <[email protected]>

* Changes on hints

* modifications

Must be plt, ax

* notebooks addition

* Update notebook

* Saving model

* mock the fit and move tests around

* fix get_distribution tests

* Correct error

* Deprecation warning and new MMM class

Co-Authored-By: Will Dean <[email protected]>

* Small changes

* updates

* add a lookup for model saving

* fix tests based on recent changes

* Deleting decimals

* test suite

* Update tests/mmm/test_budget_optimizer.py

Juan's test suggestion

* Update tests/mmm/test_budget_optimizer.py

abs and not rel

* running notebooks

* adding hint

* initial docstring check

* update docstrings 1

* Modifying notebook

* use explicit adstock and saturation in MMM

* remove reference to DelayedSaturatedMMM

* request changes

* missing total

* modify docstring

* modify eval

* Small changes

* change

* changes

* import saturation in docstring

* pre-commit

---------

Co-authored-by: Carlos Trujillo <[email protected]>
Co-authored-by: Will Dean <[email protected]>
Co-authored-by: Will Dean <[email protected]>
Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>

* CLV Plotting API (#728)

* plot_probability_alive_matrix

* docstrings

* plot_frequency_recency_matrix

* delete dead code

* docstring quick fix

* [pre-commit.ci] pre-commit autoupdate (#730)

updates:
- [github.com/astral-sh/ruff-pre-commit: v0.4.7 → v0.4.8](astral-sh/ruff-pre-commit@v0.4.7...v0.4.8)

Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>

* Fix some ParetoNBDModel docstring typos (#731)

* Removed the comma after Schmittlein

* Added missing reference

* Removed extra colon

* pass kwargs to minimizer (#737)

* Minor improvements [MMM] (#735)

* Set upper bound pymc 5.16 (#725)

* changes init

* try fix regex

* Media transformation sampling & plotting methods (#734)

* add plotting methods

* add tests for new methods

* saturation support for additional variable dims

* consolidate the logic of sampling

* change warning

* workflow from a fitted model

* change order of tests

* suggestion to use names

* because of new data

---------

Co-authored-by: Juan Orduz <[email protected]>

* improve tests mmm utils (#738)

* improve tests

* empty commit

* remove reduntant function

* `model.fit` doesn't remove prior samples (#741)

* type hint only

* more informative errors

* check for attr

* remove type ignore

* check for attr

* check for attr

* reduce indentation

* new error names

* Hierarchical Model Configuration (#743)

* some base logic and tests

* lookup function once

* add error handling

* implement for mmm and media transformations

* add examples

* add to documentation

* add to docstring

* tests for likelihood

* use deepcopy since keys are added

* set default dims and warn

* fix output_var

* migrate failing tests to model_config

* remove the moved test

* use deepcopy since keys are added

* add to docstrings from feedback

* fix handlers at initialize

* [pre-commit.ci] pre-commit autoupdate (#756)

updates:
- [github.com/astral-sh/ruff-pre-commit: v0.4.8 → v0.4.9](astral-sh/ruff-pre-commit@v0.4.8...v0.4.9)

Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>

* MMM Component Notebook (#748)

* initial notebook

* add to example

* push up some feedback

* change the number of channels

* updates

* final message

* add more feedback

* remove since wasnt working the way I wanted

* start addressing sphinx warnings and rendering issues (#750)

* start addressing sphinx warnings and rendering issues

* forgot formatting comment

* Add numpydoc as dependency

* fix dataset type

---------

Co-authored-by: Juan Orduz <[email protected]>

* Allowing Hierarchical Non Centered Parametrization (#747)

* Allowing non center parametrization

* update notebook

* Adding example in docstring

* change

* Push code changes.

* A painful and ugly change!

The things one does for democracy!

* Missing parts!

* adding missing test 2D

* Missing raise

---------

Co-authored-by: Will Dean <[email protected]>

* fix np typing (#763)

* fix typing

* empty

* empty

* empty

* dummy

* undo dummy

* empty

* add it back (#764)

* remove noqa from plots (#761)

* remove noqa

* fix escape

* empty

* empty

* Creating Time Base component for Media Contribution (#752)

* Run Ruff Notebooks (#773)

* [pre-commit.ci] pre-commit autoupdate (#779)

updates:
- [github.com/astral-sh/ruff-pre-commit: v0.4.9 → v0.4.10](astral-sh/ruff-pre-commit@v0.4.9...v0.4.10)

Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>

* `GammaGammaModel` API Improvements (#758)

* utils.customer_lifetime_value

* expected_customer_lifetime_value

* WIP clv.models.gamma_gamma.py

* gamma_gamma API

* fixed circular import

* gamma_gamma tests

* delete tests/datasets/test_summary.csv

* clv test_utils.py

* remove expected_purchases(future_t=0)

* remove monetary_value arg

* WIP docstrings

* notebooks

* docstrings

* Revert "notebooks"

This reverts commit a3154d9.

* gamma-gamma notebook

* docstrings

* Deepcopy of posterior to allow second `fit` call (#790)

* Add prior predictive example notebook (#787)

* CLV Modeling Domains and Docstrings (#785)

* quickstart

* pareto_nbd fit warning

* pareto_nbd docstrings

* clv.utils docstrings

* gamma_gamma docstrings

* beta_geo docstrings

* CLV sections in README and index

* fix nb (#793)

* Run example notebooks CI (#791)

* run notebooks init

* change kernel name

* change kernel name

* change kernel name

* rm docs

* add nb

* fix nb

* fig dpi

* update model object

* no output

* test mmm notebooks

* fix path

* graphviz

* sudo

* paralelize and Path

* reqs

* undo

* comments

* try budget allocation

* add quickstart

* ignore budget allocation

* add make command

* use make

* docs: Update model_builder.py to resolve warning in documentation build (#797)

* Time Varying Media Contribution Notebook (#778)

* Time Varaying notebook

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci

* requested changes

* quick change

* Adding to index md

* remove change mmm_budget_allocation_example.ipynb

* quick correction

* Small changes

* Small title change

---------

Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
Co-authored-by: Will Dean <[email protected]>

* prepare release (#799)

* `Prior` class to represent distributions (#759)

* reimplement hierarchical parameters with wrapper class

* write tests for prior class

* explicit with dependencies

* missing test dependency

* raise KeyError early and test dims

* add prior module to docs

* support for media transformations

* add some documentation

* migrate the delayed_saturated model to Prior class

* migrate the delayed_saturated model to Prior class

* add an example image

* add the example image

* push up some more documentation

* support for student-t

* add configuration notebook

* catch all the errrors

* add more to notebook

* use a sigma which has transform

* add parsing error catches

* add watermark

* test the parsing warnings and error handling

* have a default to dict

* default handles Prior class

* passing plotting tests

* support for lift tests

* fix two failing tests

* update notebook

* change order of pre-commit back

* migrate to Prior and parse_model_config

* migrate the clv tests

* test for warning

* run pre-commit on all files

* switch to Prior in components

* test for the components

* remove the previous dictionary function

* migrate to Prior class in mmm

* remove the unused key,value

* migrate to Prior class in clv tests

* back to dict for warning test

* isort and work out example indentation

* change the example name

* use juans suggestions

* change graph to to_graph

* rerun the time varying media

* rerun the clv notebooks I could get to work

* [pre-commit.ci] pre-commit autoupdate (#801)

updates:
- [github.com/astral-sh/ruff-pre-commit: v0.4.10 → v0.5.0](astral-sh/ruff-pre-commit@v0.4.10...v0.5.0)
- [github.com/pre-commit/mirrors-mypy: v1.10.0 → v1.10.1](pre-commit/mirrors-mypy@v1.10.0...v1.10.1)

Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>

* initial stab at some docs

* register new module in a few places

* initial stab at how-to example notebook + code + API docs

* Pull out seasonality as `YearlyFourier` and `MonthlyFourier` (#802)

* Separate Weibull adstock into CDF & PDF (#810)

* separate out weibull

* add to test suite

* Add typing and package classifiers (#811)

* add pytyped

* add the py.typed file

* add imports to mmm module (#812)

Co-authored-by: Juan Orduz <[email protected]>

* Friday progress on notebook example

* Save & load support for time varying parameters (#815)

* add missing init for save and load

* get rid of warnings from JSON parsing

* new error message without line break

* migrate to Data and non-mutable coords (#816)

Co-authored-by: Juan Orduz <[email protected]>

* improve introduction page

* updates to model priors, plotting market shares, and more

* Add in model maths for MV-ITS, saturated market assumption

* [pre-commit.ci] pre-commit autoupdate (#817)

updates:
- [github.com/astral-sh/ruff-pre-commit: v0.5.0 → v0.5.1](astral-sh/ruff-pre-commit@v0.5.0...v0.5.1)

Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>

* fix up / improve the model maths and description

* PoC: Use Pydantic as data validator (#809)

* prior with pydantic

* dependencies

* validate adstock

* make mypy happy

* add validation sample curve

* make the prior type tighter

* add test type

* add validation init mmm

* mmm

* start with Fourier

* fix type

* fix test and imprtove docstrings

* docstrings

* types

* self type

* init validator

* types model builder

* improve docstrings

* more input validations mmm init

* validation budget optimizer

* fix dummy example types

* hsgp kwargs class

* add kwargs

* undo type hint in dict

* fix fourier names

* better docs

* fix tests

* add type hint

* undo

* fix type error

* feedback2

* restrict signature

* serialize fourier

* docs and tests

* fix docs

* work on parsing

* add hsgp to parsing config

* add tests

* uncomment

* undo changes

* undo model config parser

* handle hsgp_kwargs

* add hsgp flag

* docs

* undo type hint

* improve hints

* add more sections to docs

* Update pymc_marketing/mmm/tvp.py

Co-authored-by: Will Dean <[email protected]>

* feedback 4

* fix test

---------

Co-authored-by: Will Dean <[email protected]>

* fixes + describe covariance matrix + add section on unsaturated market model

* update title of intro docs page

* add admonition box

* initial stab at modeling products individually

* fix scenario 3 sampling & commentary + allow user to supply sampler_kwargs

* Date Validation and MMM Model Hamonization (Pydantic) (#824)

* validate base mmm init class

* validate dateformat

* add comment about date

* remove ()

* close to done on docs + code?

* add tests

* enlightenment in terms of unsaturated markets

* extract data generation functions into module code

* remove old generate_data

* constrained -> saturated. parameterize a test

* turn generate_saturated_data into a test fixture

* split plot_causal_impact into plot_causal_impact_sales and plot_causal_impact_market_share

* execute notebook

* revert an accidental change from a global find & replace

* Remove warnings during tests (#823)

* address save and load tests

* catch warning on load

* remove warnings in budget optimizer

* remove plotting warnings

* remove validating warnings

* consolidate the loading function

* remove warnings in tests

* incorporate the docstring feedback

* only one deprecation warnings on DelayedSaturatedMMM

* dont have deprecation on test

* Future-proof `prior_linearized` method call (#806)

* modified: pymc_marketing/mmm/tvp.py

* Update tvp.py

Changed Xs to X in prior_linearized( as per the change in pymc)

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci

---------

Co-authored-by: sangeedutta <[email protected]>
Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
Co-authored-by: Will Dean <[email protected]>

* Create inverse_scaled_logistic_saturation and the corresponding class (#827)

* fix (#842)

* use labs theme as external (#830)

* use labs theme as external

* Update docs/source/index.md

Co-authored-by: Thomas Wiecki <[email protected]>

* install theme from pypi

---------

Co-authored-by: Thomas Wiecki <[email protected]>
Co-authored-by: Juan Orduz <[email protected]>

* specify 0.9.0 as deprecation version (#849)

* add intercept and target variable to example (#850)

Co-authored-by: Juan Orduz <[email protected]>

* Various MMM small documentation fixes (#854)

* make nb ruff astral-sh/ruff-vscode#546

* fix docs

* fix quickstart

* add types

* Update UML diagrams (#856)

* update uml diagrams

* add uml command to Makefile

* Fixing ruff commands in Makefile #825 (#859)

* updated the ruff lint command, added ruff code formating command in Makefile, updated contributing.md

* fixing the issue with the new line at the end of makefile

* updated the PHONY list in makefile

---------

Co-authored-by: Murad Khalilov <[email protected]>

* [pre-commit.ci] pre-commit autoupdate (#855)

updates:
- [github.com/pre-commit/mirrors-mypy: v1.10.1 → v1.11.0](pre-commit/mirrors-mypy@v1.10.1...v1.11.0)

Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
Co-authored-by: Juan Orduz <[email protected]>

* Fix model_builder docstrings (#861)

* Fix model_builder docstrings

* Fix whitespace error

---------

Co-authored-by: Juan Orduz <[email protected]>

* add GH discussions link to README (#866)

* add GH discussions link to README

* add link to MMM hub slack

* Update README.md

---------

Co-authored-by: Will Dean <[email protected]>

* Skip coords with scalar value (#868)

* Fix Visual for hill_saturation function (Issue #851 ) (#857)

* Fix plotting by evaluating tensors.

* Add space after sphinx directive.

* Remove indentation from blank line.

* Add shared y axis for subplots.

---------

Co-authored-by: Patrick Robotham <[email protected]>
Co-authored-by: Will Dean <[email protected]>

* Allow plot MMM components in the original scale (#870)

* add original scale implementation

* add plot nb

* change location

* undo

* make mypy happy

* test plot

* add test

* update plot readme

* fix test

* improve variable description

* Inference changed to dataset (#873)

* infernece changed to dataset

* inference changed dataset for plot_allocated_contribution_by_channel

---------

Co-authored-by: Will Dean <[email protected]>

* Add root saturation function (issue #702) (#858)

* feat: adding root_saturation to transformers.py

* feat: adding RootSaturation class to saturation.py

* chore: adding missing RootSaturation to SATURATION_TRANSFORMATIONS

* feat: adding root_saturation to transformers.py

* feat: adding RootSaturation class to saturation.py

* chore: adding missing RootSaturation to SATURATION_TRANSFORMATIONS

* chore: linting edits

* chore: adding coefficient to function

* chore: linting corrections

* chore: removed empty References section of docstring

* chore: produce visual examples of root saturation

* chore: adding root to test_saturation.py

* chore: adding RootSaturation to init file

---------

Co-authored-by: ruari.walker <[email protected]>
Co-authored-by: Will Dean <[email protected]>

* Check for missing attrs after `sample_prior_predictive` and `fit` (#867)

* separate the attr creation from attachment and perform check

* remove data for CLV

* fix model_builder tests

* fix clv tests

* more specific model builder checks

* rework with no args and kwargs

* rework common load method

* Update pymc_marketing/model_builder.py

* `json.loads` with python types bug (#881)

* loads doesnt support boolean

* defaults for the media transformation

* test for the time_varyign

* [pre-commit.ci] pre-commit autoupdate (#883)

updates:
- [github.com/astral-sh/ruff-pre-commit: v0.5.4 → v0.5.5](astral-sh/ruff-pre-commit@v0.5.4...v0.5.5)

Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>

* Fix default `ConvMode` in docstring (#864)

* Update docs in transformers.py

docs: Fix the docstring

* Update docs in transformers.py

docs: fix other docstrings

* change the plot default

---------

Co-authored-by: Will Dean <[email protected]>
Co-authored-by: Will Dean <[email protected]>

* Save off media transformations (#882)

* to_dict via lookup_name

* parse to and from dict for attrs

* improve the codecov

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci

* change test with change in default behavior

* increase the MMM model version

---------

Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>

* Update deployment docs (#887)

* update the notebook

* update the runner

* address feedback. add watermark

* Automate UML creation (#886)

* attempt to automate uml

* install graphviz binaries

* use sudo

* Update UML Diagrams

* remove the git configuration

* use the identity of the previous commit

* dont push with no change

* same triggers as others

* restrict to running with repo_changes

---------

Co-authored-by: GitHub Actions <[email protected]>
Co-authored-by: Juan Orduz <[email protected]>

* prepare release (#888)

* Update README.md (#893)

* Add url health job from streamlit app (#902)

* add url health job

* lint

* small improvements model config nb (#906)

* small improvements

* trim

* Add link model deployment to example notebook (#904)

* add link model deployment

* add pymc-marketing to watermark

* Add pymc-marketing version to some MMM notebooks (#907)

* nb part 1

* add mmm example nb

* rm example nb

* Move adstock and saturation method imports to mmm.__all__ (#908)

* Resolves #892: Move adstock and saturation method imports to mmm.__all__

* fix: patch uml GHA

* Update .github/workflows/uml.yml

* Update .github/workflows/uml.yml

* Add permissions: write-all to .github/workflows/uml.yml

---------

Co-authored-by: Will Dean <[email protected]>

* chore(Makefile): Adding a self-documenting command and light command documentation (#910)

Co-authored-by: Juan Orduz <[email protected]>

* Fix uml permissions (#913)

* [pre-commit.ci] pre-commit autoupdate (#914)

updates:
- [github.com/astral-sh/ruff-pre-commit: v0.5.5 → v0.5.6](astral-sh/ruff-pre-commit@v0.5.5...v0.5.6)
- [github.com/pre-commit/mirrors-mypy: v1.11.0 → v1.11.1](pre-commit/mirrors-mypy@v1.11.0...v1.11.1)

Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
Co-authored-by: Will Dean <[email protected]>

* Don't run tests for non-code changes (#898)

* test running budget allocator nb (#919)

* make hill pass through the origin (#920)

* Add MMM ROAS Priors Case Study (#916)

* model and data init

* make it work

* add intro

* add nb to index and references

* first complete iteration

* improve formatting

* split paragraph

* pyprojroot

* typos

* add conclusion section

* improve intro

* typos and feedback

* typo (#923)

* [pre-commit.ci] pre-commit autoupdate (#926)

updates:
- [github.com/astral-sh/ruff-pre-commit: v0.5.6 → v0.5.7](astral-sh/ruff-pre-commit@v0.5.6...v0.5.7)

Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>

* Original hill function definition (#925)

* MLflow autologging (#921)

* Adding BLAS to the env generation.

* Solving optimizer issues & typos (#933)

* Correcting typo num_days by horizon

* Correcting typo num_days by horizon and scaler

* Running notebooks

* Update UML Diagrams

* Rename horizon by periods

* Adding test requested to check budget outputs

* Running notebooks

* Update UML Diagrams

* Small notebook missing change.

* Correction in tests

* Change on name

* running notebook modifying function

* Update UML Diagrams

* Log number of posterior & tuning samples (#943)

* helper command to view the artifacts from test

* pass tune from kwargs

* test for support of all samplers

* add mlflow as a mock import

* actual import as autolog is missing from docs

* point to GH discussions (#944)

Co-authored-by: Juan Orduz <[email protected]>

* [pre-commit.ci] pre-commit autoupdate (#946)

updates:
- [github.com/astral-sh/ruff-pre-commit: v0.5.7 → v0.6.1](astral-sh/ruff-pre-commit@v0.5.7...v0.6.1)

Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>

* fix typo

* fix admonition block + markdown heading level

* collect results of plot funcs in tests

* add no cover comments

* explicit module level scope for fixture

* consistent use of product_incrementality for folder names

* background_sales -> existing_sales

* innovation_sales -> treatment_sales

* Fallback to defaults in `adstock|saturation_from_dict` (#955)

* Default saturation_from_dict to default_priors

* Default to AdstockTransformation.default priors in adstock_from_dict

* Add PyDocStyle Support (#951)

* change to normal likelihood truncated at zero

* use empirical estimate for HalfNormal sigma for observation noise std

* remove commented out lines

* use just one idata instance - removes idata_counterfactual

* make pre-commit checks pass

* break up code in __init__ into smaller methods

* remove @Property and deal with some consequences of that

* deprecate WeibullAdstock in favor of WeibullCDFAdstock and WeibullPDFAdstock (#957)

* deprecate in favor of WeibullCDFAdstock and WeibullPDFAdstock

* Update UML Diagrams

* Update UML Diagrams

* Raise informative error when including target in `X` (#962)

* add error when target is in X_df

* add test

* rename test

* format

* Ensure `fit` reproducibility (#963)

* use get instead of assign to default to sampler_config if exists

* default to what is given

* write tests based on the issue

* have defaults while not overriding

* Enforce `check_parameters` for `alpha` in `geometric_adstock` (#960)

* Enforce check_parameters in geometric_adstock

* Remove check_parameters on l_max. Revert to original test

* Add test_geometric_adstock_bad_alpha

* use ge, le instead of gt,lt

* Update tests/mmm/test_transformers.py

Co-authored-by: Will Dean <[email protected]>

* Simplify test parameters

* Update tests/mmm/test_transformers.py

Co-authored-by: Will Dean <[email protected]>

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci

---------

Co-authored-by: Will Dean <[email protected]>
Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>

* scaling should be done before hdi computation (#970)

* Fix UML permissions from forks / Run only on merge to main (#967)

* change repo and python file change to kick off

* comment to check for failure

* using the repo

* delete to trigger

* the previous trigger

* push up again

* check for the owner being pymc-labs

* remove unused

* only run after merge to main

* `DelayedSaturatedMMM` deprecations and moving files (#965)

* deprecations and moving files

* Update UML Diagrams

* change the imports in notebooks

* push up the code / test changes. need to run

* remove _get_\w*_function tests

* rerun the tvp notebook

* remove stale test

* move away from string initialization

* change the tvp media example

* Register and allow custom transform for `Prior` class (#972)

* allow register and use custom transform

* add to the example block

* Update pypi.yml (#975)

* [pre-commit.ci] pre-commit autoupdate (#977)

updates:
- [github.com/astral-sh/ruff-pre-commit: v0.6.1 → v0.6.2](astral-sh/ruff-pre-commit@v0.6.1...v0.6.2)
- [github.com/pre-commit/mirrors-mypy: v1.11.1 → v1.11.2](pre-commit/mirrors-mypy@v1.11.1...v1.11.2)

Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>

* Added deprecation warning for method keyword (#974)

Co-authored-by: Will Dean <[email protected]>
Co-authored-by: Juan Orduz <[email protected]>

* Improve and sync the README and landing page. (#978)

* add table

* sync index

* empty

* improvements

* Time slice cross validation MMM Notebook (#971)

* nb init

* add param stability

* crps init

* add crps to package

* make mypy happy

* clean

* relax req

* update path

* add crps test

* first iteration

* improvements

* improve tests

* improve tests

* improve references

* feedback 1

* add examples

* fix the wrong trigger (#983)

* Media transformation class and different transformations based on subsets of channels (#968)

* Changed pt=pt to pt_lib = None to avoid showing full module path in docs (#992)

* Changed pt=pt to pt_lib = None to avoid showing full module path in docs

* Changed pt to pt_lib inside calculate_lift_measurements_from_curve

* Linear trend (#991)

* push up some linear-trend work I had

* image for the documentation

* start the slopes at t=0

* additional checks at init

* import at mmm module

* update the image

* Move to reference section

* Move coords to single line

* Add hierarchical trend example

* add image for the example

* migrate to pydantic

* update images after seed

* Correct the latex

* fix check from tests

* revert link

* add item for new module

* Use ModelBuilder for incrementality model (#1101)

* use the modelbuilder mixin

* use Prior class for the background_distribution

* modify the tests

* test for mismatch

* add informed prior method

* begin to modify the existing notebooks

* add y to the mix

* changes to saturated notebook

* run with the accept kwargs

* rename product_incrementality module to cc

* fix typo

* add a placeholder quickstart docs page

* rename a folder (previously missed) product_incrementality -> cc

* more instances of product incrementality -> customer choice

* market_saturated -> saturated_market

* add quickstart/cc/index to the getting_started page

* avoid reusing `result` for multiple different scenarios

* notation change: rename "background" to "existing"

* add kwarg plot_total_sales to plot functions

* rename module cc -> customer_choice

* remove type hint on plot_total_sales to make tests pass

* get docs to build / pass tests

* finish the customer_choice getting started page

* fix typo

* remove manual tight_layout commands

* fix typo

* grammatical fix

* fix typo

* surpress a plot output

* fix typo

* update imports + split data generation code into different .py file

* fix imports in tests

* add docstrings to synthetic data module

* switch rng parameter to random_seed

* import from the customer_choice over submodule

* add docstrings to the model and plot function and mypy

* add file autolabeling

* allow for ax argument

* fix the test seed name

* change import

* increase tests / find some holes

* use random_seed instead of rng

* fix conflicts

* lint

---------

Co-authored-by: Juan Orduz <[email protected]>
Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
Co-authored-by: Colt Allen <[email protected]>
Co-authored-by: Will Dean <[email protected]>
Co-authored-by: Carlos Trujillo <[email protected]>
Co-authored-by: Carlos Trujillo <[email protected]>
Co-authored-by: Corey Abshire <[email protected]>
Co-authored-by: Ulf Aslak <[email protected]>
Co-authored-by: nialloulton <[email protected]>
Co-authored-by: Louis Magowan <[email protected]>
Co-authored-by: Ricardo Vieira <[email protected]>
Co-authored-by: Ivan Ugrin <[email protected]>
Co-authored-by: Will Dean <[email protected]>
Co-authored-by: Mews <[email protected]>
Co-authored-by: Oriol Abril-Pla <[email protected]>
Co-authored-by: c0d33ngr <[email protected]>
Co-authored-by: Shuvayan Das <[email protected]>
Co-authored-by: sangeedutta <[email protected]>
Co-authored-by: Arthur Mello <[email protected]>
Co-authored-by: Thomas Wiecki <[email protected]>
Co-authored-by: Murad Khalil <[email protected]>
Co-authored-by: Murad Khalilov <[email protected]>
Co-authored-by: Giannis_apost <[email protected]>
Co-authored-by: Christian Luhmann <[email protected]>
Co-authored-by: Patrick Robotham <[email protected]>
Co-authored-by: Patrick Robotham <[email protected]>
Co-authored-by: Ishaan Jolly <[email protected]>
Co-authored-by: Ruari Walker <[email protected]>
Co-authored-by: ruari.walker <[email protected]>
Co-authored-by: Maxim Kochurov <[email protected]>
Co-authored-by: GitHub Actions <[email protected]>
Co-authored-by: Pablo de Roque <[email protected]>
Co-authored-by: Dan Dean <[email protected]>
Co-authored-by: Pablo de Roque <[email protected]>
Co-authored-by: radiokosmos <[email protected]>
Co-authored-by: Juan Orduz <[email protected]>

Author: 35 people

.github/labeler.yml

@@ -15,6 +15,12 @@ CLV:
       - docs/source/notebooks/clv/*
       - pymc_marketing/clv/**
 
+"customer choice":
+  - changed-files:
+    - any-glob-to-any-file:
+      - docs/source/notebooks/customer_choice/*
+      - pymc_marketing/customer_choice/**
+
 "Prior class":
   - changed-files:
     - any-glob-to-any-file:

.github/pull_request_template.md

@@ -25,6 +25,7 @@
 <!--- Please list the modules that are affected by this PR by typing an `x` in the boxes below: -->
 - [ ] MMM
 - [ ] CLV
+- [ ] Customer Choice
 <!--- Additionally, if you are a maintainer or reviewer, please make sure that the appropriate labels are added to this PR -->
 
 ## Type of change

CONTRIBUTING.md

@@ -172,6 +172,16 @@ Classes
 
 ![](docs/source/uml/classes_clv.png)
 
+## Overview of the customer choice codebase
+
+Packages
+
+![](docs/source/uml/packages_customer_choice.png)
+
+Classes
+
+![](docs/source/uml/classes_customer_choice.png)
+
 ---
 
 This guide takes some inspiration from the [Bambi guide to contributing](https://github.com/bambinos/bambi/blob/main/CONTRIBUTING.md)

Makefile

@@ -49,6 +49,7 @@ uml: ## Install documentation dependencies and generate UML diagrams
 	pip install .[docs]
 	pyreverse pymc_marketing/mmm -d docs/source/uml -f 'ALL' -o png -p mmm
 	pyreverse pymc_marketing/clv -d docs/source/uml -f 'ALL' -o png -p clv
+	pyreverse pymc_marketing/customer_choice -d docs/source/uml -f 'ALL' -o png -p customer_choice
 
 mlflow_server: ## Start MLflow server on port 5000
 	mlflow server --backend-store-uri sqlite:///mlruns.db --default-artifact-root ./mlruns

docs/source/api/index.md

@@ -13,6 +13,7 @@
   model_config
   model_builder
   prior
+  customer_choice
   metrics
   mlflow
   deserialize

docs/source/getting_started/index.md

@@ -13,4 +13,5 @@ installation/index
 
 quickstart/clv/index
 quickstart/mmm/index
+quickstart/customer_choice/index
 :::

docs/source/getting_started/quickstart/customer_choice/index.md

@@ -0,0 +1,41 @@
+# Customer Choice Quickstart
+
+```python
+from pymc_marketing.customer_choice import (
+    MVITS,
+    generate_saturated_data,
+    plot_product,
+)
+
+# Generate simulated data
+scenario = {
+    "total_sales_mu": 1000,
+    "total_sales_sigma": 5,
+    "treatment_time": 40,
+    "n_observations": 100,
+    "market_shares_before": [[0.7, 0.3, 0]],
+    "market_shares_after": [[0.65, 0.25, 0.1]],
+    "market_share_labels": ["competitor", "own", "new"],
+    "random_seed": rng,
+}
+
+data = generate_saturated_data(**scenario)
+
+# Build a multivariate interrupted time series model
+model = MVITS(
+    existing_sales=["competitor", "own"],
+    saturated_market=True,
+)
+
+model.inform_default_prior(
+    data=data.loc[: scenario1["treatment_time"], ["competitor", "own"]]
+)
+
+# Parameter estimation
+model.sample(data[["competitor", "own"]], data["new"])
+
+# Visualize the results
+model.plot_fit();
+```
+
+See the {ref}`howto` section for more information about using the customer choice module.