0.17.0

What's Changed

Major Changes 🛠

• Add CLV Cohort ShiftedBetaGeoModel by @ColtAllen in #2005

New Features 🎉

• Add BuildModelFromDAG class for causal components by @cetagostini in #1950
• Introduce experimental TBFPC causal discovery algorithms by @cetagostini in #1947
• Add fivetran pandas accessor by @williambdean in #1924
• Add Special priors by @daniel-saunders-phil in #1939
• Add ModelSamplerEstimator for PyMC Marketing model benchmarking by @cetagostini in #1943
• Adds Optional dims arg for subsetting by dimensions when plotting a multidimensional model by @timbo112711 in #1955
• Add calibration support to YAML MMM builder by @cetagostini in #1979
• Add expected_residual_lifetime predictions to ShiftedBetaGeoModel by @ColtAllen in #2010
• Add expected_retention_elasticity to ShiftedBetaGeoModel by @ColtAllen in #2028
• Add GPU support in BudgetOptimizer by @PabloRoque in #1937

Bugfixes 🐛

• BUG [Issue with fourier] by @cetagostini-wise in #1936
• Fix Mypy - HotFix by @juanitorduz in #1963
• Fix np.testing.assert_array_almost_equal argument names by @cetagostini in #1949
• Hotfix - Fix dimension assertion for adding multidimensional event effects to multidimensional MMM by @TeemuSailynoja in #1969
• Hotfix: use target_column in create_fit_data and build_from_idata by @PabloRoque in #1918
• HOT FIX: Update image for sBG Model in gallery by @juanitorduz in #2008
• Solving issue with wrapper detection in optimizer by @cetagostini in #2021
• Fixing fixed_beta bug adding new test for nested_logit by @NathanielF in #2015

Documentation 📖

• Notebook: How to model upper funnel with causal reasoning and PyMC by @cetagostini in #1971
• Notebook: Add multi-objective optimization notebook and update optimizer by @cetagostini in #1973
• correct typo in multidimensional nb by @daniel-saunders-phil in #1919
• Add approximate_fit method for variational inference by @cetagostini in #1923
• Add lognormalprior to the multidimensional notebook by @daniel-saunders-phil in #1961
• Adding calibration by potential based on CPT (Additional calibration style) by @cetagostini in #1959
• Setting infraestructure for Translations (ES site) by @OriolAbril in #1906
• Add MMM Quickstart Guide by @juanitorduz in #1972
• Update notebook risk allocation by @cetagostini in #1982
• Split notebooks test run mmm vs others by @juanitorduz in #1986
• Adding additional translations to the spanish site by @cetagostini in #1984
• Translate to spanish certain notebooks! by @cetagostini in #1995
• Notebook: PyMC-Marketing GAM capabilities by @cetagostini in #1985
• Translate guide folder to spanish by @cetagostini in #1997
• Adding spanish translations to API docs by @cetagostini in #1999
• Finishing all translations to API in Castellano by @cetagostini in #2000
• Notebooks translations to ES by @cetagostini in #2001
• Small error on render notebook tips/warnings/notes by @cetagostini in #2003
• MMM Class Claude Sonnet 4.5 Refactor by @juanitorduz in #1974
• Solving issue in #2006 doing the optimization more obvious not close to the limit. by @cetagostini in #2007
• Updated How We Compare Section by @lfiaschi in #2012
• Notebook: Causal discovery with PyMC-Marketing by @cetagostini in #1994
• Improve documentation links by @juanitorduz in #2025

Maintenance 🔧

• Adjusting calibration by costs to work with sample posterior predictive by @cetagostini in #1988
• Fix ModelBuilder docstring formatting by @williambdean in #1926
• Keep zeroes in add_noise_to_channel_allocation by @PabloRoque in #1916
• Update gamma_control prior dimensions in MMM model by @cetagostini in #1920
• Add model merging utilities and tests for multi-model optimization by @cetagostini in #1944
• Rename LogNormalPositiveParam to LogNormalPrior by @cetagostini in #1948
• Correcting optimizer docstring by @cetagostini in #1962
• Add MaskedPrior class for masked variable creation by @cetagostini in #1945
• Clarify mean_tightness_score docstring and formula by @cetagostini in #1980
• don't use paid defaults channel for dev env by @OriolAbril in #1987
• Translations in the API documents by @cetagostini in #1998
• Last translations by @cetagostini in #2002
• Translations for the new recent merge PRs by @cetagostini in #2020
• Translations by @cetagostini in #2029
• Bump version to 0.17.0 by @juanitorduz in #2030
• Update Spanish translations for CLV and MMM API docs by @cetagostini in #2031

Full Changelog: 0.16.0...0.17.0

0.16.0

What's Changed

Major Changes 🛠

• Handle release of pymc 5.25.0 by @daniel-saunders-phil in #1844
• Migrate to prior module from pymc-extras by @williambdean in #1767

New Features 🎉

• Sensitivity analysis and marginal effects by @drbenvincent in #1673
• Add table method to ModelBuilder to display rich table by @williambdean in #1786
• Allow for indexing of apply parameters by @williambdean in #1847
• index transformation variables in higher dimension cases by @williambdean in #1857
• allow for arbitrary additive effects in MVITS model by @williambdean in #1861
• Allow Lists and NumPy Arrays as Transformation Priors by @timbo112711 in #1879
• Support HSGP instance for time_varying_media in MMM by @cetagostini in #1881
• Support HSGP instance for time_varying_intercept in MMM by @cetagostini in #1894
• Add Fivetran data loaders by @cetagostini in #1887
• Add more Basis implementations for use with EventEffect by @TeemuSailynoja in #1911

Bugfixes 🐛

• hot fix (test match) by @juanitorduz in #1852
• Update and reduce size multi_mmm.nc by @juanitorduz in #1854
• Deprecate and disable budget optimization method & Updating notebook by @cetagostini in #1832
• Handle 0-dim channel_xr in create_zero_dataset by @cetagostini in #1890
• Fix grid shape logic in saturation_curves and add tests by @cetagostini in #1889
• Updating prior predictive notebook by @cetagostini in #1897
• Streamlit Dependency Fix by @timbo112711 in #1904
• Fix FourierEffect by @PabloRoque in #1909

Documentation 📖

• fix the link to Labs YouTube by @williambdean in #1822
• Fix notebooks titles MMM in gallery by @juanitorduz in #1843
• Remove reference to linear_saturation function by @williambdean in #1833
• Updating MMM Budget allocation examples, functionalities and dependencies by @cetagostini in #1849
• Add rfm_train_test_split example to CLV Quickstart by @ColtAllen in #1855
• lift test requires 'date' in df_lift_test Data Frame with time_varying_media=True by @timbo112711 in #1818
• MMM: Fix Scaling Intercept by @juanitorduz in #1845
• Implement multidimensional scaling approach in regular MMM by @cetagostini in #1862
• Update MMM comparison table by @williambdean in #1878
• Get marginal_effects hdi estimates by @PabloRoque in #1885
• Correcting typo around yml files and csv files. by @cetagostini in #1888
• Improve saturation function docs by @daniel-saunders-phil in #1892
• Improvements on multidimensional notebook by @daniel-saunders-phil in #1876
• Clean badges by @juanitorduz in #1896
• Fourier transformation and HSGPs for streamlit app by @timbo112711 in #1898
• Add usage examples and documentation to MMM modules by @cetagostini in #1895
• Assign coords to multidimensional fit_data by @PabloRoque in #1902

Maintenance 🔧

• Add callback tracking to budget optimizer by @cetagostini in #1829
• Cosmetic changes to #1829 by @PabloRoque in #1830
• Add support for time-based budget distribution in optimizer by @cetagostini in #1841
• fix(clv): Improve validation for missing columns in CLVModel by @aaron1-z in #1851
• Adding arguments to model builder by @cetagostini in #1867
• Add causal graph support to MMM with DAG serialization by @cetagostini in #1882
• Refactor ModelBuilder into smaller classes by @ColtAllen in #1870
• Update requirements.txt [StreamlitApp] by @juanitorduz in #1900
• fix mypy overide by @juanitorduz in #1901
• Updates Streamlit App Requirements by @timbo112711 in #1903
• Multidimensional support in plot_sensitivity_analysis by @PabloRoque in #1886
• Minor convenience to override model_kwargs in build_mmm_from_yaml by @PabloRoque in #1907
• Normalize gaussian basis by @PabloRoque in #1912

New Contributors

• @daniel-saunders-phil made their first contribution in #1844
• @aaron1-z made their first contribution in #1851

Full Changelog: 0.15.1...0.16.0

0.15.1

What's Changed

New Features 🎉

• Handle static covariates in ModifiedBetaGeoModel by @PabloRoque in #1815

Bugfixes 🐛

• Handle optimization for dimensions having channel with zero spend by @PabloRoque in #1762

Documentation 📖

• adding DCM to readme by @NathanielF in #1800
• stale docstring: "likelihood" is not a model variable by @timbo112711 in #1803

Maintenance 🔧

• Compute deterministics after sampling by @PabloRoque in #1802

New Contributors

• @timbo112711 made their first contribution in #1803

Full Changelog: 0.15.0...0.15.1

0.15.0

What's Changed

New Features 🎉

• Adding Multinomial and Nested Logit Models for Consumer Choice by @NathanielF in #1654
• URL pathlike for consistent pymc_marketing.paths.data_dir API by @williambdean in #1797
• Implement add_lift_test_measurements for multidimensional MMM by @TeemuSailynoja in #1738

Bugfixes 🐛

• Automatic casting based on model types for posterior predictive by @cetagostini in #1781
• Clip noisy channel predictors by @PabloRoque in #1790

Documentation 📖

• ruff format docstring code blocks by @williambdean in #1794

New Contributors

• @NathanielF made their first contribution in #1654

Full Changelog: 0.14.1...0.15.0

0.14.1

What's Changed

New Features 🎉

• Log off traceback in MLflow by @williambdean in #1736
• multidimensional MMM initialization validation by @IlanDanial in #1725
• More granular ModelBuilder errors upon load by @williambdean in #1765

Bugfixes 🐛

• Fix LinearTrend changepoints in plot_curve method by @gdeleva in #1739
• Fix dim issue for default priors of linear trend and replace linear trend related doc pictures by @gdeleva in #1741
• LinearTrend deterministic error by @williambdean in #1748
• Bug: No need to check columns in Y for yml construction by @cetagostini in #1774
• Bug: Solving issue with hardcoded date by @cetagostini in #1780
• Bug: Sample posterior predictive raise error if dates overlap by @cetagostini in #1778
• Allow for tvp priors in multidimensional.MMM by @williambdean in #1785

Documentation 📖

• Enhances MMM Packages Comparison by @lfiaschi in #1722
• Fix time plots in CLV Quickstart by @ColtAllen in #1724
• Fix BudgetOptimizer docstrings by @TeemuSailynoja in #1705
• Updating example multidimensional to use original scale after posterior predictive by @cetagostini in #1782
• Standardize usage of singular | plural model variables by @PabloRoque in #1763

Maintenance 🔧

• upper bound pymc 5.23 by @juanitorduz in #1728
• update pytensor lower bound to 2.31.3 by @juanitorduz in #1737

New Contributors

• @lfiaschi made their first contribution in #1722
• @IlanDanial made their first contribution in #1725

Full Changelog: 0.14.0...0.14.1

0.14.0

What's Changed

Deprecations 🚨

• Deprecate CLV methods, clv_summary, method parameter, to_json and from_json methods by @ErikRingen in #1637

New Features 🎉

• Add Bass Diffusion Model by @williambdean in #1328
• Support for string in plot_curve and add to docs by @williambdean in #1650
• Expose n_samples, hdi_probs, and random_seed in plot_curve by @williambdean in #1697
• Add LinearRegression wrapper by @williambdean in #1713

Bugfixes 🐛

• support for additional dimension in LinearTrend by @williambdean in #1688
• Fix Bass model derivative by @williambdean in #1702

Documentation 📖

• Change spacing batched_convolution documentation by @williambdean in #1645
• Add utils to docs by @williambdean in #1651
• Fix Bass model example plot by @williambdean in #1666
• Create wrapper to make budget optimizer compatible with multidimensional class by @cetagostini in #1652
• Change documentation of multi-dimensional by @gdeleva in #1679
• Set html title explicitly by @OriolAbril in #1686
• Build multidimensional MMM from yaml configuration by @cetagostini in #1669
• Accelerate documentation build by @OriolAbril in #1695
• Use HDI instead of CI in MMM plotting suite by @TeemuSailynoja in #1689
• Consolidate name to PyMC-Marketing by @williambdean in #1699
• Rename all occurences of pymc-labs.io to pymc-labs.com by @twiecki in #1708

Maintenance 🔧

• Change the event function to class by @williambdean in #1675
• Issue #1696: addresses a bug in plotting the intercept by @esmailansari in #1698

New Contributors

• @ErikRingen made their first contribution in #1637
• @gdeleva made their first contribution in #1679
• @TeemuSailynoja made their first contribution in #1689
• @twiecki made their first contribution in #1708
• @esmailansari made their first contribution in #1698

Full Changelog: 0.13.1...0.14.0

0.13.1

What's Changed

Bugfixes 🐛

• Fix casting bug when x0 is not supplied to optimizer by @cluhmann in #1636

Full Changelog: 0.13.0...0.13.1

0.13.0

What's Changed

New Features 🎉

• Add Example Gallery by @juanitorduz in #1611
• Fourier and LinearTrend additive effects by @williambdean in #1580
• Add version to posterior and prior groups by @williambdean in #1591

API

• Change baseline_intercept to intercept_baseline for consistent prefix by @ango94 in #1529

Bugfixes 🐛

• Fix LinearTrend out of sample by @williambdean in #1585
• BUG: FourierEffect with multidimensional case by @williambdean in #1594

Documentation 📖

• Fix minor scaling inconsistency in hdi50 prior plot by @PabloRoque in #1579
• Allow original scales in saturation_curves_scatter by @PabloRoque in #1581
• MMM Multidimensional example linear trend by @juanitorduz in #1589
• minor improvements by @juanitorduz in #1590
• rerun mmm tvp base nb by @juanitorduz in #1595
• Add to MLflow error docs by @williambdean in #1616
• format code blocks by @williambdean in #1621
• Update "how do we compare" by @juanitorduz in #1632

Maintenance 🔧

• Cast periods value to int by @williambdean in #1597
• restrict the osqp version by @williambdean in #1614
• Permit BudgetOptimizer.allocate_budget() to take x0 as an argument by @cluhmann in #1565
• Replace batched_convolution by PyTensor native implementation by @ricardoV94 in #1583
• Log pm.sample exception to MLflow by @williambdean in #1612

New Contributors

• @ango94 made their first contribution in #1529

Full Changelog: 0.12.1...0.13.0

0.12.1

What's Changed

New Features 🎉

• Add sample_prior to sample from arbitrary VariableFactory by @williambdean in #1536
• 1553 add transform parameter for HSGP link function by @williambdean in #1555
• Add demeaned_basis flag for HSGP classes by @williambdean in #1560
• Specify target scaling dimensions by @williambdean in #1562
• Allow for media scaling settings by @williambdean in #1573

Bugfixes 🐛

• 1570 target scaled and variable need to be broadcasted with non global scaling by @williambdean in #1571

Documentation 📖

• MMM Multi-Dimensional Example by @juanitorduz in #1521

Maintenance 🔧

• Correcting small errors in multidimensional mmm by @cetagostini in #1551
• remove abstract requirement of data_setter by @williambdean in #1463

Full Changelog: 0.12.0...0.12.1

0.12.0

What's Changed

Major Changes 🛠

• excluded _prior suffix from CLV models by @arthurmello in #1498

Deprecations 🚨

• Rename CLV fit_method and t_unobserved parameters to method and t_start_eval by @ColtAllen in #1537

New Features 🎉

• [MMM] Model events as gaussian bumps by @juanitorduz in #1465
• Out-of-Box MultiDimensional MMM by @cetagostini in #1036
• Add basis bumps effect to MMM model by @wd60622 in #1475
• Allow plot_expected_purchases_pcc in BetaGeoModel and ModifiedBetaGeoModel by @PabloRoque in #1470
• Allow static covariates in BGNBDModel by @PabloRoque in #1390
• Support MediaTransformation and MediaConfigList deserialization by @wd60622 in #1504
• Check for wrong media dims by @wd60622 in #1505

Bugfixes 🐛

• load bug with y name different than output_var by @wd60622 in #1468
• Fix ModifiedBetaGeoNBDRV sim_data by @PabloRoque in #1472
• Change Basis sample_curve defaults by @wd60622 in #1500
• Check for incompatible dims in MediaTranform by @wd60622 in #1502
• Allow for extra columns in df_events by @wd60622 in #1501
• Use MediaTransformation dims in save and load by @wd60622 in #1503
• Fix _create_synth_dataset when MMM has no controls by @PabloRoque in #1513
• Use SoftPlusHSGP in multidimensional HSGP by @wd60622 in #1510
• Use SoftPlusHSGP directly in MMM class by @wd60622 in #1523
• Fix SoftPlusHSGP out of sample by @wd60622 in #1544

Documentation 📖

• Add thin wrapper around advi functionality by @PabloRoque in #1365
• Added missing paramater take_every in docs of create_log_callback by @nayan2167 in #1491
• Fixes documentation on mmm control variable scaling - Issue #1469 by @zach-brown-18 in #1499
• change parameter from samples to draws by @wd60622 in #1524
• Fix the TVP example by @wd60622 in #1533

Maintenance 🔧

• add preliz dependency by @wd60622 in #1507
• Merge value and grad in budget optimizer by @ricardoV94 in #1512

New Contributors

• @zach-brown-18 made their first contribution in #1499
• @github-actions made their first contribution in #1528

Full Changelog: 0.11.1...0.12.0