Changelog

Changelog

[0.0.1] - 2024-05-16

• First release on PyPI.

[0.0.2] - 2024-05-16

• Fixed installation issues.

[0.0.3] - 2024-05-16

• Working release on PyPI.

[0.0.4] - 2024-06-10

• New repository.
• Tests.

[0.0.5] - 2024-06-10

• Readme.
• GitHub Actions.

[0.0.6] - 2024-06-10

• Fix Codecov.

[0.0.7] - 2024-06-11

• Fix Codecov for GitHub Actions.

[0.0.8] - 2024-06-11

• Publish documentation.

[0.0.9] - 2024-06-11

• Publish documentation - next try.

[0.0.10] - 2024-06-11

• Publish documentation - next try.

[0.0.11] - 2024-06-11

• Publish documentation - next try.

[0.0.12] - 2024-06-11

• Publish documentation - next try.

[0.0.13] - 2024-06-11

• Publish documentation - next try.

[0.0.14] - 2024-06-11

• Publish documentation - next try.

[0.0.15] - 2024-06-12

• Publish documentation - next try.

[0.0.16] - 2024-06-12

• Publish documentation - next try.

[0.0.17] - 2024-06-12

• Publish documentation - next try.

[0.0.18] - 2024-06-12

• Publish documentation - next try.

[0.0.19] - 2024-06-12

• Publish documentation - next try.

[0.0.20] - 2024-06-12

• Publish documentation - next try.

[0.0.21] - 2024-06-12

• Numpy style documentation.

[0.0.22] - 2024-06-12

• Import rules.

[0.0.23] - 2024-07-12

• GPU Acceleration.

[0.0.24] - 2024-07-15

• cuDF improved import.

[0.0.25] - 2024-07-17

• Clear GPU memory pool.

[0.0.26] - 2024-07-17

• Eliminate zeros for sparse matrix.

[0.0.27] - 2024-07-19

• Readme improved.

[0.0.28] - 2024-07-23

• Predict function.

[0.0.29] - 2024-07-23

• Improved pretty notation.

[1.0.0] - 2024-07-27

• Uplift fixed.
• Docstrings and notations.
• Completely finished.

[1.0.1] - 2024-07-31

• Uplift fixed.

[1.0.2] - 2024-07-31

• Profiling.

[1.0.3] - 2024-07-31

• Comparison with another package.

[1.0.4] - 2024-08-04

• Fix predict method for cupy

[1.0.5] - 2024-08-05

• Fix empty list of stable attributes

[1.0.6] - 2024-08-06

• Feature: New fit_onehot

[1.0.7] - 2024-08-06

• Fix: Onehot data copy

[1.0.8] - 2025-01-01

• Feature: Dominant action rules

[1.0.9] - 2025-02-23

• Feature: High-Utility Action Rules Mining

[1.0.10] - 2025-03-01

• Feature: Measures - support and confidence

[1.0.11] - 2025-03-04

• Fix: Utility

[1.1.0] - 2026-05-27

• Feature: Confidence intervals for action rules via ActionRules.confidence_intervals() with three engines — bootstrap (percentile, BCa), analytic (Wald, Newcombe-Wilson, auto), and Bayesian (Beta-Binomial Monte Carlo).
• Feature: Rule categorisation (accept / reject / uncertain) based on a user-supplied decision threshold over uplift or realistic_rule_gain.
• Feature: Cross-validation via ActionRules.cross_validate() with stratified folds, configurable selection strategies, and out-of-fold metric estimation.
• Feature: New action_rules.visualization module with forest plots, bootstrap distributions, coverage calibration, and CI-width diagnostics (matplotlib lazy-imported).
• Feature: CI results are surfaced in Output.get_ar_notation(), Output.get_pretty_ar_notation(), and JSON export (get_export_notation()); NaN/Inf are serialised as null.
• Feature: New CLI options for confidence intervals (--ci-method, --ci-level, --ci-threshold, --ci-metric, …).
• Fix: df_to_array no longer creates spurious <attr>_<item_*>_nan one-hot columns for missing antecedent values; NaN is now preserved through get_dummies per the pessimistic null-value semantics of Dardzinska (2013, §2.3.2). Target column behaviour unchanged.
• Docs: Notebooks for the Telco churn end-to-end tour, rule-level CIs across three datasets (Bank Marketing, Credit Card Default, Employee Attrition), and the inference studies / article figures.

[2.0.0] - 2026-06-06

• Feature: Packed-bitset rewrite of the Action-Apriori mining core. CPU and GPU CUDA path.
• Compatibility: Public API preserved so existing 1.x call sites keep working unchanged.