v0.4.0

July 2023

Following this release the deprecation policy remains suspended. Future releases may have breaking changes, so it may be wise to set an upper bound on the package version.

Highlights

• Data downloading functions for popular classification, regression and forecasting archives are now available

• Implementations for the RSTSF interval-based and the RDST shapelet-based classifiers have been added to the classification module

• Time series adaptations for the PAM, CLARA, CLARANS clustering algorithms have been added to the clustering module

• The interval-based forests in the classification module have been reworked to use a new base class, including speed-ups and also allowing the implementation of regression versions (RISERegressor, CIFRegressor and DrCIFRegressor)

• ResNet and FCN deep learning regressors are now available

• Large portions of the website documentation have been overhauled, including the examples page and introduction notebooks for data types and data loading

Annotation

Documentation

• [DOC] Restructure annotation and transformation notebooks (#531) @TonyBagnall

Benchmarking

Enhancements

• [ENH] Result loader from tsc.com (#452) @TonyBagnall

• [ENH] Data downloaders for classification/regression and forecasting (#478) @TonyBagnall

• [ENH] Full classifier results (#571) @TonyBagnall

Classification

Bug Fixes

• [BUG] HIVECOTEV1 cross-validation bugfix (#518) @MatthewMiddlehurst

• [BUG] RotationForest bugfix and doc improvements (#517) @MatthewMiddlehurst

• [ENH] Updates to Catch22 and fixes to collection converters (#509) @MatthewMiddlehurst

• [BUG] Add non-deterministic tag to RDSTClassifier (#577) @MatthewMiddlehurst

Documentation

• [BUG] RotationForest bugfix and doc improvements (#517) @MatthewMiddlehurst

• [DOC] Add correct images for classification (#514) @TonyBagnall

Enhancements

• [ENH] Updates to Catch22 and fixes to collection converters (#509) @MatthewMiddlehurst

• [ENH] Update interval extraction transformers and pipelines (#511) @MatthewMiddlehurst

• [ENH] Implement RSTSF classifier (#544) @MatthewMiddlehurst

• [ENH] Add deep networks/classifiers/regressors to init (#554) @TonyBagnall

• [ENH] Implement configurable interval base class with classifier and regressor (#545) @MatthewMiddlehurst

• [ENH] RDSTClassifier and transformer tests (#529) @baraline

• [ENH] Replace TSF estimators with base class version and remove series_as_features package (#575) @MatthewMiddlehurst

• [ENH] Full classifier results (#571) @TonyBagnall

• [ENH] Move CIF and DrCIF to use the interval base class (#573) @MatthewMiddlehurst

• [ENH] Updates the RISE classifier to use new interval base class (#584) @MatthewMiddlehurst

• [ENH] Updates the STSF classifier to use new interval base class (#585) @MatthewMiddlehurst

Maintenance

• [MNT] Avoid CI fail on deep test by generating at least 2 classes in random data (#485) @hadifawaz1999

Refactored

• [REF] change single class use case (#497) @TonyBagnall

Clustering

Bug Fixes

• [BUG] DBA numba error (#541) @chrisholder

Documentation

• [DOC] clustering and regression notebooks (#516) @TonyBagnall

Enhancements

• [ENH] Kmedoids refactor and speed up (#498) @chrisholder

• [ENH] Pam kmedoids clusterer (#500) @chrisholder

• [ENH] CLARA clusterer (#501) @chrisholder

• [ENH] CLARANS clusterer (#503) @chrisholder

Refactored

• [ENH] Kmedoids refactor and speed up (#498) @chrisholder

• [ENH] Pam kmedoids clusterer (#500) @chrisholder

Distances

Bug Fixes

• [BUG] RDST bugfixes (#483) @baraline

• [BUG] DBA numba error (#541) @chrisholder

Documentation

• [DOC] Update distances notebook (#491) @TonyBagnall

• [DOC] Tidy up the datasets and distances examples (#513) @TonyBagnall

• [DOCS] Add docs for function transformer with distances, and tests for distances with sklearn (#581) @TonyBagnall

• [DOC] Distance function template (#583) @TonyBagnall

Forecasting

Documentation

• [DOC] Move forecasting notebooks (#536) @TonyBagnall

Maintenance

• [MNT] BATS parameter test (#568) @TonyBagnall

• [MNT] Add statsforecast back to all_extras (#552) @MatthewMiddlehurst

Regression

Bug Fixes

• [BUG] RotationForest bugfix and doc improvements (#517) @MatthewMiddlehurst

• [ENH] Updates to Catch22 and fixes to collection converters (#509) @MatthewMiddlehurst

• [BUG] Regression base class score function (#549) @TonyBagnall

Documentation

• [BUG] RotationForest bugfix and doc improvements (#517) @MatthewMiddlehurst

• [DOC] clustering and regression notebooks (#516) @TonyBagnall

Enhancements

• [ENH] Updates to Catch22 and fixes to collection converters (#509) @MatthewMiddlehurst

• [ENH] Update interval extraction transformers and pipelines (#511) @MatthewMiddlehurst

• [ENH] Add deep networks/classifiers/regressors to init (#554) @TonyBagnall

• [ENH] Add ResNet deep learning regressor (#558) @hadifawaz1999

• [ENH] Add FCN deep learning regressor (#560) @hadifawaz1999

• [ENH] Implement configurable interval base class with classifier and regressor (#545) @MatthewMiddlehurst

• [MNT-ENH] Fixed stuff in CNN deep learning regressor (#559) @hadifawaz1999

• [ENH] Replace TSF estimators with base class version and remove series_as_features package (#575) @MatthewMiddlehurst

• [ENH] Move CIF and DrCIF to use the interval base class (#573) @MatthewMiddlehurst

• [ENH] Updates the RISE classifier to use new interval base class (#584) @MatthewMiddlehurst

Maintenance

• [MNT] Avoid CI fail on deep test by generating at least 2 classes in random data (#485) @hadifawaz1999

• [MNT-ENH] Fixed stuff in CNN deep learning regressor (#559) @hadifawaz1999

Other

• [ENH] Removes the delegate structure from regression (#580) @TonyBagnall

Transformations

Bug Fixes

• [BUG] RDST bugfixes (#483) @baraline

• [BUG] reduce test size for MultiRocket (#520) @TonyBagnall

• [ENH] Updates to Catch22 and fixes to collection converters (#509) @MatthewMiddlehurst

Documentation

• [DOC] Restructure annotation and transformation notebooks (#531) @TonyBagnall

Enhancements

• [ENH] Collection transformer wrapper and basic base class testing (#467) @MatthewMiddlehurst

• [ENH] Updates to Catch22 and fixes to collection converters (#509) @MatthewMiddlehurst

• [ENH] Add some simple collection transformers (#508) @MatthewMiddlehurst

• [ENH] Update interval extraction transformers and pipelines (#511) @MatthewMiddlehurst

• [ENH] RDSTClassifier and transformer tests (#529) @baraline

Maintenance

• [MNT] Add test for transformer numpy input (#555) @MatthewMiddlehurst

Refactored

• [ENH] Refactoring PAA (#489) @hadifawaz1999

Other

Bug Fixes

• [MNT] tidy up resnet and fix failures (#486) @hadifawaz1999

• [BUG] Update gunpoint arff (#496) @TonyBagnall

• [BUG] Doctest formatting bug 546 (#548) @TonyBagnall

• [BUG] Scenarios string fix (#594) @TonyBagnall

Documentation

• [DOC] Change README logo link (#482) @MatthewMiddlehurst

• [DOC] InceptionTime docstrings (#494) @TonyBagnall

• [DOCS] Update the API (#495) @TonyBagnall

• [DOC] Add transformations to getting started page and add opengraph sphinx package (#522) @MatthewMiddlehurst

• [DOC] Added Deep Learning notebook and images (#493) @hadifawaz1999

• [DOC] Remove the tutorials and users webpages (#525) @MatthewMiddlehurst

• [DOC] Remove stray user_guide dir and _mission webpage (#526) @MatthewMiddlehurst

• [DOC] fixed path of image (#535) @hadifawaz1999

• [BUG] Doctest formatting bug 546 (#548) @TonyBagnall

• [DOC] Rewrite current user installation guide and move dev installation guide to its own file (#524) @MatthewMiddlehurst

• [DOC] Remove mentoring and estimator overview webpages (#527) @MatthewMiddlehurst

• [DOC] Auto generate about page teams (#532) @MatthewMiddlehurst

• [DOC] Utils notebook directory (#561) @TonyBagnall

• [DOCS] attempt to minimise docstring warnings (#572) @TonyBagnall

• [DOC] Datatypes docs (#538) @TonyBagnall

• [DOC] purge AA_datatypes_and_datasets.ipynb (#540) @TonyBagnall

• [DOC] Tidy up the docs (#591) @TonyBagnall

• [DOC] Full examples debug (#598) @TonyBagnall

Enhancements

• [ENH] Add function to make example 3D numpy problems (#510) @MatthewMiddlehurst

• [ENH] Tidy up test exclusions (#473) @TonyBagnall

Maintenance

• [MNT] Loosen upper bound for esig and update Python version for notebook tests. (#487) @MatthewMiddlehurst

• [MNT] Revert holidays and dash dependency changes (#445) @MatthewMiddlehurst

• [MNT] Add non-deterministic and cant-pickle tag for estimators (#371) @hadifawaz1999

• [MNT] Replace handles-missing-data tag with capability:missing_values (#505) @MatthewMiddlehurst

• [MNT] Set upper bound for tslearn to fix CI (#534) @MatthewMiddlehurst

• [MNT] Change esig dependency version requirements to fix CI (#539) @MatthewMiddlehurst

• [MNT] Setting more upper bounds to fix CI (#542) @MatthewMiddlehurst

• [MNT] Exclude results loading notebook (#562) @MatthewMiddlehurst

• [MNT] estimator checks deprecation (#556) @TonyBagnall

• [MNT] Add periodic test workflow and remove Makefile (#553) @MatthewMiddlehurst

• [MNT] Update codeowners (#547) @TonyBagnall

• [MNT] DL OOM test failure fix and remove Makefile (#593) @MatthewMiddlehurst

Contributors

The following have contributed to this release through a collective 79 GitHub Pull Requests:

@baraline, @chrisholder, @hadifawaz1999, @MatthewMiddlehurst, @TonyBagnall