v0.8.1

April 2024

The following modules are still considered experimental, and the deprecation policy does not apply:

annotation, anomaly_detection, benchmarking, segmentation, similarity_search, testing, transformations/series, visualisation

Highlights

• Mostly bug fixes, documentation improvements and new deprecations

• Subgradient barycentre averaging is available for clustering (@chrisholder)

Anomaly Detection

Enhancements

• [ENH] BaseSeriesEstimator rework (#1394) @MatthewMiddlehurst

Classification

Documentation

• [DOC] Correct WeightedEnsembleClassifier parameter docstring (#1418) @wayneadams

Enhancements

• [ENH] fit_predict for FreshPRINCE and RotationForest (#1456) @MatthewMiddlehurst

Other

• [BUG] Fix RDST doctest failure & numba deprecration warning (#1445) @baraline

Clustering

Enhancements

• [ENH] Subgradient barycentre averaging (#1339) @chrisholder

• [ENH] Kmeans allow empty clusters (#1400) @chrisholder

Maintenance

• [MNT] split up and simplified ba test (#1432) @chrisholder

Datasets

Bug Fixes

• [BUG] fix loader (#1428) @TonyBagnall

Documentation

• [DOC] PR #1452 (#1462) @RishavKumarSinha

Distances

Enhancements

• [ENH] Subgradient barycentre averaging (#1339) @chrisholder

• [ENH] Distance module pairwise unequal length support (#1356) @chrisholder

Forecasting

Deprecation

• [ENH] replace convert/convert_to with convert_series (#1338) @TonyBagnall

• [DEP] deprecate check_scorer (#1439) @TonyBagnall

• [DEP] deprecate probabilistic performance metric classes (#1436) @TonyBagnall

Enhancements

• [ENH] replace convert/convert_to with convert_series (#1338) @TonyBagnall

• [ENH] remove conversion in get_cutoff (#1342) @TonyBagnall

• [ENH] remove convert_to and vectorized import from index_functions (#1433) @TonyBagnall

Other

• [COV] Tests for testing/utils/data_gen (#1347) @TonyBagnall

Regression

Enhancements

• [ENH] fit_predict for FreshPRINCE and RotationForest (#1456) @MatthewMiddlehurst

Segmentation

Enhancements

• [ENH] BaseSeriesEstimator rework (#1394) @MatthewMiddlehurst

Similarity Search

Enhancements

• [ENH] BaseSeriesEstimator rework (#1394) @MatthewMiddlehurst

Transformations

Bug Fixes

• [BUG] Fix MultiRocketMultivariate failures (#1453) @baraline

Deprecation

• [DEP] deprecate check_scorer (#1439) @TonyBagnall

Documentation

• [DOC] PR #1452 (#1462) @RishavKumarSinha

Enhancements

• [ENH] BaseSeriesTransformer maintains axis for user determined output with BaseSeriesEstimator function (#1372) @TonyBagnall

• [ENH,REF] Refactor univariate-only tag (#1363) @itsdivya1309

• [ENH] BaseSeriesEstimator rework (#1394) @MatthewMiddlehurst

• [ENH] remove datatypes call from transformations test function (#1434) @TonyBagnall

• [ENH] Add support for array-like and statistical inputs as arguments in fill_value of Padding Transformer (#1373) @aadya940

• [ENH] Autocorrelation SeriesTransformer (#1380) @TonyBagnall

Refactored

• [ENH,REF] Refactor univariate-only tag (#1363) @itsdivya1309

Other

• [BUG] Fix RDST doctest failure & numba deprecration warning (#1445) @baraline

Unit Testing

Maintenance

• [MNT] Add Classifiers For Result Comparison (#1379) @harshithasudhakar

Other

• [COV] Tests for testing/utils/data_gen (#1347) @TonyBagnall

Other

Bug Fixes

• [BUG] allow np.ndarray with ndim>1 to convert to pd.series when squeezable to 1D (#1427) @TonyBagnall

Deprecation

• [DEP] Make VectorizedDF private (#1369) @MatthewMiddlehurst

Documentation

• [DOCS] Update the mentoring projects (#1366) @TonyBagnall

• [DOC] Fixes mis-wording/typo in mentoring docs (#1396) @wayneadams

• [DOC] Adds previously implemented classifiers to docs/api_references (#1401) @wayneadams

• [DOC] About Us webpage update (#1419) @MatthewMiddlehurst

• [DOC] Add estimator overview table with capabilities (#1426) @baraline

• [DOC] Remove optional from docstrings (#1458) @griegner

Enhancements

• [DEP] Deprecate _serialize file and load function (#1378) @TonyBagnall

• [ENH] remove datatypes call from test_interval_wrappers (#1435) @TonyBagnall

Maintenance

• [MNT] Notebook test exclusion tag and stale branch removal (#1316) @MatthewMiddlehurst

• [MNT] Update branch clean action (#1422) @MatthewMiddlehurst

• [MNT] dask CI fix, ReadTheDocs fix and expanded full pytest actions tag (#1420) @MatthewMiddlehurst

• [MNT] Update cache restore action (#1431) @aadya940

Refactored

• [REF] Move tag deprecation to registry (#1460) @MatthewMiddlehurst

Contributors

The following have contributed to this release through a collective 38 GitHub Pull Requests:

@aadya940, @baraline, @chrisholder, @griegner, @harshithasudhakar, @itsdivya1309, @MatthewMiddlehurst, @RishavKumarSinha, @TonyBagnall, @wayneadams

v0.8.0

April 2024

Highlights

• Deprecations from 0.7.0 have been removed, see the package deprecation sections for more details

• typing-extensions has been added as a core dependency

• The following deep learners are now available for regression: IndividualLITERegressor, LITETimeRegressor and EncoderRegressor ((@aadya940) & @AnonymousCodes911)

• The HydraRegressor and MultiRocketHydraRegressor algorithms have been implemented for regression module (@MatthewMiddlehurst)

• A wrapper for the tslearn LearningShapelets classifier has been added (@itsdivya1309)

• Support for unequal length in pairwise distance calculation for the SBD and MSM distances is now available, this will be expanded to other distances in time (@CodeLionX)

Benchmarking

Documentation

• [DOC] Fix benchmarking notebooks (#1305) @TonyBagnall

Maintenance

• [MNT] xfail website reads that return a connection error (#1357) @TonyBagnall

Classification

Bug Fixes

• [BUG] Fix random state for deep learning models in classification/regression and clustering (#1271) @hadifawaz1999

• [BUG] ElasticEnsemble with euclidean and twe distance measures (#1288) @itsdivya1309

• [BUG] fixed sqr error (#1240) @AnonymousCodes911

• [BUG] LearningShapeletClassifier fixes (#1370) @MatthewMiddlehurst

Deprecation

• [DEP] Remove deprecated overloaded multiplication and sklearnpipeline (#1265) @TonyBagnall

• [DEP] remove deprecated the save_train_predictions parameter in BOSSEnsemble, CBOSS, T… (#1333) @TonyBagnall

• [DEP] remove support probabilities from weasel2 (#1335) @TonyBagnall

• [DEP] Deprecating ShapeDTW classifier, to be replaced with a newer correct version (#1367) @hadifawaz1999

Documentation

• [DOCS] hard code shapelet classifiers in notebook (#1331) @TonyBagnall

Enhancements

• [ENH] Update regression pipeline (#1279) @MatthewMiddlehurst

• [ENH] Loading unequal length, no missing values classification problems (#1157) @TonyBagnall

• [ENH] Tidy dummy estimators for classification and regression (#1281) @MatthewMiddlehurst

• [ENH] Learning Shapelet Classifier (#1247) @itsdivya1309

Maintenance

• [MNT] Changing deep learning save format from .hdf5 to .keras (#1292) @hadifawaz1999

Clustering

Bug Fixes

• [BUG] fixed dba test (#1319) @chrisholder

• [BUG] Fix random state for deep learning models in classification/regression and clustering (#1271) @hadifawaz1999

Maintenance

• [MNT] Changing deep learning save format from .hdf5 to .keras (#1292) @hadifawaz1999

• [MNT] Distance test refactor (#1294) @chrisholder

• [MNT] Type hint distance module (#1326) @chrisholder

Datasets

Bug Fixes

• [BUG] write_to_tsfile not writing header (#1329) @vNtzYy

• [BUG] Stop joining Nones (#1353) @TonyBagnall

Enhancements

• [ENH] Loading unequal length, no missing values classification problems (#1157) @TonyBagnall

Maintenance

[MNT] xfail website reads that return a connection error (#1357) @TonyBagnall

Distances

Bug Fixes

• [BUG] Unequal length bounding matrix fixes (#1161) @chrisholder

Enhancements

• [ENH] SBD and MSM: Support pairwise distance calculation for unequal length time series (#1287) @CodeLionX

• [ENH] Capability to use shape-dtw on precomputed transformation (#1323) @hadifawaz1999

Maintenance

• [MNT] Distance test refactor (#1294) @chrisholder

• [MNT] Type hint distance module (#1326) @chrisholder

Regression

Bug Fixes

• [BUG] Fix random state for deep learning models in classification/regression and clustering (#1271) @hadifawaz1999

Enhancements

• [ENH] Remove regression delegate file (#1280) @MatthewMiddlehurst

• [ENH] Updated EncoderRegressor (#1226) @AnonymousCodes911

• [ENH] Hydra and MultiRocketHydra for regression (#1207) @MatthewMiddlehurst

• [ENH] Refactored _TestRegressor in regression testing module to be a MockRegressor in testing/mock_estimators (#1213) @tvilight4

• [ENH] Add IndividualLITERegressor and LITETimeRegressor (#1221) @aadya940

• [ENH] Update regression pipeline (#1279) @MatthewMiddlehurst

• [ENH] Tidy dummy estimators for classification and regression (#1281) @MatthewMiddlehurst

• [ENH] Initializing metrics to “mean_squared_error” In Deep Learning Models (#1317) @harshithasudhakar

Maintenance

• [MNT] Changing deep learning save format from .hdf5 to .keras (#1292) @hadifawaz1999

Segmentation

Deprecation

• [ENH] Deprecate annotators (#1306) @TonyBagnall

Transformations

Bug Fixes

• [BUG] RandomShapeletTransform stdev fix (#1324) @MatthewMiddlehurst

Deprecation

• [ENH] Deprecate transformers (#1307) @TonyBagnall

• [DEP] deprecate ElbowClass prior to move to dedicated package (#1336) @TonyBagnall

Enhancements

• [ENH] Enhancing BaseCollectionTransformers+Tidying Utils dir (#1328) @AnonymousCodes911

• [ENH] remove call to update in BaseCollectionTransformer (#1344) @TonyBagnall

Maintenance

• [MNT] Modified test_base_rocketGPU.py (#1325) @AnonymousCodes911

Unit Testing

Enhancements

• [ENH] Refactored _TestRegressor in regression testing module to be a MockRegressor in testing/mock_estimators (#1213) @tvilight4

Maintenance

• [MNT] added unequal length utils function (#1303) @chrisholder

• [MNT] Move estimator_checks to the Testing Module (#1312) @aadya940

Visualisations

Enhancements

• [ENH] Add plot_spectrogram (#1237) @aadya940

Deprecation

• [DEP] remove plotting from utils (#1334) @TonyBagnall

Other

Deprecation

• [DEP] Remove mlflow (#1285) @MatthewMiddlehurst

• [ENH] Remove deprecated performance metric classes (#1261) @TonyBagnall

• [DEP] remove deprecated ‘save_transformed_data’ and ‘transformed_data_’ in BaseIntervalForest, FreshPrince, STC, (#1332) @TonyBagnall

Documentation

• [DOC] Missing link for aeon-neuro (#1299) @harshithasudhakar

• [DOC] Updates to conform to ruff - pydocstyle hook (#1277) @AnonymousCodes911

• [DOC] Add clustering example to the README (#1354) @chrisholder

• [DOC] Update papers page and remove GSoC advertisements (#1368) @MatthewMiddlehurst

Enhancements

• [ENH] Remove deprecated performance metric classes (#1261) @TonyBagnall

• [ENH] Refactor dask converters (#1302) @TonyBagnall

• [ENH] Replace calls to check_is_scitype that return metadata with validate_input (#1290) @TonyBagnall

• [ENH] reduce warnings (#1337) @TonyBagnall

• [ENH] Replace mtype to scitype function (#1298) @TonyBagnall

Maintenance

• [MNT] Refactored any occurrence of ‘_instances’ to ‘_cases’. ‘series_length’ to ‘n_timepoints’ and ‘n_dims’ to ‘n_channels’ (#1304) @chrisholder

• [MNT] Changes to pyarrow and dask in all_extras to fix CI (#1309) @MatthewMiddlehurst

• [MNT] Cleanup code in aeon.util.conversion module (#1289) @CodeLionX

• [MNT] Add new dependency: typing-extensions (#1327) @chrisholder

Refactored

• [ENH] Replace calls to check_is_scitype that return metadata with validate_input (#1290) @TonyBagnall

Contributors

The following have contributed to this release through a collective 58 GitHub Pull Requests:

@aadya940, @AnonymousCodes911, @chrisholder, @CodeLionX, @hadifawaz1999, @harshithasudhakar, @itsdivya1309, @MatthewMiddlehurst, @TonyBagnall, @tvilight4, @vNtzYy