Tsfresh

==============
- Bugfixes/Typos/Documentation:
* Fixes issue 1073: Updated setup.cfg to require newer scipy version (1081)
* extract_relevant_features now passes chunksize to extract_features (1083)
* Fix code and tests for numpy >= 2.0 (1085)
* Update tsfresh.feature_extraction.feature_calculators.skewness to make it consistent with the design principle of not ignoring nan (1066)
* Fix spelling/grammar in pipeline notebook (1082)
* Added recommendation to revert thread limitations (1069)
* Fix the 01 example notebook to not leak information between train and test set
* Feature calculator return type documentation (1070)

==============
- Added Features
- Make Dask and Distributed optional dependencies (1061)
- View and Set N Jobs (1029)

- Bugfixes/Typos/Documentation:
- Extra notes on parallelization efficiencies (1046)
- Update doc extraction settings for clarity and formatting (1033)
- Typos (1031, 1034, 1049, 1048)

==============

- Added Features
- Make tsfresh compatible with numpy 1.24 (1018) and pandas 2.0 (1028)

- Bugfixes/Typos/Documentation:
- Use pandas Index.equals in check_if_pandas_series (963)
- Updates to package layout, CI/CD and developer setup

==============

- Breaking Change
- The matrixprofile package becomes an optional dependency

- Bugfixes/Typos/Documentation:
- Fix feature extraction of Friedrich coefficients for pandas>1.3.5
- Fix file paths after example notebooks were moved

==============

- Breaking Change
- Drop Python 3.6 support due to dependency on statsmodels 0.13

- Added Features
- Improve documentation (831, 834, 851, 853, 870)
- Add absolute_maximum and mean_n_absolute_max features (833)
- Make settings pickable (845, 847, 910)
- Disable multiprocessing for `n_jobs=1` (852)
- Add black, isort, and pre-commit (876)

- Bugfixes/Typos/Documentation:
- Fix conversion of time-series into sequence for lempel_ziv_complexity (806)
- Fix range count config (827)
- Reword documentation (893)
- Fix statsmodels deprecation issues (898, 912)
- Fix typo in requirements (903)
- Bump statsmodels to v0.13 (
- Updated references

==============

- Added Features
- Allow arbitrary rolling sizes (766)
- Allow for multiclass significance tests (762)
- Add multiclass option to RelevantFeatureAugmenter (782)
- Addition of matrix_profile feature (793)
- Added new query similarity counter feature (798)
- Add root mean square feature (813)
- Bugfixes/Typos/Documentation:
- Do not send coverage of notebook tests to codecov (759)
- Fix typos in notebook (757, 780)
- Fix output format of `make_forecasting_frame` (758)
- Fix badges and remove benchmark test
- Fix BY notebook plot (760)
- Ts forecast example improvement (763)
- Also surpress warnings in dask (769)
- Update relevant_feature_augmenter.py (779)
- Fix column names in quick_start.rst (778)
- Improve relevance table function documentation (781)
- Fixed 789 Typo in "how to add custom feature" (790)
- Convert to the correct type on warnings (799)
- Fix minor typos in the docs (802)
- Add unwanted filetypes to gitignore (819)
- Fix build and test failures (815)
- Fix imputing docu (800)
- Bump the scikit-learn version (822)