Pycytodata

This is a minor release that fixes a critical bug that affects all previous releases.
Update is strongly recommended.

Bug Fixes

- Fixed a critical issue with ``PyCytoData.load_dataset()`` not working due to old web source down ([3](https://github.com/kevin931/PyCytoData/issues/3))
- This issue stems from upstream web source, thus affecting all previous releases.
- This update fixes the issue with a different implementation.
- There is no implementation change.


Changes and New Features

- No new feature added.

This is a minor release with various bug fixes and documentation improvents.

Bug Fixes

- Fixed a potential issue with loading benchmark datasets' samples out of order (This behavior is not guaranteed, but a implementation detail)
- Fixed an issue with ``DataLoader.load_dataset`` not recognizing downloaded datasets
- Fixed an issue with ``preprocess.bead_normalization`` having uninitialized array (9)

Changes and New Features

- No new feature added
- Improved documentations with streamlined front page and updated links (4 )
- Added docstrings for `+` and `+=` operators
- Updated references for CytofDR paper publication in Nature Communications (4 )

Known Issues
- Reference datasets still cannot be downloaded due to broken source links (3 )
- `data.load_expression` does not report mismatched columns when loading multiple datasets (7)

datasets.rev.1
This release contains the three built-in datasets in PyCytoData: Levine13, Levine32, and Samusik. The datasets are originally obtained from FlowRepository (Access Number: [FR-FCM-ZZPH](http://flowrepository.org/id/FR-FCM-ZZPH)) and distributed in accordance with the licensing terms. You do not need to manually download these datasets yourself: the updated package (starting from v0.1.2) will automatically handle the data processing.

This is the first official release of ``PyCytoData`` with LTS.

Bug Fixes

- Fix the default cofactor of ``PyCytoData.preprocess``
- Fix lineage channels for ``levine32`` and ``samusik``

Changes and New Features

- ``PyCytoData.run_dr_methods`` now runs on lineage channels only
- Add tutorials and references for documentation
- Improve docstrings and general documentations
- Add preprocessing options for loading benchmark datasets
- Relax ``numpy`` dependency to >=1.20
- Add descriptors for PyPI releases

Deprecations

- Permanently remove ``FileIO.make_dir`` function for safety reasons

Release Notes

- This is the first official pre-release of the ``PyCytoData`` package.
- We have proper support for the following workflows, including:
- Downloading data
- Using PyCytoData as CyTOF data analysis pipeline
- FileIO
- CyTOF DR Integration
- Releases on PyPI and conda

Known Issue

There is a potential issue of compatibility with ``CytofDR`` on ``conda``. If a problem occurs, try
using pip instead.