v0.6.2

• Hot fix of wrong documentation url

v0.6.1

• Migrated the documentation to a new host
• Fixed minor typos in the documentation
• Fixed a minor type hinting bug

v0.6

• Re-designed the API of adtk.visualization.plot

• Removed adtk.data.resample because its functionality is highly overlapped with pandas resampler module

• Made adtk.data.expand_event accept events in the form of pandas Series/DataFrame

• Made adtk.data.expand_event accept time delta in the form of str or int

• Changed the output type of adtk.data.split_train_test from a 2-tuple of lists to a list of 2-tuples

• Turned the following model parameters required from optional

  • window in adtk.detector.LevelShiftAD
  • window in adtk.detector.VolatilityShiftAD
  • window in adtk.transformer.RollingAggregate
  • window in adtk.transformer.DoubleRollingAggregate
  • model in adtk.detector.MinClusterDetector
  • model in adtk.detector.OutlierDetector
  • target and regressor in adtk.detector.RegressionAD
  • target and regressor in adtk.transformer.RegressionResidual
  • aggregate_func in adtk.aggregator.CustomizedAggregator
  • detect_func in adtk.detector.CustomizedDetector1D
  • detect_func in adtk.detector.CustomizedDetectorHD
  • transform_func in adtk.transformer.CustomizedTransformer1D
  • transform_func in adtk.detector.CustomizedTransformer1D
  • steps in adtk.pipe.Pipeline

• Added consistency check between training and testing inputs in multivariate models

• Improved time index check in time-dependent models

• Turned all second-order sub-modules private, and a user now can only import from the following first-order modules

  • adtk.detector
  • adtk.transformer
  • adtk.aggregator
  • adtk.pipe
  • adtk.data
  • adtk.metrics
  • adtk.visualization

• Refactored the inheritance structure of model components (see https://arundo-adtk.readthedocs-hosted.com/en/latest/inheritance.html#inheritance)

• Added Python 3.8 support

• Fixed compatibility issues with statsmodels v0.11

• Fixed compatibility issues with pandas v1.0

• Created an interactive demo notebook in Binder

• Added type hints, and added type checking in CI/CD test

• Added Black and isort to developer requirement and CI/CD check

• Optimized release process by publishing package to PyPI through GitHub Actions

• Improved docstrings and API documentation

• Fixed many minor bugs and typos

v0.5.5

• Fixed a bug that empty lists were ignored by AndAggregator
• Fixed some typo in the documentation

v0.5.4

• Optimized the workflow of how a univariate model is applied to pandas DataFrame
  • Added more informative error messages
  • Fixed some bugs resulting in model-column matching error due to inconsistency between output Series names and DataFrame columns
  • Clarified the workflow in the documentation

v0.5.3

• Quick hotfix to avoid errors caused by statsmodels v0.11 by requiring statsmodels dependency <0.11

v0.5.2

• Formalized the management of releases and pre-releases, including rules of branches and versioning
• Added more rules for developers to the documentation

v0.5.1

• Added many new unit tests, and modified some old unit test
• Removed seaborn from dependencies (use matplotlib built-in style now)
• Fixed a bug in the metric module of dict objects as input
• Fixed a bug in the detector OutlierDetector that output series has dtype object if NaN is present
• Fixed a bug in transformer pipeline that detect and transform methods are confused
• Fixed a bug in pipenet that an aggregator node may crash if its input is from a node where subset contains a single item
• Fixed a bug in pipenet summary that subset column are always "all" even if not
• Some minor optimization of code

v0.5

• Changed the parameter steps of pipenet from list to dict

• Added method summary to pipenet

• Corrected some major algorithmic issues on seasonal decomposition

  • Removed STL decomposition transformer, and hence the corresponding option in SeasonalAD detector
  • Recreated classic seasonal decomposition transformer

• Updated the demo notebook in the documentation

• Added an option to hide legend in the plotting function

• Added some package setup options for developers

• Fixed an issue of tracking Travis and Coveralls status

• Some minor internal optimization in the code

• Fixed some format issues and typos in the documentation

v0.4.1

• Fixed an issue of tox environments
• Minor spelling/grammar fix in documentation