Changelog

: Something that you couldn’t do before.
: A miscellaneous minor improvement.
: An existing feature now may not require as much computation or memory.
: Something that previously didn’t work as documented or as expected should now work.
: An update to the documentation.
: Miscellaneous updates such as package structure or GitHub quality of life updates.

: Froze all dependency versions; included threadpoolctl version requirement.
: Fixed dependency naming and requirements in setup.py and requirements.txt.
: Updated model naming schema when loading a model to use the file name as the model name.
: Updated relevant documentation to incorporate changes.

: Updated README to reflect freezing of dependencies.
: Specified exact versions of dependencies in requirements.txt and in setup.py.

: Remove min_impurity_split from GradientBoostingClassifier as it no longer exists in the underlying sklearn class.
: Improve organization of the build directory.

: Sleep Ensemble is released! Hurray!
: IO interface selection for read and write implemented. Interfaces for .mat exports from Spike2, .smr/.smrx from Spike2, and .xls "datasets" have been included.
: End-user command line interface application developed. Includes ability to classify data, train/validate models, load/export builds.
: Protocol management embedded into CLI application similar to IO interface integration.
: Sleep Ensemble 4 protocol added to sleepens.protocols.
: README.md includes more details on installation and usage.
: BUILDS.md written as a build directory to easily find pre-trained build binaries from release history.

Provide feedback