v0.5.9

This release adds Python 3.13 and drops Python 3.9 support, along with a few quality of life improvements (for example, the NSRR token can now be stored in the config and/or in an environment variable).

v0.5.8

This release provides the three built-in classifiers as Keras 3 models.

v0.5.7

This release adds support for Python 3.12.

v0.5.6

This release switches to edfio for loading EDF files.

v0.5.5

This is a minor release, which only changes how imports are handled within the package. It contains no changes relevant for users.

v0.5.4

This release removes an unnecessary pandas dependency, fixes issues with reading SHHS data, and gets rid of expected warnings during feature extraction.

v0.5.3

This release adds a new function which plots ECG records together with (multiple) annotations. We also tweaked the Pan-Tompkins ECG detector to better handle flat data segments at the beginning of a recording. We also switched our documentation to MkDocs, which looks really nice! Finally, we fixed some minor issues and generally improved consistency of our functions – see our CHANGELOG.md for more details.

v0.5.2

Include classifiers and tests.

v0.5.1

This release adds more documentation for the Pan-Tompkins R peak detector. Also, the preprocess_rri function now returns a copy instead of working in-place.

v0.5.0

This release adds functions for evaluating, saving, and loading sleep stage classification models. WAKE-SLEEP and WAKE-REM-NREM classifiers trained on the MESA dataset are bundled with the package. Further changes include metadata feature extraction and support for Python 3.10.

The full changelog is available here.