Skip to content

Releases: cbrnr/sleepecg

v0.5.8

21 May 09:38
Compare
Choose a tag to compare

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

v0.5.7

08 Feb 08:36
Compare
Choose a tag to compare

This release adds support for Python 3.12.

v0.5.6

22 Dec 13:43
Compare
Choose a tag to compare

This release switches to edfio for loading EDF files.

v0.5.5

01 Jun 08:47
Compare
Choose a tag to compare

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

13 Apr 05:50
Compare
Choose a tag to compare

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

05 Dec 11:08
Compare
Choose a tag to compare

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

02 Aug 08:05
Compare
Choose a tag to compare

Include classifiers and tests.

v0.5.1

21 Jul 14:06
Compare
Choose a tag to compare

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

22 Mar 10:15
Compare
Choose a tag to compare

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.

v0.4.1

14 Jan 11:38
Compare
Choose a tag to compare

Hotfix release to fix reader_dispatch in examples/benchmark/utils.py not yielding anything (#68)