Demos#
Runnable examples are in the demo/ folder of the repository (setup and data description: demo/README.md). Each one runs against brainmaze-eeg 3.1.0 and brainmaze-utils 3.0.0 from PyPI; matplotlib is only needed for the figures.
- eeg_wave_detection
Slow-wave downslope features with
WaveDetectoron a 6.8-h recording (Carvalho et al. 2024): the default andtrough='paper'modes. See EEG Wave Detection and Analysis.- spike_detection
The Janca (
'spike'preset) and Barkmeier detectors scored against planted spikes;GapAwareSpikeDetectoron NaN gaps and on constant-value dropouts (synthetic and a real amplifier dropout); the'ripple'preset. See Spike detectors.- sleep_classification
KDEBayesianModeltrained on half a night and applied to the other half: skipped epochs, and out-of-distribution epochs labelled'UNKNOWN', with a table of which artefacts are flagged, which silently change the stage and which change nothing; accuracy, balanced accuracy and kappa. See Classifiers.