LorenFrankLab / spyglass

Neuroscience data analysis framework for reproducible research built by Loren Frank Lab at UCSF
https://lorenfranklab.github.io/spyglass/
MIT License
89 stars 41 forks source link

Sparse Waveforms from spikeinterface #1058

Open samuelbray32 opened 1 month ago

samuelbray32 commented 1 month ago

Describe the bug spikeinterface.extract_waveforms has an argument sparse. If this is True the waveform extractor will only return the waveforms on a subset of channels that it deems important on a per-unit basis through one of a set of algorithms here. This is an issue for clusterless feature extraction for probes, as all events are a single "unit" and can have features on very different channels.

Problem: The default value for sparse changed from False to True between spikeinterface versions 0.98.2 and 0.99.0

Solution:

samuelbray32 commented 1 month ago

See new amplitude_probe parameter set in WaveformFeatureParams table for example solution