Search is not available for this dataset
sampling_rate_hz float64 20k 30k | n_channels int64 32 64 | n_samples int64 3M 4.5M |
|---|---|---|
20,000 | 64 | 3,000,000 |
30,000 | 32 | 4,500,000 |
20,000 | 64 | 3,000,000 |
30,000 | 32 | 4,422,830 |
20,000 | 64 | 3,000,000 |
30,000 | 32 | 4,500,000 |
20,000 | 64 | 3,000,000 |
30,000 | 32 | 4,500,000 |
20,000 | 64 | 3,000,000 |
30,000 | 32 | 4,500,000 |
20,000 | 64 | 3,000,000 |
30,000 | 32 | 4,500,000 |
Paired spike sorting — benchmark segments
Two time segments of six paired (extracellular + juxtacellular/loose-patch) recordings from
SpikeForest, labeled A–F:
visible/<X>/: 0–150 s of the recording.heldout/<X>/: 150–300 s of the recording (recordingDends at about 297.4 s).
Each folder has traces.npy (int16, shape (n_samples, n_channels), unaltered, source channel order),
geometry.csv (channel positions) and meta.json (sampling_rate_hz, n_channels, n_samples).
No ground truth is included. These are benchmark segments.
Sources and license
Recordings come from SpikeForest PAIRED_ENGLISH (English et al., 2017, Neuron) and
PAIRED_MEA64C_YGER (Yger et al., 2018, eLife; Zenodo 10.5281/zenodo.1205233),
both distributed under CC BY 4.0. This derived dataset is released under CC BY 4.0; please cite the original authors.
- Downloads last month
- 149