cortex-lab / KiloSort

GPU code for spike sorting
GNU General Public License v2.0
175 stars 100 forks source link

Erroneous templates matched exclusively for part of recording excluding matches of real templates #241

Open McKenzieNeuro opened 3 years ago

McKenzieNeuro commented 3 years ago

I have a strange problem that I am hoping someone has seen since I have not been able to debug it. For some recordings, I see that for the first part of my recording, noise templates are defined and matched over and over and no real templates ever match. Then at some time point where nothing remarkable happens in the raw data, the bad templates never match and the real templates can give me actual unit spike times. The transition point occurs at different moments for different sessions. I attach a screen shot of the issue with two units plotted over time.

Any hints? image

Thank you sam