Spyking Circus 2 Too Many Redundant Units

[msj220001]

Hello,

I am using Spyking Circus 2, an internal sorter.
The data set I am using is .plx. It is a 10 minute recording from a low-density 16 channel probe.

Spyking Circus 2 is outputing upwards of 300 clusters, but many of those clusters are the same units and it takes a long time to manually curate and merge them together.

What parameters should I adjust that would reduce the number of redundant units?
I've identified some adjustable parameters in spyking_circus2.py file, hdbscan.py file, and random_projections.py file.
Would any of the parameters in each of these files accomplish this?

[yger]

This has been fixed in recent PR #2574 and it should be merged soon. SC2 was clearly performing badly for low channel count electrodes, and it should behave now more robustly