Feature request: Criteria for good units (Dohoung Kim)

[jamesjun]

Criteria for good units: Is there any criteria that you use to decide whether the unit is good or not? I previously used MClust and was heavily dependent on isolation distance over 18~20 and L-ratio below 0.05. In JRCLUST, I found some cases that the unit seems pretty well separated, but the calculated index is not very good. For example, the unit in the attached file seems clearly isolated, L-ratio is really high (1.5). I thought that L-ratio might be not adequate for our IMEC recording settings, but I don’t know the exact reason. So I thought that displaying mean rho and delta value might be helpful to decide the quality of the units. If you think it’s reasonable, and if it could be plotted as a circled point in the ‘Cluster rho-delta’ window, it would be awesome.

-Dohoung Kim