spikeinterface
spikeinterface copied to clipboard
Sorting analyzer behavior with multi probe recording
Hi, I have a problem with the following pipeline: I have a 2 NPX1 recording. I sort each probe individually. I curate each one in phy. I load each sorting with read_phy and aggregate them with aggregate_units. I compute extensions on this aggregated sorting.
The waveforms from this aggregated sorting analyzer are, however, off. If I plot them using plot_units_waveforms, I see four columns of waveforms, two columns are fine, two columns are flat lines (I'll post a pic later). Notably, if I plot using a sorting_analyzer created from an individual sorting (not the aggregated one) waveforms are fine. I guess the waveform extraction is not noticing the probe group of each channel and picking the channels corresponding to the peak template in both probes? Not sure though.
Should I use the sorting analyzer on not aggregated sorting objects only?if yes shouldn't there be some correspondent method to aggregate_units for the sorting analyzers too?
Thanks!