Export time-dependent templates

[rossant]

It would be great if KS2 could export the templates at different times of the recording in order to better track the drift
cortex-lab/phy#887

[marius10p]

Due to recent developments in the algorithm, I don't think that will be necessary anymore.

[czuba]

OH, bleeding edge! ...any chance recent KS2 developments related to solving the temporal stuttering issue that we were digging into at the workshop? [crosses fingers]

I ask because I just came into a block of time/need to find a solution to this sorting hiccup, and one lingering feature that we wanted to look at was time-dependent template evolution (esp. with respect to 'real' vs 'batch reordered' time).

...I just tried running the current KS2 master branch {last commit June 14, 2020}, but I'm still seeing strong batch-size-dependent stuttering in the output clusters.

[marius10p]

I don't know, the new drift correction strategy doesn't use batch re-ordering any more or any kind of tracking, but I've only tested it on dense Neuropixels probes. I don't want to give you false hope, it might be that your best sorting would still be with the old version.