I have a bare-bones version where I calculate a best fit line and see whether it is significantly different than zero. On either side of zero AHV. This can be done on binned values or all (raw) spikes. All spikes is more conservative. Need to add a slope threshold, like Taube lab does. Is it worth doing curve fitting with sigmoids? Breaking the TC up into line segments seems clunky.
I have a bare-bones version where I calculate a best fit line and see whether it is significantly different than zero. On either side of zero AHV. This can be done on binned values or all (raw) spikes. All spikes is more conservative. Need to add a slope threshold, like Taube lab does. Is it worth doing curve fitting with sigmoids? Breaking the TC up into line segments seems clunky.