Closed Sepidak closed 4 years ago
A new function called "Normalised Kinematic Spiking Frequency" has been added to the "Rotation Analysis" function groups.
As per the requests above the user can plot the normalised speed spiking frequencies (spiking frequencies have been normalised to the 0-5 deg/s bin). The user can also specify whether they want to smooth the mean signal (Note - the size of the smoothing window must be an odd number otherwise the user will receive an error popup message).
After multiple rounds of testing with normalisation to the peak, baseline, or average, none seems to be giving a meaningful outcome. I think we should have done this differently, by first subtracting the baseline from all bins (so the first bin will be 0 for all cells), and then for each cell normalising to the maximum firing across "all the 3 conditions". When we normalised to the max, we did it within each condition and that must be the reason why the outcome was weird. We want to compare the 3 conditions so the normalisation (when using the peak) should be based on all the 3. Otherwise they will all go from 0 to 1 and look the same.
1-for each cell calculate the firing rate per speed bin (combined positive and negative velocities) 2- normalize the firing rates to that of the first bin (0-5 deg) 3- show mean +sem of all cells