Smoothing functions applied to spike data result in still-jagged outputs

NeuralAnalysis / PyalData

Repository for the Python implementation of the TrialData analysis library.

GNU General Public License v3.0

7 stars 9 forks source link

Smoothing functions applied to spike data result in still-jagged outputs #116

Closed raeedcho closed 3 years ago

raeedcho commented 3 years ago

Applying any of the available smoothing functions (anything that uses util.norm_gaussian_window) to spike data seems to result in slightly jagged output (example image below)

I'm pretty sure this is because the default window for util.norm_gaussian_window is 5 standard deviations long, which results in tails that aren't at zero (see below)

I propose we increase the length of the default window to 10 standard deviations long, which seems to get rid of the jaggedness for me.

bagibence commented 3 years ago

Good catch, makes sense, merged the PR. Thank you! 👍