search

AllenInstitute / visual_behavior_analysis

Python package for analyzing behavioral data for Brain Observatory: Visual Behavior
Other
21 stars 6 forks source link

Event Oscillations #697

Closed alexpiet closed 2 years ago

alexpiet commented 3 years ago

The mesoscope events have oscillations before and after real events. We believe this is due to the way Peter L resampled time. From Peter: “Yes, likely the down-sampling from 30 Hz to 10Hz. We may have to revisit the method”

dougollerenshaw commented 3 years ago

Tagging @ledochowitsch

dougollerenshaw commented 3 years ago

Sample plot that @alexpiet generated: image

alexpiet commented 3 years ago

Legend for that plot. Dots are the detected event magnitudes for an example mesoscope-cell and scientifica-cell. I looked for a time point where there happened to be a similar amplitude event for both cells at about the same time.

The solid lines are the filtered_events computed by VBA. We see the mesoscope cell has oscilations in the events.

alexpiet commented 2 years ago

I'm tracking down loose issues, and I can't replicate this. I'm spot checking sessions for (1) negative event values, (2) comparing dff and events for random cells and looking for oscillations.

alexpiet commented 2 years ago

I dont know how to systematically check for oscillations, but I can systematically check for negative values. So I'm running a script over night to screen for negative event values.

alexpiet commented 2 years ago

I found no sessions with negative event values. I'm going to close this issue