Closed DowntonCrabby closed 2 years ago
Probably what Pika needs to do is make BehaviorSession
automatically detect the presence/absence of eyetracking data and gracefully return None
if it is absent. The current API requires users to specify skip_eyetracking
by hand when instantiating the BehaviorSession from LIMS.
Hi all. I opened issue #2566 to address Scott's above comment. Can confirm that eyetracking is associated with the session, we just need to handle the behavior only cases a bit more cleanly. Closing this issue.
Describe the use case that is addressed by this feature. Currently the eye tracking data (specifically the
eye_tracking
table attribute) is only associated withophys_experiment
orecephys_session
objects. This is confusing pedagogically and is better organized at thebehavior_session
level.The
behavior_session
objects should contain mouse behavior level data streams such as running, licks, task performance etc which eye tracking definitely falls in to- and is more logically grouped there. Then theophys_experiment
or theecephys_session
objects contain all the behavior information + the physiology.Describe the solution you'd like The eye tracking data, specifically the
eye_tracking
table attribute should be loaded as part of thebehavior_session
dataset object (for both vbo and vbn).Additional context @matchings or @corbennett might have additional context or notes to add.