search

Tyelab / bruker_control

Repository for code running the lab's multi-photon imaging experiments at the Bruker Ultima Investigator scope. Unites Arduino, Prairie View, and GenTL machine cameras in Python.
Mozilla Public License 2.0
8 stars 3 forks source link

Not Enough Airpuffs During Pre-Opto Sessions. #157

Closed jmdelahanty closed 1 year ago

jmdelahanty commented 1 year ago

While doing data analysis of the newest CMS Cohort, Jim has run into problems where a mean is being attempted upon just one datapoints. There is currently no check that the pre and post opto stages have adequate numbers of airpuff trials in particular. While there are checks to ensure that there are at least the appropriate amount of airpuffs overall and that enough are present in the opto stimulation trials themselves, there aren't checks in place ensuring that there's a certain amount scheduled before and after.

This can be solved by both increasing the amount of airpuff trials and, mostly, generating trials as described in #148 as Autopilot does.

jmdelahanty commented 1 year ago

Code in develop appears to solve this issue. Doesn't do it well, but added new configurability for pre-post opto airpuff trial numbers.

jmdelahanty commented 1 year ago

It looks like this might not have been an issue after all, which is lucky honestly, but rather due to errant triggers being sent for either LEDs or shutter triggers somehow. Having what's in develop will make sure it's never an issue.

It doesn't make any sense yet how these triggers were sent. The configuration files don't have any scheduled prematurely and the only way it could happen in the Arduino code would be if there are pins set high errantly as well for a certain trial type, but there doesn't yet appear to be a pattern in it that would indicate a bug like this is the problem.

The only guess I have is either prairie view settings were incorrectly set, which has happened before as fixed in #139 (but that was solved months before and these LED triggers are happening at strange times in the middle of the recording ) or there was some strange grounding issue that was causing triggers in strange places when something was moved or a nearby pin was triggered...