QC capsule for dynamic foraging behavior (modality:behavior) raw data acquired together with HARP/Bonsai-based behavior
run_capsule.py : main script
Following the "alternate-workflow" with which you don't need to make a new asset, instead directly pushing QC.json to DocDB.
https://github.com/AllenNeuralDynamics/aind-qc-portal?tab=readme-ov-file#alternate-workflow
QC tests implemented
- Check for dropped frames in cameras
- If dropped frames exist, report number dropped for each camera
- Check that experimenter name is not the default name
- Add QC alert for dirty files
- check for side bias (extreme bias < 1, and average bias < .75)
- should add check for lick interval less than 50ms
Steps:
1.reading rawdata
2.generating figures and metrics
3.submitting figures to kachery to obtain unique url
4.Composing QC/QCevals/QCmetrics
5.Pushing QC,josn to DocDB
6.visualizing, manual QCing under AIND-QCportal-app