mouse performance for session 783927872

[gkocker]

@nicain

Code to calculate hit rate for this session. It looks really low, and gives a negative d prime.

from allensdk.brain_observatory.behavior.behavior_ophys_session import BehaviorOphysSession
from allensdk.brain_observatory.behavior.behavior_ophys_api.behavior_ophys_nwb_api import BehaviorOphysNwbApi
from allensdk.internal.api.behavior_ophys_api import BehaviorOphysLimsApi

ophys_experiment_id = 783927872
filepath = 'behavior_ophys_session_783927872.nwb'.format(ophys_experiment_id)
nwb_api = BehaviorOphysNwbApi(filepath)
session = BehaviorOphysSession(nwb_api)

hit_rate = session.trials['hit'].sum() / session.trials['stimulus_change'].sum()
fa_rate = session.trials['false_alarm'].sum() / session.trials['stimulus_change'].sum()

print(hit_rate)

[nicain]

@Gocker Thanks for QCing this experiment. Given this observation, what would be the course of action that would resolve this issue? Failing the experiment, or something else?

[gkocker]

Yes, I think the possible course of action is failing the experiment. I thought I'd heard that performance (d' > 1) was a QC factor for the behavior experiments, but am not 100% sure. @matchings ?