Adding stimulus annotations and calculating with different conditions takes too much time - instead, caching the data enables faster analysis with multiple different combinations of stimulus (potentially with other conditions such as licking, running, pupil dynamics, engagement, face and body motion, etc)
Having session-wide traces with conditions such as task, gray screen, auto-rewarded, fingerprinting, etc help analyzing interaction between neurons with different conditions easily.
Need to implement for both events and dffs
Tasks
[ ] Cache response with stimulus annotations, for both dff and events
[ ] Cache whole-session traces with condition annotations, such as gray screen, auto-rewarded, and fingerprint, for both dff and events
Considerations
Caching window: baseline and response duration: For now, cache responses to each image presentation including omissions at [-3, 3.1] (for visualization purposes later on; free to pass a window as input arg)
Max estimated file size ~ 2.3 GB (max # of neurons per exp: 500. 4800 stimuli (60 min with 0.75 s interval) X 20 Hz X 6 s X 500 cells X 8 byte (double for each point) ~ 2.3 GB)
Analysis window: For each analysis, again free to pass a window as input arg, but default as [-0.2, 0.75]
Analysis parameters: can be added in the future.
Sampling rate and interpolation: Have options to be passed as input args, default to be interpolation with 20 Hz
Cache file name should indicate either dff or events, and the sampling rate (in case of interpolation) or 'no_interp' (when not interpolated, the sampling rate will be same as the imaging frequency)
Purpose:
Tasks
Considerations