Feat metadata mapping

[micahwoodard]

Adds metadata mapping for acquisition, instrument, and data description as well as a cli. Based of vr foraging metadata mappers. Also some fixes for coupled trial generator. There are some things in data description like funding and investigators that need to be worked out

[bruno-f-cruz]

You can validate this with the TrialOutcome pydantic class to make your life easier

[micahwoodard]

@arielleleon @saskiad this is the pr mapping the dynamic foraging output into an acquisition and instrument model.

[micahwoodard]

Updates mapper to classes to match vr foraging. resolves #97

[micahwoodard]

acquisition.json
instrument.json

@alexpiet these are new files generated with fixes for issues found in #107. There a couple of things I noted in the issue that will take more work

[alexpiet]

Taking another look at these, more feedback

• acquisition, the water calibration has repeated information under the input/output, and then under fit/fit_parameters/measurements. Its confusing to me. I'm not sure if this is the way the schema is designed, or a choice you are making.
• The water calibration is duplicated in both acquisition and instrument, to me it makes more sense in instrument. But someone from scicomp should weigh in
• This is feedback on the calibration itself, but it appears we do 200 repeats at valve_open_time of 0.02, and 1000 repeats at 0.03 and 0.04. Why the difference?
• instrument/components/camera, why is the camera model, frame_rate, recording_software not listed?
• instrument/components/lens, why is there no model information?
• instrument/components/behavior board, why is firmware/hardware/core/tag version not listed? (similarly for the white rabbit and sound card and environment sensor)
• lickspout stage has no information about model or manufacturer

[micahwoodard]

merging since this has been running as acquisition code for testing for about ~2 months. Can continue to iterate on metadata mapping as issues arise