Write co-registration settings to JSON param file

[alex-l-kong]

What is the purpose of this PR?

Closes #54. Currently, after running update_coregistration_params.ipynb, the new conversion factors are appended directly to settings.py. This causes a lot of clutter and causes settings.py to diverge from main. We should write the params to untracked temp .json file instead.

How did you implement your changes

The following changes have been added:

• write_coreg_params has been nuked: instead we save to a file called stage_optical_coreg_params.json in toffy directly in update_coregistration_params.ipynb.
• convert_stage_to_optical now includes a a default argument stage_optical_coreg_path set to stage_optical_coreg_params.json. If this path exists, then we read the coregistration conversion factors directly from the created .json file. Otherwise, we use the default factors in settings.py.

[review-notebook-app]

Check out this pull request on 

See visual diffs & provide feedback on Jupyter Notebooks.

---

Powered by ReviewNB

[ngreenwald]

Okay, this is a good start.

I think we should standardize the coreg file naming, and not give the user the option to modify it. In addition, lets completely remove the coreg setting defaults from settings.py. If there's no coreg file detected, the fucntions that rely on one should give the user an error message and direct them to run the notebook. The defaults are not going to transfer across instruments, so I think it makes more sense to force the user to run it once on each instrument.

If we have a structured name scheme (i.e. coreg_params_yymmdd), then we can search for the files in the expected places.

The alternative is to append successive coreg updates to the same file (coreg_params.json) with some text in the JSON about when they were added. Whichever you prefer.

[alex-l-kong]

@ngreenwald I think it's easier to write successive coreg updates to the same file. We'll store each different co-registration parameter set in list format such as:

{
    'coreg_params': [
        {
            'STAGE_TO_OPTICAL_X_MULTIPLIER': ...,
            'STAGE_TO_OPTICAL_X_OFFSET': ...,
            ...
            'date': '220322'
        },
        {
            ...
            'date': '220323'
        }
    ]
}

In this way, we have one JSON file (as opposed to multiple that would create more clutter in toffy and be harder to keep track of) and simply take the most recent set of params created (the last element in 'coreg_params').

This will allow us to remove the STAGE_TO_OPTICAL_... params from settings.py. However, we'll still need to keep the MICRON_TO_STAGE_... constants around somewhere. I think it's easiest to leave these in settings.py...alternatively, we could put these in a separate .json file as well. I'm open to other ideas here.

[alex-l-kong]

Per our meeting, the following have been updated:

1. settings.py no longer contains STAGE_TO_OPTICAL_ constants, however we'll leave the MICRON_TO_STAGE_ constants there
2. The STAGE_TO_OPTICAL_ constants generated by update_coregistration_params.ipynb will now be saved to toffy/coreg_params.json. It will contain a key called coreg_params which defines a list of every set of conversion factors generated by the user. Each set of co-registraion params will also contain a key called 'date' which will indicate the timestamp they were generated. Only the most recent (the last one appended to the coreg_params list) is used.
3. If the user has not run update_coregistration_params.ipynb before autolabel_tma_cores.ipynb, an error message will direct them to run the co-registration script first.

[ngreenwald]

Looks good, just a typo