Take the code in https://github.com/lsst/obs_subaru/blob/master/python/lsst/obs/subaru/crosstalk.py and:
- Make it camera-agnostic
- Make it comply with coding standards
- Put it in cp_pipe (the old code can be removed later, so this needn't be considered a move, it's just a new cp_pipe task for now)
Then, use this task and some darks from the AuxTel chip to calc xtalk coefficients for
If it looks like the code if specifically for working with darks then the task should be named as such. If it looks like it will be sufficiently generic to work with saturated stars/CBP exposures then it should have a more generic name.
Output should be a written with a butler.put() with an appropriate dataset type defined in obs_base datasets.yaml.