Fix Version/s: None
Team:Data Release Production
Write a PTC Task to calculate the gains for a bunch of flat-pairs.
A very rough sketch/some of the features that it should include:
- Run isr on the images as a subtask. Make sure the steps being done are acceptable.
- Have options to do both paired/unpaired PTCs (use paired taskrunner from
DM-19535for the paired version)
- Have options to calculate it the raw mean/variance way, or use "Robert's method" (already implemented in atmospec but will move from there (and that needs doing anyway))
- Options for binning the images
- Options for subtracting correlations (mutually exclusive with binning? Enforce doing one of the two? (probably not the latter))
Once we have the actual PTC curve:
- Variable fit order (default p3?) (force a zero intercept on the fit?)
- For densely sampled PTCs, get the linearity out as well as the gain (must be careful to separate BF from NL)
- Option for both the gains and the linearity to be butler.put() for the detector
- Have the task be able to output plots of the PTC rather than just dumping the products
- Some internal sanity checking/quality control stuff, e.g. we should persist the fit qualities as well, ideally as errors on the measured quantities