Location:Comment on this issue
Quantum efficiency curves are being measured as a function of wavelength on LSST sensors. For purposes of verification, the commissioning team would like to track these measurements as a function of time. Since we already do this with defects and have a mechanism in place to do that, my proposal is to track these in a similar way.
This will require some work outlined by the following steps:
1) As with the defects, we will define an ascii representation for the QE curves in e.g. ecsv. This will require translation from the native FITS format they are currently in.
2) Again following the model of defects, we will store them in obs_lsst_data with an infinite validity range (since we will initially only have one set)
3) Once the format is defined, we can create a new butler dataset for access via gen2. Gen3 will need to be considered as well.
4) Add a free function in obs_base (we will try to avoid a specialization in obs_lsst) that takes a set of QE curves as retrieved by the Butler, a Detector object, a Point2D, and a wavelength that returns a scalar of the QE at that point and wavelength.
I believe all of these steps are reasonably straight forward and does not imply a change to any APIs (only the addition of a single free function).