Currently the fiber input geometric losses are treated as being the same for all objects of a given objtype -- i.e. the loss for the median size for that object under median seeing conditions. This really should be a function of seeing, object size, object profile, and fiber mis-centering. Find a way to efficiently parameterize that and a data model to store it.
The inputs to throughput should then include seeing, objsize, and fiber offsets in addition to objtype.
objsize: pick a name that makes it clear if this is radius, FWHM, diameter, etc.
profile: should support Gaussian, Moffat, and Sersic with various n (de Vaucouleurs, exponential).
Currently the fiber input geometric losses are treated as being the same for all objects of a given objtype -- i.e. the loss for the median size for that object under median seeing conditions. This really should be a function of seeing, object size, object profile, and fiber mis-centering. Find a way to efficiently parameterize that and a data model to store it.
The inputs to throughput should then include seeing, objsize, and fiber offsets in addition to objtype.
objsize: pick a name that makes it clear if this is radius, FWHM, diameter, etc.
profile: should support Gaussian, Moffat, and Sersic with various n (de Vaucouleurs, exponential).