post-Cycle 9 Roman module development

[rmandelb]

Now that the galsim.roman module is up to date with Cycle 7, there are some additional tasks that could be done:

• Make Roman module components accessible in config
• Include dither sequences
• Individual pixel effects: ability to put in a flat field including effects down to the sub-pixel level. Cross-hatching pattern could then be simulated.
• Defects like cosmic rays
• correlated read noise: should store a correlation function in Roman module and use it to generate the read noise
• brighter-fatter: empirically-based Antilogus-type implementation (CDModel) using coefficients measured by Chris
• understanding WCS offsets from Chris's code

Other potential improvements to that module can be collected here.

[rmjarvis]

I noticed when writing text for Troxel's paper that the IPC coefficients here don't match what we are using.

I couldn't actually find the reference for our coefficients, but at least we should probably update our IPC numbers to match this one:

[0.21, 1.66, 0.22]
[1.88, 91.59, 1.87]
[0.21, 1.62, 0.20]

(unnormalized obviously)

Also, it doesn't make much sense to me that this array would not be 180 degree rotationally symmetric. Shouldn't the average correlation between a pixel and the one above it be the same as with the one below it? Likewise upper-left and lower-right, etc. I would have thought there should only be 5 independent numbers here. Am I wrong about this?

[rmjarvis]

Still more things to do here, but I turned the above into a task list and checked off a couple items.
In particular, I think the WCS is pretty solid here, so I don't think whatever that was referring to is still pending. And the Roman stuff is pretty much all config-usable now, so that's been done for a while. (Since #1121)
Also, cross-referencing PR with Cycle 9 changes: #1251