This is necessary if we are going to have multi-wavelength stuff work naturally. Or at least hogghogg thinks so.

Leslie confirmed to me that the finite-element method is not a good idea. He will suggest something involving Green functions

If I set up the default geometry well, the trivial config (all sources have the same amp and phase) will lead to an in-focus point-source image in the center of the one-d field.

Need a routine that visualizes not just the output intensity but also the input sources, with the amplitudes and phases (possibly shown as phasors?). Imagining a two-panel figure, one panel of which shows the intensities at the image points, and the other shows the amplitudes and phases at the source points.

Need code that "takes a snapshot" of an intensity field and gets noise added.

wrap everything and fire it into leastsq for optimization

In preparation for making a toy model coronograph.

and they should not include low-order distortions, given the AO system working out front.

Needs to either get Gaussian noise or else do a Poisson sampling based on the Intensity-implied rate.