Handle defocused and in-focus observations

[tobias-liaudat]

We want to handle a dataset containing in-focus and defocus observations of possibly the same stars.

The underlying WFE PSF model we are constraining is the same one, the difference is that the defocused observations will have a supplementary delta Zerkine contribution that has to be taken into account when optimising the model.

Having a fixed Zernike contribution can be handled with the physical layer. This will require properly building the associated Zernike prior to the dataset composed of all the stars (in-focus and defocused). The corresponding contribution of the defocused star is to be defined as the moving of one mirror in a complex optical system adds a more complex contribution than only the Zk4.

The first thing to do is to simulate a focus+defocus simulated dataset to validate the approach, which should include the associated zernike prior.