Finalize Jones Pupil to FITS file conversion

[Jashcraf]

poke.writing.JonesToFITS() needs some updating with keywords that are meaningful.

[Jashcraf]

I think just some simple headers for wavelength + pixelscale would be helpful

[Jashcraf]

@kenjim21 here's another fairly self-contained project that requires actually using Poke and stuff. I might need to generate some data for you.

Poke creates numpy arrays containing "Jones Pupil" data, which kind of tells you how the polarization state evolves through a given optical system (like a telescope!). It'd be convenient to transfer these numpy arrays with a file format that telescopey people use, so having a dedicated function in poke.writing that does this would be ideal.

Enter the FITS file format: https://fits.gsfc.nasa.gov/fits_primer.html (primer)

This is sort of used as a catch-all file transfer format in astronomy, because we work with a lot of data that takes the shape of images. The big takeaways are that the aforementioned numpy arrays are the "images" that I want to save. And I want them to include a header full of information about these data.

I've started working on this but it's barely functional: https://github.com/Jashcraf/poke/blob/a15947c959ded991f1d152927fc24c1edfc5ba42/poke/writing.py#L97

Would you be willing to complete this function? If so, give the FITS file primer a read and we can start talking about how to do a good translation with Poke.

[Jashcraf]

@kenjim21 you said you had some Q's on this part, would you like to ask them here or set up a meeting to discuss the more physics-y side of Poke?

[kenjim21]

@Jashcraf I'd appreciate setting up a meeting to discuss the tree physics-y side of Poke just so I can get a better idea of what is required in terms of formatting and conversion

[Jashcraf]

Totes lets do it! I'll send ya a message to set some stuff up. I'm also including below a list of notes for this feature that I thought of

Desired Header keywords

• pixel scale in pupil
• wavelength
• field of view
• raytrace filename (optional)
• polynomial decomposition fit error

Methods

• if regularly spaced in the entrance pupil, just pass those coordinates
• if not, do a zernike decomposition and report the fit error

[Jashcraf]

Adding to the headers a name for each dimension of the numpy array

[Jashcraf]

Have @kenjim21 write a demo and test file showing these things off

[Jashcraf]

@kenjim21 Just added what you'll need to get started in the most recent commit to branch 49-finalize-jones-pupil-to-fits-file-conversion

It comes with a more fleshed out poke.writing.jones_to_fits function, and more importantly the beginnings of a Demo that I was hoping you could flesh out. You can find it in poke/docs/notebooks/jones_to_fits.ipynb

It basically shows you how to load a sample Rayfront and grab its Jones pupil. This should let you play around with some data.

[kenjim21]

@Jashcraf Just finished updating the jones_to_fits function and the notebook for base demonstration. Let me know what you think of this as a start and we can build up from there!

[kenjim21]

@Jashcraf Just made some minor style changes for the demonstration notebook. Let me know what you think!

[Jashcraf]

Looks good I think we are basically ready to make a PR for this feature. There's just a couple things I would like edited:

1. realimag: defaults to true (meaning imaginary values will be saved as opposed to saving the absolute values)

Should instead be something like

1. realimag: defaults to true (meaning complex values are saved as real and imaginary components, if false, saves as absolute value and phase)

And for anything in text that is expressing something that you write in code, could you format it so that it looks like code? Example:

• Primary Header (hdul[0]):
• Primary Header (hdul[0])

Other places where this change should be made include import astropy.io, the FITS headers, and the with fits.open('sample_jones.fits') as hdul:.

After that make a PR and I think we are set :) Thanks for doing this!