Accept FITS big-endian byte order

[bryce-wedig]

This PR allows JAXtronomy to handle raw FITS data which uses big-endian byte order.

While modeling JWST strong lenses with dolphin, I got the following error:

TypeError: Dtype >f8 is not a valid JAX array type. Only arrays of numeric types are supported by JAX.

The fix, dtype=float, is a no-op for regular little-endian arrays but will byte-swap big-endian arrays so JAX can handle them.

[bryce-wedig]

The failing CI is due to Astropy v8.0.0 which was released just yesterday. Pinning astropy<8.0.0 should work for now, but obviously that's a temporary fix and lenstronomy will need to be patched to be compatible. Please let me know how you want to deal with this.

[ahuang314]

@bryce-wedig Thanks for this PR, setting astropy<8.0.0 is fine for now, we can address that later

[sibirrer]

@bryce-wedig @ahuang314 I am making a PR in lenstronomy aimed at fixing this: lenstronomy/lenstronomy#868

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 99.98%. Comparing base (676b9f3) to head (73b730b).

Additional details and impacted files

@@           Coverage Diff           @@
##             main      #98   +/-   ##
=======================================
  Coverage   99.98%   99.98%           
=======================================
  Files          80       80           
  Lines        7330     7331    +1     
=======================================
+ Hits         7329     7330    +1     
  Misses          1        1           

Files with missing lines	Coverage Δ
jaxtronomy/Data/image_noise.py	100.00% <100.00%> (ø)
jaxtronomy/Data/imaging_data.py	100.00% <100.00%> (ø)

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[ahuang314]

looks good to me