Prepare for GPU support

[tyler-a-cox]

finufft is preparing to release a version of the type 3 transform on GPUs by the end of the month. This PR adds the utility functions required to run fftvis on GPUs once the type 3 transform is released later this month. In the past, we've tried to keep the code mostly similar to matvis with the core measurement equation computation replaced with a nufft. However, cufinufft uses cupy instead of pycuda, so we'll need to diverge from matvis a little. We should still be able to keep the functionality mostly the same though.

[codecov-commenter]

⚠️ Please install the to ensure uploads and comments are reliably processed by Codecov.

Codecov Report

Attention: Patch coverage is 0% with 54 lines in your changes missing coverage. Please review.

Project coverage is 63.52%. Comparing base (f0e5142) to head (0827abd).

Files	Patch %	Lines
fftvis/gpu.py	0.00%	54 Missing ⚠️

❗ Your organization needs to install the Codecov GitHub app to enable full functionality.

Additional details and impacted files

@@             Coverage Diff             @@
##             main      #22       +/-   ##
===========================================
- Coverage   80.59%   63.52%   -17.07%     
===========================================
  Files           5        6        +1     
  Lines         201      255       +54     
  Branches       47       58       +11     
===========================================
  Hits          162      162               
- Misses         31       85       +54     
  Partials        8        8               

Flag	Coverage Δ
unittests	63.52% <0.00%> (-17.07%)	⬇️

Flags with carried forward coverage won't be shown. Click here to find out more.

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[steven-murray]

@tyler-a-cox we should think about this together a bit -- I have a version of matvis that is done in cupy, which does seem nicer than pycuda. I haven't merged it yet, but we could try to align our goals a bit here.