Optimize reconstructions

[talonchandler]

Replaced by #525. Close this when #525 merges

This PR enables optimizing reconstructions using torch gradients with respect to:

• NA_illumination + NA_detection
• defocus offset
• illumination tilts
• detection aberrations (via Zernike modes)

These optimization have been trialed in the following reconstruction modes:

• 3D-to-2D phase, and
• 3D widefield fluoresence.

Quickstart:

• on the HPC with access to a screen, install from this branch with pip install waveorder[all] and run python ./waveorder/docs/examples/visuals/optimize_phase_recon.py
• script will start a short 10-iteration optimization loop. Try modifying NUM_ITERATIONS and OPTIMIZABLE_PARAMS.
• view traces with tensorboard --logdir runs

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 36.59%. Comparing base (57a2004) to head (fd7c79e).

Additional details and impacted files

@@           Coverage Diff           @@
##             main     #455   +/-   ##
=======================================
  Coverage   36.59%   36.59%           
=======================================
  Files          49       49           
  Lines       10224    10225    +1     
=======================================
+ Hits         3741     3742    +1     
  Misses       6483     6483           

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

🚀 New features to boost your workflow:

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