Form wsclean sub-band cubes, convol and then linmos

[tjgalvin]

wsclean deals with the wide-bandwidth of visibility data by imaging some number of sub-channel intervals. In a previous pull request these sub-band images were all concatenated together into a single cube - which introduced a modest reduction in the number of files on disk throughout processing.

This pull request extends that to convolve them to a common resolution across each sub-band image, and then linmos the set of spectral cubes together. The convolution is performed using the racs-tools beamcon_3D script, and is currently configured to convolve each channel to the best common resolution in isolation of others.

Tests have been added along the way.

[codecov-commenter]

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

Codecov Report

Attention: Patch coverage is 58.15217% with 77 lines in your changes missing coverage. Please review.

Project coverage is 59.54%. Comparing base (d59d357) to head (7eea24d).

Files with missing lines	Patch %	Lines
flint/convol.py	39.62%	32 Missing ⚠️
flint/prefect/common/imaging.py	20.00%	32 Missing ⚠️
flint/prefect/flows/continuum_pipeline.py	10.00%	9 Missing ⚠️
flint/coadd/linmos.py	93.54%	4 Missing ⚠️

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

Additional details and impacted files

@@            Coverage Diff             @@
##             main     #173      +/-   ##
==========================================
+ Coverage   59.17%   59.54%   +0.37%     
==========================================
  Files          33       33              
  Lines        5536     5686     +150     
==========================================
+ Hits         3276     3386     +110     
- Misses       2260     2300      +40     

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