Fill in NaN for missing EK80 coefficients [all tests ci]

[leewujung]

This PR change EK80 set_vendor to fill in np.nan for missing filter coeffs across all channels.

Currently we have below in our test file:

• D20210330-T123857.raw that does not contain any filter coeffs
• three_ensemble-Phase0-D20240506-T053349-0.raw that only contains filter coeffs from 1 of the 2 channels, which seems to be due to ping sequencing. This file is from Error storing filter coefficients for sequenced ek80 file #1317.
  • [Data conversion] ValueError when attempting to import raw EK80 data #1413 seems to be due to a similar reason.

This addresses #1317 and #1413.

[leewujung]

The failed tests do not appear to be related to the changes made in this PR. I'll check soon.

[ctuguinay]

Yeah, I tried one of the failed tests with an older scipy version and it passed, but after upgrading to the latest scipy that test failed.

[codecov-commenter]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 85.33%. Comparing base (9f56124) to head (a5a92f2).
Report is 186 commits behind head on main.

Additional details and impacted files

@@            Coverage Diff             @@
##             main    #1458      +/-   ##
==========================================
+ Coverage   83.52%   85.33%   +1.81%     
==========================================
  Files          64       72       +8     
  Lines        5686     6579     +893     
==========================================
+ Hits         4749     5614     +865     
- Misses        937      965      +28     

Flag	Coverage Δ
unittests	85.33% <100.00%> (+1.81%)	⬆️

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.

[ctuguinay]

So the conversion runs well, but when trying to calibrate the three ensemble raw file I'm getting the following error:

I'ma grab some lunch and investigate this.

[ctuguinay]

I pushed something small to cast numpy float arrays to integers when indexing.

[ctuguinay]

@leewujung I think what you have looks good. Let me know what you think of these small changes I made.

[leewujung]

@ctuguinay : Thanks for the changes and the added test! Great that you caught this indexing problem. I had forgotten that having nan would force the array to be float.

The only thing I have is github warns that there's no @pytest.mark.test -- I can't find it anywhere either. Was this a typo?

[ctuguinay]

Oh yes my bad that was for singling out a single test. Let me remove that and squash merge this. It was hidden at line 619: