Multiple BC Fix

[lamkina]

Purpose

This PR addresses a bug where adding multiple BCs of the same type didn't work as expected. Previously, only the first BC of a given type would be properly added and tracked, but then any subsequent BC of the same type would be ignored. This resulted in incorrect functionals and derivatives with zero seeds.

Using the suggested fixes from PR #323 this should now be fixed. I added a new mesh to input_files and tests to make sure this is working correctly.

Expected time until merged

1-3 weeks

Type of change

• Bugfix (non-breaking change which fixes an issue)
• New feature (non-breaking change which adds functionality)
• Breaking change (non-backwards-compatible fix or feature)
• Code style update (formatting, renaming)
• Refactoring (no functional changes, no API changes)
• Documentation update
• Maintenance update
• Other (please describe)

Testing

Run testflo tests/reg_tests/test_multiple_bc.py to test the changes. You can also run the complex tests in the same file to verify the derivatives are working as expected.

Checklist

• I have run flake8 and black to make sure the Python code adheres to PEP-8 and is consistently formatted
• I have formatted the Fortran code with fprettify or C/C++ code with clang-format as applicable
• I have run unit and regression tests which pass locally with my changes
• I have added new tests that prove my fix is effective or that my feature works
• I have added necessary documentation

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 27.79%. Comparing base (4a0f01c) to head (54108f9).

Additional details and impacted files

@@             Coverage Diff             @@
##             main     #354       +/-   ##
===========================================
- Coverage   41.45%   27.79%   -13.67%     
===========================================
  Files          13       13               
  Lines        4069     4069               
===========================================
- Hits         1687     1131      -556     
- Misses       2382     2938      +556     

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

[lamkina]

@sseraj @eirikurj The failing tests are passing locally so I'm not sure what is happening. This specific test is just checking the totals of BC integrated functions computed with the adjoint against the totals that are saved in the training data. Locally it matches to 1e-12 absolute and relative tolerance, but on the images it's at worst getting 1e-6 absolute and 1e-5 relative.

Any debugging tips on what might be happening?

[sseraj]

I have had tests fail like this when the flow converges fine locally but does not fully converge on the Docker images. I would check that the flow and adjoint both converge to the same level on the Docker images as on your local machine. You might have to increase nCycles