🎨 Control coil enforcement, constraint stabilisation and general cleanup

[je-cook]

Description

• Fixes some bugs where control coils were not explicitly selected in parts of the equilibria optimisation problems.
• Added rounding to field and force constraints (to 16 dp) which seems to stabilise them in local testing. I have it on good authority that the condition matrix of force constraint is infinite therefore very unstable. This seems to help a bit.
• Some general cleanup around properties on component managers

This was done while doing some work towards passive coils

Interface Changes

Checklist

I confirm that I have completed the following checks:

• Tests run locally and pass pytest tests --reactor
• Code quality checks run locally and pass pre-commit run --from-ref develop --to-ref HEAD
• Documentation built locally and checked sphinx-build -W documentation/source documentation/build

[codecov]

Codecov Report

Attention: Patch coverage is 83.15789% with 16 lines in your changes missing coverage. Please review.

Project coverage is 76.40%. Comparing base (60d6be1) to head (6dfb5d9).
Report is 3 commits behind head on develop.

Files with missing lines	Patch %	Lines
bluemira/equilibria/run.py	30.43%	16 Missing ⚠️

Additional details and impacted files

@@             Coverage Diff             @@
##           develop    #3518      +/-   ##
===========================================
+ Coverage    76.38%   76.40%   +0.01%     
===========================================
  Files          230      230              
  Lines        26677    26700      +23     
===========================================
+ Hits         20378    20399      +21     
- Misses        6299     6301       +2     

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

[github-actions]

⚠️ Warning Report

Found 0 new warnings, 1 fixed warning. 🎉

All warnings (6)

On collect

• /home/runner/miniconda3/envs/bluemira/lib/python3.10/site-packages/ufl/core/ufl_type.py:56: DeprecationWarning: attach_operators_from_hash_data deprecated, please use UFLObject instead.

On runtest

• /home/runner/miniconda3/envs/bluemira/lib/python3.10/site-packages/ffcx/element_interface.py:23: DeprecationWarning: Use of elements created by UFL is deprecated. You should create elements directly using Basix.
• /home/runner/miniconda3/envs/bluemira/lib/python3.10/site-packages/ffcx/element_interface.py:26: DeprecationWarning: Converting elements created in UFL to Basix elements is deprecated. You should create the elements directly using basix.ufl.element instead
• /home/runner/miniconda3/envs/bluemira/lib/python3.10/site-packages/ufl/core/expr.py:275: DeprecationWarning: Expr.ufl_domain() is deprecated, please use extract_unique_domain(expr) instead.
• /home/runner/miniconda3/envs/bluemira/lib/python3.10/site-packages/basix/ufl.py:1909: DeprecationWarning: Converting elements created in UFL to Basix elements is deprecated. You should create the elements directly using basix.ufl.element instead
• /home/runner/work/bluemira/bluemira/bluemira/radiation_transport/radiation_tools.py:731: DeprecationWarning: interp2dis deprecated!interp2d` is deprecated in SciPy 1.10 and will be removed in SciPy 1.12.0.

For legacy code, nearly bug-for-bug compatible replacements are
RectBivariateSpline on regular grids, and bisplrep/bisplev for
scattered 2D data.

In new code, for regular grids use RegularGridInterpolator instead.
For scattered data, prefer LinearNDInterpolator or
CloughTocher2DInterpolator.

For more details see
https://gist.github.com/ev-br/8544371b40f414b7eaf3fe6217209bff
`

[oliverfunk]

I've gone through it quite carefully.

Two things:

• All the spherical harmioncs stuff is using sig_fig_round, where the constraints use np.round. They obviously do different things, but I find the inconsistency slightly worrying.

• The 15 sig figs used in the SH stuff should be a param.

[sonarcloud]

Quality Gate passed

Issues
0 New issues
0 Accepted issues

Measures
0 Security Hotspots
0.0% Coverage on New Code
0.0% Duplication on New Code

See analysis details on SonarCloud

[geograham]

Been through the changes, all looks sensible and ready to merge.

[oliverfunk]

Okay lgtm!

Perpahs in some doc. it could be stated that rounding constraint return values is generally recommended for improving numeric stability.

As a recommendation for constraint implementers.

[je-cook]

As we talked about I think its very constraint specific for those constraints that are numerically unstable. As you said we could do something as part of a wider refactor of the eq opt stuff