GEOPY-1889: Return nan on predicted where data is nan

[domfournier]

GEOPY-1889 - Return nan on predicted where data is nan

[Copilot]

🟡 Changes recommended

Scalar DC/IP normalizations cause the new masking assignment to raise a TypeError.

1 open finding

• Broadcast normalization before boolean mask assignment · New

What changed in this PR

Ensures saved predicted data preserves NaN positions from observed data.

Changes:

• Masks prediction transforms using observed-data NaNs.
• Adds gravity and magnetotelluric regression coverage.
• Updates magnetotelluric workspace names and expected metrics.

File	Description
directives_factory.py	Adds NaN masking to saved data transforms.
driver_grav_test.py	Expects one NaN in predicted gravity data.
driver_mt_test.py	Tests NaN propagation and updates workspace paths.

🧠 Review effort: Balanced

---

💡 Add a code-review agent skill or configure MCP servers for context-aware, tailored reviews. Learn more in the docs.

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 90.51%. Comparing base (f6929c3) to head (0966590).

Additional details and impacted files

@@           Coverage Diff            @@
##           develop     #493   +/-   ##
========================================
  Coverage    90.51%   90.51%           
========================================
  Files          111      111           
  Lines         7134     7137    +3     
  Branches       886      887    +1     
========================================
+ Hits          6457     6460    +3     
  Misses         456      456           
  Partials       221      221           

Files with missing lines	Coverage Δ
simpeg_drivers/components/data.py	95.91% <ø> (-0.09%)	⬇️
...drivers/components/factories/directives_factory.py	95.88% <100.00%> (+0.11%)	⬆️

🚀 New features to boost your workflow:

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