Fix youngAge mutual exclusivity in makeMutuallyExclusive() and fireSenseCovariatesCreate()

[eliotmcintire]

makeMutuallyExclusive() (R/makeMutuallyExclusive.R:19-28) recomputed the rows to zero from the key column inside its pattern loop, so once a pattern matched the key column itself youngAge got zeroed instead of staying 1, and every later pattern then found no rows left to zero. fireSenseCovariatesCreate() (R/fuelClassPrep.R) also applied this exclusivity step before youngAge was finalized for non-forest pixels, so those pixels kept their nfLCC_* land-cover value, and treedWetland was never cleared for young pixels at all.

The fix computes the rows to zero once per key column before any zeroing, excludes the key column from its own pattern matches, and moves the exclusivity step in fireSenseCovariatesCreate() to run after youngAge and treedWetland are both final. A new exported helper, youngAgeExclusiveCols(), lets fireSense_SpreadPredict derive the identical rule the fit uses instead of keeping a second copy.

Verified with new failing-first tests, matching the symptom seen on real ELF 14.3 cached inputs: before the fix, 4624 pixels had youngAge = 1 going into the fit's objective, and 0 after the (broken) exclusivity step. devtools::load_all() + testthat::test_dir(): 291 tests / 770 expectations pass on development, 296 tests / 790 expectations pass on this branch (0 failures either way; the new/changed tests fail on development and pass here). R CMD check --no-tests is clean (0 errors, 0 warnings, 0 notes). Version bumped to 0.2.3.9047.

Merge this before PredictiveEcology/fireSense_spreadFit#46 and PredictiveEcology/fireSense_spreadPredict#21, which both floor their fireSenseUtils dependency on this version.

🤖 Generated with Claude Code

https://claude.ai/code/session_01CwcjqqK59FmTJscyi7xUqv

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 44.35%. Comparing base (e9ff8ba) to head (b48ac45).
⚠️ Report is 1 commits behind head on development.

Additional details and impacted files

@@               Coverage Diff               @@
##           development      #92      +/-   ##
===============================================
+ Coverage        42.96%   44.35%   +1.38%     
===============================================
  Files               40       40              
  Lines             3798     3803       +5     
===============================================
+ Hits              1632     1687      +55     
+ Misses            2166     2116      -50     

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

🚀 New features to boost your workflow:

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