calibratePowerLaw(): gamma and p_c50 of the powerLaw link from an ELF's flammable map

[eliotmcintire]

Adds a way to set the spread model's one fixed shape number for each ELF from that ELF's own landscape, instead of one value for all ELFs. Nothing changes until fireSense_spreadFit calls it (PredictiveEcology/fireSense_spreadFit, same change).

Affects: nothing existing; new functions and constants only.

Technical details

gamma of the "powerLaw" link was the fixed spreadPowerLawGamma (1.9, R/constants.R:116), although it is confounded with the coefficient scale and depends on the landscape. New calibratePowerLaw() (R/calibratePowerLaw.R) spreads fires with SpaDES.tools::spreadCpp() as the fit and fireSense_burn do (8 directions, escapeSizePixels(), jumpTries, jumpMeanDist) at constant p on the flammable pixels from random starts in the largest connected patch, finds p_c50 where half the fires burn at least 25% of that patch (isotonic regression, interpolated), and gives gamma as the slope of log mean ha on log(p_c50 - p) below p_c50, with R2 and the per-p table. Results are set by the seed only (one L'Ecuyer-CMRG stream per fire), not by cores, and the caller's RNG state is restored. New powerLawPcBounds() (floor + 0.01 to p_c50 + margin; spreadPowerLawPcBounds stays as the outer limits), spreadPowerLawPGrid, and ledger column names spreadFitPowerLawPc50Txt, spreadFitPowerLawR2Txt, spreadFitPowerLawPcMarginTxt. On a real ELF mask (14.3, 100 fires per p, 16 cores, 80 s) it gave p_c50 0.264, gamma 3.08, R2 0.98, against 0.265 / 3.25 from the original GAM scripts. Tests: test-calibratePowerLaw.R (uniform map p_c50 near 0.25, breaks raise it, seed and cores, RNG state, error cases, bounds). Full suite 2168 passed, 0 failed (base 2150); R CMD check --no-manual --no-build-vignettes (scratch library): 0 errors, 0 notes, 2 vignette warnings that come only from skipping the vignette build.

🤖 Generated with Claude Code

https://claude.ai/code/session_014G18Mik3fH4z5M5CyMSUS6

[codecov]

Codecov Report

❌ Patch coverage is 89.83051% with 6 lines in your changes missing coverage. Please review.
✅ Project coverage is 89.48%. Comparing base (bd81487) to head (a5db849).

Files with missing lines	Patch %	Lines
R/calibratePowerLaw.R	89.83%	6 Missing ⚠️

Additional details and impacted files

@@             Coverage Diff              @@
##           development     #168   +/-   ##
============================================
  Coverage        89.47%   89.48%           
============================================
  Files               51       52    +1     
  Lines             4466     4525   +59     
============================================
+ Hits              3996     4049   +53     
- Misses             470      476    +6     

☔ 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.