Feature/issue 217 snow depth

[Raakshass]

Pull Request

Description

Adds snow depth as a predictor feature to improve solar forecast accuracy during winter conditions.

Snow on PV panels reduces output through:

• Reflection: Snow reflects incoming radiation
• Physical coverage: Snow blocks panel surface
• Albedo effect: Ground snow can slightly increase diffuse radiation (positive effect)

Changes Made:

1. open_meteo.py: Added snow_depth to Open-Meteo API variables
2. v3_lightgbm.py: Added _add_snow_features() method with:
   • snow_depth: Raw value in meters (NaN → 0)
   • has_snow: Binary indicator (0/1)
   • snow_impact: Normalized factor (0-1, capped at 0.5m)
3. test_v3_lightgbm.py: Added 7 unit tests for snow feature engineering

The implementation is backward compatible - defaults to 0 if snow data is unavailable.

Fixes #217

How Has This Been Tested?

• Yes - Python syntax validation (py_compile) passes for all 3 modified files
• Yes - 7 new unit tests in TestSnowFeatures class cover:
  • Snow features added with data
  • Values with/without snow
  • Impact capped at 1.0 for heavy snow
  • NaN handling
  • Backward compatibility when snow_depth missing

Checklist:

• My code follows OCF's coding style guidelines
• I have performed a self-review of my own code
• I have made corresponding changes to the documentation
• I have added tests that prove my fix is effective or that my feature works
• I have checked my code and corrected any misspellings