i.aster.toar: add test file

[jayneel-shah18]

This PR adds a comprehensive test suite for the i.aster.toar module.

The test suite focuses on regression sensitivity and minimal test execution, ensuring that any change in computation results is detected with high precision while keeping the number of i.aster.toar executions as low as possible. It also includes a theoretical validation based on the reflectance correction with respect to solar zenith angle.

Tests added

• Radiance mode (-r) for thermal bands (bands 10–14)
  • Validates against strict univariate statistics.
  • Checks that output is floating-point and within a reasonable radiance range.
• High gain mode (-a -b) for bands 1–9
  • Regression check with precomputed min, max, sum statistics.
• Low gain mode (-c -d -e) for bands 1–9
  • Regression check with precomputed statistics.
• Theoretical test: reflectance vs. sun elevation
  • Verifies that reflectance scales according to:

    reflectance ∝ 1 / cos(90° - sun_elevation)
    

  • Compares outputs from runs with sun_elevation = 45° and 60°.
  • Expected ratio: cos(45°)/cos(30°) ≈ 0.8165.

Performance

• macOs: 2.919s
• Windows: 5.005s
• Ubuntu: 3.966s

Looking forward to feedback and potential refinements!

[echoix]

The code itself looks fine, nothing wrong stood out. I find it interesting to validate the ratio of two runs to check the proportion.

[echoix]

It will probably show a failure for macOS, which all PRs have since yesterday, blocking all merges.

I'd also like someone like Anna or other to look at more the GIS side of the tests, if they want. Otherwise, after a few business days, I'll merge it anyways

[echoix]

Since there wasn't any other objections on the GIS side of the tests in a week and a half, I don't see any problem with this. We are better with a test than without.