[1.0.0] Interpolation

[brendancol]

1.0.0 readiness: Interpolation

Area: Stats Functions: idw, kriging, spline
Source: xrspatial/interpolate/

Scope

• Public surface frozen against the README feature matrix
• Dark-corner decisions resolved (see [1.0.0] Scope freeze: reconcile dark corners #2930)

Sweep clearance, no CRITICAL or HIGH (accuracy and security first)

• Accuracy /deep-sweep xrspatial/interpolate
• Security
• Performance
• API consistency

Last swept: none yet

Examples

• Notebook example
• (optional) standalone script

Linked bug issues

• kriging(): nugget lands on the kriging-matrix diagonal (should be gamma(0)=0) #2915 (High) kriging(): nugget lands on the kriging-matrix diagonal (should be gamma(0)=0)
• Style: fix flake8 E128 and isort drift in interpolate/_kriging.py #2916 (Low) Style: fix flake8 E128 and isort drift in interpolate/_kriging.py
• kriging() raises an opaque numpy error for a single input point #2917 (Low) kriging() raises an opaque numpy error for a single input point
• idw(): GPU backends (cupy, dask+cupy) have no test coverage #2919 (Medium) idw(): GPU backends (cupy, dask+cupy) have no test coverage
• kriging() crashes on single-point input with "zero-size array to reduction maximum" #2920 (Medium) kriging() crashes on single-point input with "zero-size array to reduction maximum"
• Test coverage gaps in interpolate/_kriging.py (backend, params, edge cases) #2921 (Medium) Test coverage gaps in interpolate/_kriging.py (backend, params, edge cases)
• kriging() memory guard raises spurious MemoryError on dask-backed templates #2923 (High) kriging() memory guard raises spurious MemoryError on dask-backed templates