pyinterp.rtree.kriging

Contents

pyinterp.rtree.kriging#

pyinterp.rtree.kriging(tree, coordinates, config=None, *, k=None, covariance_model=None, sigma=None, lambda_=None, nugget=None, drift_function=None, radius=None, boundary_check=None, num_threads=None)[source]#

Kriging interpolation.

Parameters:

  • tree (core.RTree3DHolder | core.geometry.geographic.RTree) – R*Tree containing the scattered data.

  • coordinates (np.ndarray[TwoDims, np.dtype]) –

    Query coordinates as a NumPy array with shape:

    • (N, 2): For geographic.RTree, representing (longitude, latitude).

    • (N, 2) or (N, 3): For RTree3DHolder, representing (longitude, latitude[, altitude]).

  • config (rtree.Kriging | None) – Configuration object (if provided, keyword args are ignored)

  • k (int | None) – Number of nearest neighbors to use

  • covariance_model (CovarianceFunction | None) – Covariance function type

  • sigma (float | None) – Variance parameter

  • lambda – Length scale parameter

  • nugget (float | None) – Nugget effect

  • drift_function (DriftFunction | None) – Drift/trend function

  • radius (float | None) – Search radius (None = no limit)

  • boundary_check (BoundaryCheck | None) – Boundary checking mode

  • num_threads (int | None) – Number of threads (0 = auto)

Returns:

Tuple of (interpolated values, neighbor counts)

Return type:

tuple[np.ndarray[OneDim, np.dtype], NDArray1DUInt32]

Examples

>>> result, counts = kriging(
...     tree, coords, k=10, covariance_model="matern_32"
... )
Contents