geoutils.operators.interpolator.Kriging

class geoutils.operators.interpolator.Kriging(model, *, backend='gstools', neighborhood=None, max_overlap=None, exact=True, pseudo_inverse=True, cache_size=256)

The Kriging interpolator predicts values from nearby observations and their spatial correlation.

Pass a fitted variogram to describe how correlation changes with distance, then use this interpolator in Raster.interp_points() or PointCloud.grid(). The backend selects GSTools or GPyTorch for ordinary kriging; neighborhood and max_overlap control which nearby values are used.

We reuse weights when points have the same positions relative to the requested coordinate (common in rasters). coefficients() returns those weights, and kriging_variance() gives the prediction variance from the model.

Parameters:

• model (Variogram | VariogramModel) – Fitted GeoUtils variogram or its VariogramModel representation.

• backend (Literal['gstools', 'gpytorch']) – Library used to solve the kriging system, "gstools" or "gpytorch".

• neighborhood (PointNeighbours | GridNeighbours | None) – Points or raster cells considered for each prediction. By default, select points/cells within the model’s longest effective range (the distance over which its values are correlated).

• max_overlap (float | None) – Optional distance limit for the effective range and the extra data read around chunk edges.

• exact (bool) – Include the nugget when predicting at a known source location.

• pseudo_inverse (bool) – Let GSTools use a pseudo inverse for redundant or coincident observations.

• cache_size (int) – Maximum number of arrangements of nearby points for which calculated weights are stored.

__init__(model, *, backend='gstools', neighborhood=None, max_overlap=None, exact=True, pseudo_inverse=True, cache_size=256)

Check the covariance model, search radius, selected library, and cache size.

Methods

__init__(model, *[, backend, neighborhood, ...])	Check the covariance model, search radius, selected library, and cache size.
coefficients(data)	Return ordinary kriging weights in the same order as the supplied source values.
evaluate(data, *[, nodata_propagation])	Check nodata and weights before predicting a value.
kriging_variance(data)	Return the model interpolation variance for the same source points used by coefficients().
predict(data)	Predict a value from the nearby source observations.
predict_batch(data, *[, nodata_propagation])	Predict several values by calling evaluate() for each group of source observations.
validate_geospatial_support(coordinate_count)	Check that the variogram uses all coordinates included in the neighbour search.

Attributes

accepts_sample_weights
default_neighborhood
default_nodata_propagation
error_structure
interpolation_order
minimum_inputs