DeclusteringParameters

evo.compute.tasks.geostatistics.declustering.DeclusteringParameters

Parameters for the declustering task.

Computes grid-based declustering weights by measuring each sample's influence on evaluation locations.

The power parameter controls the estimation mode:

power=2.0 (default) — inverse-distance weighting (IDW)
power=None — arithmetic-mean KNN (equal neighbour weights)

Example::

>>> params = DeclusteringParameters(
...     source=pointset,
...     grid=regular_grid,
...     target=Target.new_attribute(pointset, "weights"),
...     neighborhood=SearchNeighborhood(
...         ellipsoid=Ellipsoid(ranges=EllipsoidRanges(200, 150, 100)),
...         max_samples=20,
...     ),
... )

source

source: AnyDeclusteringSource

Source sample locations (accepts a geoscience object directly).

grid

grid: AnyDeclusteringGrid

Evaluation grid for measuring sample influence (accepts a geoscience object directly).

target

target: AnyTargetAttribute

Target object and attribute to write declustering weights to.

neighborhood

neighborhood: SearchNeighborhood

Search neighborhood parameters.

power

power: float | None = Field(default=2.0, gt=0.0)

Inverse-distance weighting power.

Controls how strongly distance affects influence: higher values give nearer samples more weight. Set to None to use KNN mode (arithmetic-mean, equal neighbour votes).

Default is 2.0 (standard IDW). Must be positive when provided.