Simulations

SGS

Bases: Simulations

Sequential Gaussian Simulations Solver class.

Parameters:

Name	Type	Description	Default
covariance_model	Covariance | CovarianceElem	Covariance used for distance computation.	required
neighborhood_model	Neighborhood	Neighborhood model used for weights computation.	required
axes	list[Coord]	If solver is 2D or 3D.	[U, V, W]
backend	GeostatsBackend	Computational backend to use. Defaults to RUST.	RUST
mean	float	Known Simple Kriging mean. Defaults to 0.0 (Julia C API default).	0.0
seed	int	Base RNG seed for the Rust backend. Realization i uses seed + i.	0
search_tree	SearchTrees	Spatial index used by the Rust backend.	OCTREE

Attributes:

Name	Type	Description
covariance_model	Covariance | CovarianceElem	Covariance used for distance computation.
neighborhood_model		Neighborhood model used for weights computation.
axes		If solver is 2D or 3D.

backend property

Return the computational backend.

covariance_model property

Return Covariance Model.

Returns:

Type	Description
Covariance | CovarianceElem	Covariance model.

neighborhood_model property

Return Neighborhood Model.

Returns:

Type	Description
Neighborhood	Neighborhood model.

solve(obj, obj_region, obj_attribute, support, support_region, support_attribute, simulations_number)

Solve SGS on a support using a conditioning object.

Parameters:

Name	Type	Description	Default
obj	GeoRefObject | None	Conditioning Object.	required
obj_region	str | None	Spatial subset of conditioning.	required
obj_attribute	str | None	Property to estimate.	required
support	GeoRefObject	Object to estimate onto.	required
support_region	str | None	Spatial subset of estimation.	required
support_attribute	str	Property name to add in support.	required
simulations_number	int	Number of simulations to be realized.	required

Raises:

Type	Description
Warning	Maximum allowed neighbors is higher than number of conditioning data.

solve_api(neighborhood_model, covariance_model, dim, support_coords, simulations_number, obj_coords=None, obj_data=None, backend=GeostatsBackend.RUST, mean=0.0, seed=0, search_tree=SearchTrees.OCTREE) staticmethod

Low-level SGS solving method.

Parameters:

Name	Type	Description	Default
covariance_model	Covariance | CovarianceElem	Covariance used for distance computation.	required
neighborhood_model	Neighborhood	Neighborhood model used for weights computation.	required
dim	int	Number of dimensions.	required
support_coords	ndarray	Support coordinates in region along specified axes.	required
simulations_number	int	Number of simulations to be realized.	required
obj_coords	ndarray | None	Conditioning object coordinates in region along specified axes.	None
obj_data	ndarray | None	Values in selected region.	None
backend	GeostatsBackend	Computational backend to use. Defaults to RUST.	RUST
mean	float	Known Simple Kriging mean. Defaults to 0.0.	0.0
seed	int	Base RNG seed for the Rust backend.	0
search_tree	SearchTrees	Spatial index used by the Rust backend.	OCTREE

Returns:

Type	Description
ndarray	SGS results of shape (n_targets, simulations_number).

Raises:

Type	Description
Warning	Maximum allowed neighbors is higher than number of conditioning data (Julia) or simulation targets (Rust).
ValueError	If backend is not Julia or Rust.

Simulations

Bases: Entity

Simulations Solver class.

Parameters:

Name	Type	Description	Default
covariance_model	Covariance | CovarianceElem	Covariance used for distance computation.	required
neighborhood_model	Neighborhood	Neighborhood model used for weights computation.	required
axes	list[Coord]	If solver is 2D or 3D.	[U, V, W]

Attributes:

Name	Type	Description
covariance_model	Covariance | CovarianceElem	Covariance used for distance computation.
neighborhood_model		Neighborhood model used for weights computation.
axes		If solver is 2D or 3D.

covariance_model property

Return Covariance Model.

Returns:

Type	Description
Covariance | CovarianceElem	Covariance model.

neighborhood_model property

Return Neighborhood Model.

Returns:

Type	Description
Neighborhood	Neighborhood model.

solve() abstractmethod

Use the defined method to interpolate onto target locations.