Struct OrdinaryKriging 

pub struct OrdinaryKriging { /* private fields */ }

Ordinary Kriging interpolator

Ordinary Kriging assumes the mean is unknown but constant within a local neighborhood. It provides the Best Linear Unbiased Estimator (BLUE) for spatial data.

Implementations

impl OrdinaryKriging

pub fn new( points: &ArrayView2<'_, f64>, values: &ArrayView1<'_, f64>, variogram: VariogramModel, ) -> SpatialResult<Self>

Create a new Ordinary Kriging interpolator

Arguments

• points - Array of point coordinates, shape (n_points, ndim)
• values - Array of values at points, shape (n_points,)
• variogram - Variogram model to use

Returns

• New OrdinaryKriging instance

Examples

use scirs2_spatial::kriging::{OrdinaryKriging, VariogramModel};
use ndarray::array;

let points = array![[0.0, 0.0], [1.0, 0.0], [0.0, 1.0]];
let values = array![1.0, 2.0, 3.0];
let variogram = VariogramModel::spherical(1.0, 0.5, 0.1);

let kriging = OrdinaryKriging::new(&points.view(), &values.view(), variogram).unwrap();

pub fn fit(&mut self) -> SpatialResult<()>

Fit the Kriging model by precomputing the covariance matrix inverse

This step is optional but recommended for multiple predictions as it avoids recomputing the matrix inverse each time.

pub fn predict(&self, location: &[f64]) -> SpatialResult<KrigingPrediction>

Predict value at a new location

Arguments

• location - Point where to predict, shape (ndim,)

Returns

• KrigingPrediction with value, variance, and weights

Examples

use scirs2_spatial::kriging::{OrdinaryKriging, VariogramModel};
use ndarray::array;

let points = array![[0.0, 0.0], [1.0, 0.0], [0.0, 1.0], [1.0, 1.0]];
let values = array![1.0, 2.0, 3.0, 4.0];
let variogram = VariogramModel::spherical(1.5, 1.0, 0.1);

let mut kriging = OrdinaryKriging::new(&points.view(), &values.view(), variogram).unwrap();
kriging.fit().unwrap();

let prediction = kriging.predict(&[0.5, 0.5]).unwrap();
println!("Predicted: {:.3} ± {:.3}", prediction.value, prediction.variance.sqrt());

pub fn predict_batch( &self, locations: &ArrayView2<'_, f64>, ) -> SpatialResult<Vec<KrigingPrediction>>

Predict values at multiple locations efficiently

Arguments

• locations - Array of locations, shape (n_locations, ndim)

Returns

• Vector of KrigingPrediction results

pub fn variogram(&self) -> &VariogramModel

Get the variogram model

pub fn n_points(&self) -> usize

Get the number of data points

pub fn points(&self) -> &Array2<f64>

Get the data points

pub fn values(&self) -> &Array1<f64>

Get the data values

pub fn cross_validate(&self) -> SpatialResult<Array1<f64>>

Cross-validation: leave-one-out prediction errors

Returns

• Array of prediction errors (predicted - actual)

Trait Implementations

impl Clone for OrdinaryKriging

fn clone(&self) -> OrdinaryKriging

Returns a duplicate of the value.

fn clone_from(&mut self, source: &Self)

Performs copy-assignment from source.

impl Debug for OrdinaryKriging

fn fmt(&self, f: &mut Formatter<'_>) -> Result

Formats the value using the given formatter.