petektools 0.2.8

Standalone numerics & geostatistics kernels for Rust: scattered-data gridding (minimum-curvature, IDW, nearest) and a curated numeric front-door. Pure leaf with PyO3 bindings.
Documentation 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25

//! Collocated-secondary (Markov-1) support: standardising the secondary field
//! before it is folded into each node's kriging as a collocated cokriging datum.

use ndarray::Array2;

/// Standardise a secondary field to zero mean / unit variance over its finite
/// nodes. A constant (zero-variance) field is mapped to all-zeros (it then
/// carries no drift).
pub(super) fn standardize(field: &Array2<f64>) -> Array2<f64> {
    let finite: Vec<f64> = field.iter().cloned().filter(|v| v.is_finite()).collect();
    let n = finite.len();
    if n == 0 {
        return field.clone();
    }
    let mean = finite.iter().sum::<f64>() / n as f64;
    let var = finite.iter().map(|v| (v - mean).powi(2)).sum::<f64>() / n as f64;
    let sd = var.sqrt();
    field.mapv(|v| {
        if v.is_finite() && sd > 0.0 {
            (v - mean) / sd
        } else {
            0.0
        }
    })
}