Struct Gaussian 

pub struct Gaussian { /* private fields */ }

The SGS propagation spec (ps.gaussian(...)): the spatial-continuity model, the RNG seed, the moving-neighbourhood search, and an optional collocated secondary (trend cokriging, Phase 3).

The variogram is fitted in data space; SGS transforms to normal scores internally (petektools NormalScore), so its total sill need not be 1.

Implementations

impl Gaussian

pub fn new(variogram: Variogram, seed: u64) -> Self

A plain (no-secondary) SGS spec from a data-space variogram and a seed. The search neighbourhood defaults to a bounded window (DEFAULT_MAX_NEIGHBOURS nodes within a variogram-range/lattice-scaled radius) — seconds on real-scale lattices where the old whole-grid default took >15 min/cube. Beyond a covariance range a conditioning node’s kriging weight is ~0, so the bounded result matches the unbounded one within simulation tolerance.

pub fn with_search(self, max_neighbours: usize, radius_m: f64) -> Self

Override the moving-neighbourhood search: at most max_neighbours conditioning nodes within radius_m (world units).

pub fn with_unbounded_search(self) -> Self

Restore the legacy unbounded (whole-grid) search window — every previously-simulated node in the lattice is a search candidate. Pathologically slow on real-scale lattices; kept as an explicit opt-in for small grids or exact reproduction of pre-bounded-default results.

pub fn allow_mean_fill(self) -> Self

Opt into filling a simulated layer that carries no conditioning data with the conditioned mean (a flat layer) rather than erroring — for a model where a data-less layer is expected and a structureless mean-fill is acceptable. The default errors loudly instead (task_petekstatic_canonical_fixes item 4).

pub fn with_trend(self, trend: TrendSurface, corr: f64) -> Self

Steer the simulation with a collocated secondary variable (trend cokriging, Markov-1): the TrendSurface is resampled to the model areal lattice and folded into each node’s kriging at correlation corr. corr == 0 recovers plain SGS bit-for-bit; corr -> 1 pulls the field toward the trend pattern.

Trait Implementations

impl Clone for Gaussian

fn clone(&self) -> Gaussian

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Debug for Gaussian

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

Formats the value using the given formatter.