Struct GaussianConditioner 

pub struct GaussianConditioner { /* private fields */ }

Reusable factorization of a symmetric positive-definite linear system.

The constructor accepts a dense matrix in row-major order, optionally adds a fixed non-negative jitter value to its diagonal, and computes a Cholesky factorization. Solves reuse that factorization and never form an explicit matrix inverse.

Implementations

impl GaussianConditioner

pub fn new( matrix: &[f64], dimension: usize, jitter: f64, ) -> Result<Self, ConditioningError>

Construct a Gaussian conditioning system from a flattened square matrix.

matrix is interpreted in row-major order. The supplied jitter is added exactly once to each diagonal entry before factorization.

Errors

Returns ConditioningError when dimensions are inconsistent, values are non-finite, jitter is negative, or Cholesky factorization fails.

pub const fn dimension(&self) -> usize

Return the system dimension.

pub const fn jitter(&self) -> f64

Return the fixed diagonal jitter used during factorization.

pub fn solve(&self, rhs: &[f64]) -> Result<Vec<f64>, ConditioningError>

Solve K x = rhs using the stored Cholesky factorization.

Errors

Returns ConditioningError when the right-hand side has the wrong dimension or contains non-finite entries.

Trait Implementations

impl Clone for GaussianConditioner

fn clone(&self) -> GaussianConditioner

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Debug for GaussianConditioner

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

Formats the value using the given formatter.