Struct ParallelPcgSolver 

pub struct ParallelPcgSolver {
    pub max_iterations: usize,
    pub tolerance: f64,
    pub abs_tolerance: f64,
}

Preconditioned Conjugate Gradient with parallel SpMV.

Uses a Jacobi (diagonal) preconditioner by default. The SpMV step uses CsrMatrix::spmv_par and the dot products use rayon::par_iter.

Fields

max_iterations: usize

Maximum number of CG iterations.

tolerance: f64

Relative residual tolerance for convergence.

abs_tolerance: f64

Minimum absolute residual (avoids infinite loop on singular systems).

Implementations

impl ParallelPcgSolver

pub fn new(max_iterations: usize, tolerance: f64) -> Self

Create a new solver with explicit parameters.

pub fn solve(&self, a: &CsrMatrix, b: &[f64], x: &mut [f64]) -> PcgStats

Solve A * x = b using Parallel PCG.

x is used as the initial guess (zero-initialise for a cold start).

Returns PcgStats with convergence information.

Trait Implementations

impl Clone for ParallelPcgSolver

fn clone(&self) -> ParallelPcgSolver

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Debug for ParallelPcgSolver

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

Formats the value using the given formatter.

impl Default for ParallelPcgSolver

fn default() -> Self

Returns the “default value” for a type.