Struct SolverRouter 

pub struct SolverRouter { /* private fields */ }

Stateless algorithm selector.

Given a SparsityProfile and a QueryType, the router walks a decision tree (documented in the module-level docs) to pick the Algorithm with the best expected cost.

Example

use ruvector_solver::router::{SolverRouter, RouterConfig};
use ruvector_solver::types::{Algorithm, QueryType, SparsityProfile};

let router = SolverRouter::new(RouterConfig::default());
let profile = SparsityProfile {
    rows: 500,
    cols: 500,
    nnz: 1200,
    density: 0.0048,
    is_diag_dominant: true,
    estimated_spectral_radius: 0.4,
    estimated_condition: 10.0,
    is_symmetric_structure: true,
    avg_nnz_per_row: 2.4,
    max_nnz_per_row: 5,
};

let algo = router.select_algorithm(&profile, &QueryType::LinearSystem);
assert_eq!(algo, Algorithm::Neumann);

Implementations

impl SolverRouter

pub fn new(config: RouterConfig) -> Self

Create a new router with the provided configuration.

pub fn config(&self) -> &RouterConfig

Return a shared reference to the active configuration.

pub fn select_algorithm( &self, profile: &SparsityProfile, query: &QueryType, ) -> Algorithm

Select the optimal algorithm for the given matrix profile and query.

This is a pure function with no side effects – it does not touch the matrix data, only the precomputed profile.

Trait Implementations

impl Clone for SolverRouter

fn clone(&self) -> SolverRouter

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Debug for SolverRouter

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

Formats the value using the given formatter.

impl Default for SolverRouter

fn default() -> Self

Returns the “default value” for a type.