Struct SolverOrchestrator 

pub struct SolverOrchestrator { /* private fields */ }

High-level solver facade that combines routing with execution.

Owns a SolverRouter and delegates to the appropriate solver backend. Provides a solve_with_fallback method that automatically retries with progressively more robust (but slower) algorithms when the first choice fails.

Example

use ruvector_solver::router::{SolverOrchestrator, RouterConfig};
use ruvector_solver::types::{ComputeBudget, CsrMatrix, QueryType};

let orchestrator = SolverOrchestrator::new(RouterConfig::default());

let matrix = CsrMatrix::<f64>::from_coo(3, 3, vec![
    (0, 0, 2.0), (0, 1, -0.5),
    (1, 0, -0.5), (1, 1, 2.0), (1, 2, -0.5),
    (2, 1, -0.5), (2, 2, 2.0),
]);
let rhs = vec![1.0, 0.0, 1.0];
let budget = ComputeBudget::default();

let result = orchestrator
    .solve(&matrix, &rhs, QueryType::LinearSystem, &budget)
    .unwrap();
assert!(result.residual_norm < 1e-6);

Implementations

impl SolverOrchestrator

pub fn new(config: RouterConfig) -> Self

Create a new orchestrator with the provided routing configuration.

pub fn router(&self) -> &SolverRouter

Return a reference to the inner router.

pub fn solve( &self, matrix: &CsrMatrix<f64>, rhs: &[f64], query: QueryType, budget: &ComputeBudget, ) -> Result<SolverResult, SolverError>

Auto-select the best algorithm and solve Ax = b.

Analyses the sparsity profile of matrix, routes to the best algorithm via SolverRouter::select_algorithm, and dispatches.

Errors

Returns SolverError if the selected solver fails (e.g. non-convergence, dimension mismatch, numerical instability).

pub fn solve_with_fallback( &self, matrix: &CsrMatrix<f64>, rhs: &[f64], query: QueryType, budget: &ComputeBudget, ) -> Result<SolverResult, SolverError>

Solve with a deterministic fallback chain.

Tries the routed algorithm first. On failure, falls back through:

1. Selected algorithm (from routing)
2. CG (robust iterative)
3. Dense (direct, always works for small systems)

Each step is only attempted if the previous one returned an error.

Errors

Returns the error from the last fallback attempt if all fail.

pub fn estimate_complexity( &self, matrix: &CsrMatrix<f64>, query: &QueryType, ) -> ComplexityEstimate

Estimate the computational complexity of solving with the routed algorithm, without actually solving.

Useful for admission control, cost estimation, or deciding whether to batch multiple queries.

pub fn analyze_sparsity(matrix: &CsrMatrix<f64>) -> SparsityProfile

Analyse the sparsity profile of a CSR matrix.

Performs a single O(nnz) pass over the matrix to compute structural and numerical properties used by the router. This is intentionally cheap so it can be called on every solve request.

Trait Implementations

impl Clone for SolverOrchestrator

fn clone(&self) -> SolverOrchestrator

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Debug for SolverOrchestrator

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

Formats the value using the given formatter.

impl Default for SolverOrchestrator

fn default() -> Self

Returns the “default value” for a type.