numra-ode

ODE and DAE solvers for the Numra workspace — with forward sensitivity, event detection, and automatic stiffness handling.

Solves initial-value problems dy/dt = f(t, y), y(t0) = y0 with explicit Runge-Kutta, implicit Runge-Kutta, and multistep methods. Includes forward sensitivity analysis, DAE consistent-initialization, structural index reduction, event handling, and parameter-uncertainty propagation.

Example

use numra_ode::{DoPri5, OdeProblem, Solver, SolverOptions};

// Lorenz system
let problem = OdeProblem::new(
    |_t, y: &[f64], dydt: &mut [f64]| {
        let (sigma, rho, beta) = (10.0, 28.0, 8.0 / 3.0);
        dydt[0] = sigma * (y[1] - y[0]);
        dydt[1] = y[0] * (rho - y[2]) - y[1];
        dydt[2] = y[0] * y[1] - beta * y[2];
    },
    0.0, 20.0, vec![1.0, 1.0, 1.0],
);

let opts = SolverOptions::default().rtol(1e-8);
let result = DoPri5::solve(&problem, 0.0, 20.0, &[1.0, 1.0, 1.0], &opts).unwrap();

What's in this crate

Solver families:

Additional capabilities:

• Forward sensitivity: ParametricOdeSystem, solve_forward_sensitivity, SensitivityResult
• DAEs: DaeProblem, compute_consistent_initial, structural index reduction (reduce_dae_problem, analyze_dae_index)
• Events: zero-crossing detection with state resets (bouncing_ball-style)
• Uncertainty: UncertainParam, solve_with_uncertainty, solve_monte_carlo
• Diagnostics: SolverStats on every result (accepted/rejected steps, function/Jacobian/LU evaluations)

Composes with

ODE trajectories from this crate flow into:

• numra-interp for cubic-spline / Akima resampling of the trajectory onto arbitrary points
• numra-integrate for quadrature of functions defined on the trajectory
• numra-fft for spectral analysis of the solution
• numra-signal for peak detection and filtering on the solution
• numra-ocp for parameter estimation of ODE models against measured data

See interop workflows for verified end-to-end examples.

Install

[dependencies]
numra-ode = "0.1"

Or pull in the whole Numra workspace via the umbrella crate:

[dependencies]
numra = "0.1"

Documentation

• API: https://docs.rs/numra-ode
• Book: Solving ODEs · Stiff systems
• Source: https://github.com/moussaoutlook/numra-rs/tree/main/numra-ode

License

Numra Academic & Research License (Non-Commercial). Academic and research use is free; commercial use requires a separate license — contact contact@spectralautomata.com. See LICENSE.