Module robust 

Robust optimization with worst-case constraint reformulation.

When problem parameters are uncertain, robust optimization tightens constraints to ensure feasibility at a specified confidence level. For each inequality constraint g(x, p) <= 0, the solver determines the worst-case parameter values (within the confidence ellipsoid) and enforces g(x, p_worst) <= 0 instead.

Example

use numra_optim::robust::RobustProblem;

let result = RobustProblem::<f64>::new(1)
    .x0(&[5.0])
    .objective(|x: &[f64], _p: &[f64]| (x[0] - 5.0) * (x[0] - 5.0))
    .param("target", 5.0, 1.0)
    .solve()
    .unwrap();

Author: Moussa Leblouba Date: 8 February 2026 Modified: 2 May 2026

Structs

RobustOptions

Options for robust optimization.

RobustProblem

Declarative builder for robust optimization problems.

RobustResult

Result of robust optimization.

UncertainParam

An uncertain parameter for robust optimization.

Functions

normal_quantile

Compute the inverse of the standard normal CDF (quantile function).