Crate numra_optim 

Optimization algorithms for Numra.

Algorithms

• Bfgs – BFGS quasi-Newton method (unconstrained)
• Lbfgs – L-BFGS limited-memory quasi-Newton method (unconstrained)
• lm_minimize – Levenberg-Marquardt nonlinear least squares
• lbfgsb_minimize – L-BFGS-B bound-constrained optimizer
• augmented_lagrangian_minimize – Augmented Lagrangian constrained optimizer
• OptimProblem – Declarative problem builder with auto solver selection

Author: Moussa Leblouba Date: 5 March 2026 Modified: 2 May 2026

Re-exports

pub use augmented_lagrangian::augmented_lagrangian_minimize;

pub use augmented_lagrangian::AugLagOptions;

pub use auto::SolverChoice;

pub use bfgs::bfgs_minimize;

pub use bfgs::Bfgs;

pub use cmaes::cmaes_minimize;

pub use cmaes::CmaEsOptions;

pub use derivative_free::nelder_mead;

pub use derivative_free::powell;

pub use derivative_free::NelderMeadOptions;

pub use derivative_free::PowellOptions;

pub use error::OptimError;

pub use global::de_minimize;

pub use global::DEOptions;

pub use lbfgs::lbfgs_minimize;

pub use lbfgs::Lbfgs;

pub use lbfgs::LbfgsOptions;

pub use lbfgsb::lbfgsb_minimize;

pub use lbfgsb::LbfgsBOptions;

pub use levenberg_marquardt::lm_minimize;

pub use levenberg_marquardt::LmOptions;

pub use lp::simplex_solve;

pub use lp::LPOptions;

pub use milp::milp_solve;

pub use milp::MILPOptions;

pub use multiobjective::nsga2_optimize;

pub use multiobjective::NsgaIIOptions;

pub use optim_sensitivity::compute_param_sensitivity;

pub use problem::Constraint;

pub use problem::ConstraintKind;

pub use problem::LinearConstraint;

pub use problem::ObjectiveKind;

pub use problem::OptimProblem;

pub use problem::ProblemHint;

pub use problem::VarType;

pub use qp::active_set_qp_solve;

pub use qp::QPOptions;

pub use robust::RobustOptions;

pub use robust::RobustProblem;

pub use robust::RobustResult;

pub use robust::UncertainParam;

pub use sqp::sqp_minimize;

pub use sqp::SqpOptions;

pub use stochastic::StochasticOptions;

pub use stochastic::StochasticParam;

pub use stochastic::StochasticProblem;

pub use stochastic::StochasticResult;

pub use types::IterationRecord;

pub use types::OptimOptions;

pub use types::OptimResult;

pub use types::OptimStatus;

pub use types::ParamSensitivity;

pub use types::ParetoPoint;

pub use types::ParetoResult;

Modules

augmented_lagrangian

Augmented Lagrangian method for constrained optimization.

auto

Automatic solver selection for optimization problems.

bfgs

BFGS quasi-Newton optimizer.

cmaes

CMA-ES (Covariance Matrix Adaptation Evolution Strategy).

derivative_free

Derivative-free optimization methods: Nelder-Mead and Powell.

error

Error types for optimization algorithms.

global

Global optimization via Differential Evolution (DE/rand/1/bin).

lbfgs

L-BFGS (Limited-memory BFGS) optimizer.

lbfgsb

L-BFGS-B: bound-constrained limited-memory BFGS.

levenberg_marquardt

Levenberg-Marquardt algorithm for nonlinear least squares.

lp

Revised Simplex method for Linear Programming.

milp

Branch-and-Bound solver for Mixed-Integer Linear Programs (MILP).

multiobjective

NSGA-II multi-objective optimizer.

optim_sensitivity

Parametric sensitivity analysis for optimization problems.

problem

Declarative optimization problem builder.

qp

Active Set method for convex Quadratic Programming.

robust

Robust optimization with worst-case constraint reformulation.

sqp

Sequential Quadratic Programming (SQP) for constrained nonlinear optimization.

stochastic

Stochastic optimization via Sample-Average Approximation (SAA).

types

Common types for optimization algorithms.