9 releases
0.2.11 | Mar 2, 2025 |
---|---|
0.2.10 | Feb 1, 2025 |
0.2.9 | Jan 9, 2025 |
0.2.7 | Dec 21, 2024 |
0.2.4 | Oct 30, 2024 |
#162 in Math
43,324 downloads per month
Used in 57 crates
(2 directly)
175KB
4K
SLoC
microlp
This is a fork of the archived minilp crate, which was made to fix some bugs, add features and allow the community to make issues and PRs.
⚠️ Warning ⚠️
If you were using the library prior to 0.2.11, please use the latest version of the library as there was a major bug for integer variables.
A fast linear programming solver library.
Linear programming is a technique for finding the minimum (or maximum) of a linear function of a set of variables subject to linear equality and inequality constraints.
Getting started
You can use microlp on its own, but it's recommended to use it with goodlp or with rooc modeling language as it makes it easier to create models. Look at the examples below on how to use microlp on its own.
Features
- Pure Rust implementation.
- Able to solve problems with hundreds of thousands of variables and constraints.
- Incremental: add constraints to an existing solution without solving it from scratch.
- Problems can be defined via an API or parsed from an MPS file.
- Allows for continuous, integer and boolean variables
Warning: this is an early-stage project. Although the library is already quite powerful and fast, it will probably cycle, lose precision or panic on some harder problems. Please report bugs and contribute code! Models with integer or binary variables are solved using a simple branch & bound method.
Examples
Basic usage
use microlp::{Problem, OptimizationDirection, ComparisonOp};
// Maximize an objective function x + 2 * y of two continuous variables x >= 0 and 0 <= y <= 3
let mut problem = Problem::new(OptimizationDirection::Maximize);
let x = problem.add_var(1.0, (0.0, f64::INFINITY));
let y = problem.add_var(2.0, (0.0, 3.0));
// subject to constraints: x + y <= 4 and 2 * x + y >= 2.
problem.add_constraint(&[(x, 1.0), (y, 1.0)], ComparisonOp::Le, 4.0);
problem.add_constraint(&[(x, 2.0), (y, 1.0)], ComparisonOp::Ge, 2.0);
// Optimal value is 7, achieved at x = 1 and y = 3.
let solution = problem.solve().unwrap();
assert_eq!(solution.objective(), 7.0);
assert_eq!(solution[x], 1.0);
assert_eq!(solution[y], 3.0);
For a more involved example, see examples/tsp, a solver for the travelling salesman problem.
License
This project is licensed under the Apache License, Version 2.0.
Dependencies
~2MB
~45K SLoC