lmi-solver-rs

LMI (Linear Matrix Inequality) solver using the Ellipsoid Method.

This crate provides an implementation of an LMI feasibility oracle that can be used with the ellipsoid method from ellalgo-rs to solve optimization problems with Linear Matrix Inequality constraints.

Modules

• [ldlt_mgr] - LDL^T factorization manager for positive definiteness checking
• [lmi_oracle] - Linear Matrix Inequality (LMI) feasibility oracle

Usage

use ellalgo_rs::arr::Arr;
use ellalgo_rs::cutting_plane::OracleFeas;
use lmi_solver_rs::lmi_oracle::LMIOracle;
use ndarray::Array2;

let f = vec![
    Array2::from_shape_vec((2, 2), vec![-7.0, -11.0, -11.0, 3.0]).unwrap(),
    Array2::from_shape_vec((2, 2), vec![7.0, -18.0, -18.0, 8.0]).unwrap(),
    Array2::from_shape_vec((2, 2), vec![-2.0, -8.0, -8.0, 1.0]).unwrap(),
];
let b = Array2::from_shape_vec((2, 2), vec![33.0, -9.0, -9.0, 26.0]).unwrap();
let mut oracle = LMIOracle::new(f, b);
let x = Arr::from(vec![1.0, 1.0, 1.0]);
let cut = oracle.assess_feas(&x);
assert!(cut.is_some());