Struct BundleMethodSolver 

pub struct BundleMethodSolver {
    pub mu: f64,
    pub m_l: f64,
    pub max_bundle_size: usize,
    pub max_iter: usize,
    pub tol: f64,
}

Bundle method for nonsmooth convex optimization.

Maintains a bundle of subgradients and uses a stability center with a proximal term to regularize the QP subproblem. Supports serious steps (descent) and null steps (bundle update only).

Fields

mu: f64

Proximal parameter μ > 0 (controls step aggressiveness).

m_l: f64

Descent tolerance m_L ∈ (0, 1) for serious step acceptance.

max_bundle_size: usize

Maximum bundle size (older cuts are dropped when exceeded).

max_iter: usize

Maximum iterations.

tol: f64

Convergence tolerance.

Implementations

impl BundleMethodSolver

pub fn new( mu: f64, m_l: f64, max_bundle_size: usize, max_iter: usize, tol: f64, ) -> Self

Create a new bundle method solver.

pub fn minimize<F: ConvexFunction>( &self, f: &F, x0: &[f64], ) -> (Vec<f64>, f64, usize)

Solve min f(x) starting from x0. Returns (x_star, f(x_star), iterations).

Trait Implementations

impl Clone for BundleMethodSolver

fn clone(&self) -> BundleMethodSolver

Returns a duplicate of the value.

fn clone_from(&mut self, source: &Self)

Performs copy-assignment from source.

impl Debug for BundleMethodSolver

fn fmt(&self, f: &mut Formatter<'_>) -> Result

Formats the value using the given formatter.