Struct ProjectedGradient 

pub struct ProjectedGradient {
    pub learning_rate: f64,
    pub max_iter: usize,
    pub tol: f64,
    pub lb: Vec<f64>,
    pub ub: Vec<f64>,
}

Projected gradient descent with box constraints lb ≤ x ≤ ub.

Fields

learning_rate: f64

Fixed learning rate.

max_iter: usize

Maximum number of iterations.

tol: f64

Convergence tolerance.

lb: Vec<f64>

Lower bound per coordinate.

ub: Vec<f64>

Upper bound per coordinate.

Implementations

impl ProjectedGradient

pub fn new( lr: f64, max_iter: usize, tol: f64, lb: Vec<f64>, ub: Vec<f64>, ) -> Self

Create a new projected gradient optimizer.

pub fn project(&self, x: &[f64]) -> Vec<f64>

Project x onto the box [lb, ub].

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

Minimize f starting from x0 with box-constraint projection.

Returns (x_star, f(x_star)).

Trait Implementations

impl Clone for ProjectedGradient

fn clone(&self) -> ProjectedGradient

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Debug for ProjectedGradient

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

Formats the value using the given formatter.