Struct ConvexFunction 

pub struct ConvexFunction {
    pub name: String,
    pub lipschitz_grad: Option<f64>,
    pub strong_convexity: Option<f64>,
    /* private fields */
}

Represents a convex function f : ℝ^n → ℝ with optional gradient.

Fields

name: String

Name of the function.

lipschitz_grad: Option<f64>

Lipschitz constant of gradient (if known).

strong_convexity: Option<f64>

Strong convexity modulus (if known).

Implementations

impl ConvexFunction

pub fn new( name: impl Into<String>, eval_fn: fn(&[f64]) -> f64, grad_fn: Option<fn(&[f64]) -> Vec<f64>>, ) -> Self

Construct a new convex function.

pub fn eval(&self, x: &[f64]) -> f64

Evaluate f at x.

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

Compute gradient of f at x (finite differences if no exact gradient provided).

pub fn fenchel_conjugate_approx( &self, y: &[f64], radius: f64, steps: usize, ) -> f64

Compute the Fenchel conjugate f*(y) = sup_x {⟨y,x⟩ - f(x)} approximately via grid search on [-R, R]^n with resolution steps.

pub fn in_epigraph(&self, x: &[f64], t: f64) -> bool

Compute epigraph membership: (x, t) ∈ epi(f) iff f(x) ≤ t.

pub fn in_level_set(&self, x: &[f64], alpha: f64) -> bool

Compute level set membership: x ∈ lev_α(f) iff f(x) ≤ α.

Trait Implementations

impl Clone for ConvexFunction

fn clone(&self) -> ConvexFunction

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Debug for ConvexFunction

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

Formats the value using the given formatter.