Struct CoordinateDescentOptimizer

pub struct CoordinateDescentOptimizer<F: Float> {
    pub max_iter: usize,
    pub tol: F,
    pub random: bool,
}

Coordinate Descent optimizer for separable problems

Fields

max_iter: usize

Maximum number of iterations

tol: F

Convergence tolerance

random: bool

Whether to use random coordinate selection

Implementations

impl<F: Float + ScalarOperand> CoordinateDescentOptimizer<F>

pub fn new() -> Self

Create a new coordinate descent optimizer

pub fn random() -> Self

Create a new random coordinate descent optimizer

pub fn minimize<Obj, Grad1D>( &self, obj_fn: Obj, grad_1d_fn: Grad1D, x0: &Array1<F>, ) -> Result<OptimizeResult<F>, OptimizeError>

where Obj: Fn(&ArrayView1<'_, F>) -> F, Grad1D: Fn(&ArrayView1<'_, F>, usize) -> F, F: Into<f64> + Copy,

Minimize using coordinate descent

Trait Implementations

impl<F: Clone + Float> Clone for CoordinateDescentOptimizer<F>

fn clone(&self) -> CoordinateDescentOptimizer<F>

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl<F: Debug + Float> Debug for CoordinateDescentOptimizer<F>

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

Formats the value using the given formatter.

impl<F: Float + ScalarOperand> Default for CoordinateDescentOptimizer<F>

fn default() -> Self

Returns the “default value” for a type.