Struct argmin::solver::trustregion::Steihaug

pub struct Steihaug<P, F> { /* private fields */ }

Steihaug method

The Steihaug method is a conjugate gradients based approach for finding an approximate solution to the second order approximation of the cost function within the trust region.

Reference

Jorge Nocedal and Stephen J. Wright (2006). Numerical Optimization. Springer. ISBN 0-387-30303-0.

Implementations

impl<P, F> Steihaug<P, F>

where P: ArgminMul<F, P> + ArgminDot<P, F> + ArgminAdd<P, P>, F: ArgminFloat,

pub fn new() -> Self

Construct a new instance of Steihaug

Example

let sh: Steihaug<Vec<f64>, f64> = Steihaug::new();

pub fn with_epsilon(self, epsilon: F) -> Result<Self, Error>

Set epsilon

The algorithm stops when the residual is smaller than epsilon.

Must be larger than 0 and defaults to 10^-10.

Example

let sh: Steihaug<Vec<f64>, f64> = Steihaug::new().with_epsilon(10e-9)?;

pub fn with_max_iters(self, iters: u64) -> Self

Set maximum number of iterations

The algorithm stops after iter iterations.

Defaults to u64::MAX.

Example

let sh: Steihaug<Vec<f64>, f64> = Steihaug::new().with_max_iters(100);

Trait Implementations

impl<P: Clone, F: Clone> Clone for Steihaug<P, F>

fn clone(&self) -> Steihaug<P, F>

Returns a copy of the value.

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

Performs copy-assignment from source.

impl<P: Default, F: Default> Default for Steihaug<P, F>

fn default() -> Steihaug<P, F>

Returns the “default value” for a type.

impl<'de, P, F> Deserialize<'de> for Steihaug<P, F>

where P: Deserialize<'de>, F: Deserialize<'de>,

fn deserialize<__D>(__deserializer: __D) -> Result<Self, __D::Error>

where __D: Deserializer<'de>,

Deserialize this value from the given Serde deserializer.

impl<P, F> Serialize for Steihaug<P, F>

where P: Serialize, F: Serialize,

fn serialize<__S>(&self, __serializer: __S) -> Result<__S::Ok, __S::Error>

where __S: Serializer,

Serialize this value into the given Serde serializer.

impl<P, O, F, H> Solver<O, IterState<P, P, (), H, (), F>> for Steihaug<P, F>

where P: Clone + ArgminMul<F, P> + ArgminL2Norm<F> + ArgminDot<P, F> + ArgminAdd<P, P> + ArgminZeroLike, H: ArgminDot<P, P>, F: ArgminFloat,

fn name(&self) -> &str

Name of the solver. Mainly used in Observers.

fn init( &mut self, _problem: &mut Problem<O>, state: IterState<P, P, (), H, (), F>, ) -> Result<(IterState<P, P, (), H, (), F>, Option<KV>), Error>

Initializes the algorithm.

fn next_iter( &mut self, _problem: &mut Problem<O>, state: IterState<P, P, (), H, (), F>, ) -> Result<(IterState<P, P, (), H, (), F>, Option<KV>), Error>

Computes a single iteration of the algorithm and has access to the optimization problem definition and the internal state of the solver. Returns an updated state and optionally a KV which holds key-value pairs used in Observers.

fn terminate( &mut self, state: &IterState<P, P, (), H, (), F>, ) -> TerminationStatus

Used to implement stopping criteria, in particular criteria which are not covered by (terminate_internal.

fn terminate_internal(&mut self, state: &I) -> TerminationStatus

Checks whether basic termination reasons apply.

impl<P, F: ArgminFloat> TrustRegionRadius<F> for Steihaug<P, F>

fn set_radius(&mut self, radius: F)

Set current radius.

Needed by TrustRegion.

Example

use argmin::solver::trustregion::{Steihaug, TrustRegionRadius};
let mut sh: Steihaug<Vec<f64>, f64> = Steihaug::new();
sh.set_radius(0.8);