Struct NonlinearConjugateGradient

pub struct NonlinearConjugateGradient<P, L, B, F> { /* private fields */ }

Non-linear Conjugate Gradient method

A generalization of the conjugate gradient method for nonlinear optimization problems.

Requires an initial parameter vector.

Requirements on the optimization problem

The optimization problem is required to implement CostFunction and Gradient.

Reference

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

Implementations

impl<P, L, B, F> NonlinearConjugateGradient<P, L, B, F>

where F: ArgminFloat,

pub fn new(linesearch: L, beta_method: B) -> Self

Construct a new instance of NonlinearConjugateGradient.

Takes a LineSearch and a NLCGBetaUpdate as input.

Example

let nlcg: NonlinearConjugateGradient<Vec<f64>, _, _, f64> =
    NonlinearConjugateGradient::new(linesearch, beta_method);

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

Specify the number of iterations after which a restart should be performed.

This allows the algorithm to “forget” previous information which may not be helpful anymore.

Example

let nlcg = nlcg.restart_iters(100);

pub fn restart_orthogonality(self, v: F) -> Self

Set the value for the orthogonality measure.

Setting this parameter leads to a restart of the algorithm (setting beta = 0) after consecutive search directions stop being orthogonal. In other words, if this condition is met:

|\nabla f_k^T * \nabla f_{k-1}| / | \nabla f_k |^2 >= v

A typical value for v is 0.1.

Example

let nlcg = nlcg.restart_orthogonality(0.1);

Trait Implementations

impl<P: Clone, L: Clone, B: Clone, F: Clone> Clone for NonlinearConjugateGradient<P, L, B, F>

fn clone(&self) -> NonlinearConjugateGradient<P, L, B, F>

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl<'de, P, L, B, F> Deserialize<'de> for NonlinearConjugateGradient<P, L, B, F>

where P: Deserialize<'de>, L: Deserialize<'de>, B: 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, L, B, F> Serialize for NonlinearConjugateGradient<P, L, B, F>

where P: Serialize, L: Serialize, B: 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<O, P, G, L, B, F> Solver<O, IterState<P, G, (), (), (), F>> for NonlinearConjugateGradient<P, L, B, F>

where O: CostFunction<Param = P, Output = F> + Gradient<Param = P, Gradient = G>, P: Clone + ArgminAdd<P, P> + ArgminMul<F, P>, G: Clone + ArgminMul<F, P> + ArgminDot<G, F> + ArgminL2Norm<F>, L: Clone + LineSearch<P, F> + Solver<O, IterState<P, G, (), (), (), F>>, B: NLCGBetaUpdate<G, P, F>, 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, G, (), (), (), F>, ) -> Result<(IterState<P, G, (), (), (), F>, Option<KV>), Error>

Initializes the algorithm.

fn next_iter( &mut self, problem: &mut Problem<O>, state: IterState<P, G, (), (), (), F>, ) -> Result<(IterState<P, G, (), (), (), 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_internal(&mut self, state: &I) -> TerminationStatus

Checks whether basic termination reasons apply.

fn terminate(&mut self, _state: &I) -> TerminationStatus

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