Struct recless::Rls

pub struct Rls<F> { /* fields omitted */ }

The parameters of recursive least squares algorithm.

This struct contains all parameters involved in a recursive least squares algorithm with exponential forgetting. For a review see [Haykin's Adaptive Filter Theory][http://www.isbnsearch.org/isbn/9780132671453]. The implementation here does not implicitly take time into account. By making a choice of the forgetting factor λ < 1 and shifting down old values of the input vector manually, the user can get this algorithm to behave accordingly.

Methods

impl<F: NdFloat> Rls<F>

fn new(initialization_factor: F, forgetting_factor: F, n: usize) -> Self

Constructs a new Rls object with initialization factor δ and a weight vector of length n.

fn with_weight(
    initialization_factor: F,
    forgetting_factor: F,
    weight: Array1<F>
) -> Self

Constructs a new Rls object with initialization factor δ and pre-defined weight w.

impl Rls<f32>

fn update<S>(&mut self, input: &ArrayBase<S, Ix1>, target: f32) where
    S: Data<Elem = f32>, 

Performs a recursive update of inverse correlation matrix and weight vector.

impl Rls<f64>

fn update<S>(&mut self, input: &ArrayBase<S, Ix1>, target: f64) where
    S: Data<Elem = f64>, 

Performs a recursive update of inverse correlation matrix and weight vector.

impl<T> Rls<T>

fn gain_ref(&self) -> &Array1<T>

Returns a reference to the gain vector.

fn inverse_correlation_ref(&self) -> &Array2<T>

Returns a reference to the inverse correlation matrix.

fn inv_forgetting_factor_ref(&self) -> &T

Returns a reference to the inverse forgetting factor.

fn weight_ref(&self) -> &Array1<T>

Returns a reference to the (tap) weight vector.

fn prior_error_ref(&self) -> &T

Returns a refernce to the prior error.

Trait Implementations

impl<F: Clone> Clone for Rls<F>

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

Returns a copy of the value.

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

Performs copy-assignment from source.

impl<F: Debug> Debug for Rls<F>

fn fmt(&self, __arg_0: &mut Formatter) -> Result

Formats the value using the given formatter.