Struct LowRankMatrixRecovery 

pub struct LowRankMatrixRecovery<State = Untrained> {
    pub algorithm: RecoveryAlgorithm,
    pub rank: Option<usize>,
    pub max_iter: usize,
    pub tol: f64,
    pub lambda: f64,
    pub mu: f64,
    pub random_state: Option<u64>,
    /* private fields */
}

Low-rank Matrix Recovery algorithms for robust decomposition

These methods are designed to recover low-rank matrices from corrupted observations, handling outliers and sparse corruption better than standard matrix completion.

Fields

algorithm: RecoveryAlgorithm

Algorithm to use for recovery

rank: Option<usize>

Rank of the low-rank component (if known)

max_iter: usize

Maximum number of iterations

tol: f64

Convergence tolerance

lambda: f64

Regularization parameter for sparse component

mu: f64

Regularization parameter for nuclear norm

random_state: Option<u64>

Random state for reproducibility

Implementations

impl LowRankMatrixRecovery<Untrained>

pub fn new() -> Self

Create a new low-rank matrix recovery transformer

pub fn algorithm(self, algorithm: RecoveryAlgorithm) -> Self

Set the recovery algorithm

pub fn rank(self, rank: usize) -> Self

Set the rank (if known)

pub fn max_iter(self, max_iter: usize) -> Self

Set maximum iterations

pub fn tol(self, tol: f64) -> Self

Set tolerance

pub fn lambda(self, lambda: f64) -> Self

Set sparsity regularization parameter

pub fn mu(self, mu: f64) -> Self

Set nuclear norm regularization parameter

pub fn random_state(self, random_state: u64) -> Self

Set random state

impl LowRankMatrixRecovery<TrainedLowRankMatrixRecovery>

pub fn low_rank_component(&self) -> &Array2<f64>

Get the recovered low-rank component

pub fn sparse_component(&self) -> &Array2<f64>

Get the sparse component (outliers/corruption)

pub fn estimated_rank(&self) -> usize

Get the estimated rank

pub fn factors(&self) -> &Option<(Array2<f64>, Array1<f64>, Array2<f64>)>

Get SVD factors if available

pub fn recover_low_rank_component(&self, x: &Array2<f64>) -> Result<Array2<f64>>

Recover low-rank component from new corrupted data

pub fn reconstruct(&self) -> Array2<f64>

Reconstruct the original matrix (low-rank + sparse)

Trait Implementations

impl<State: Clone> Clone for LowRankMatrixRecovery<State>

fn clone(&self) -> LowRankMatrixRecovery<State>

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl<State: Debug> Debug for LowRankMatrixRecovery<State>

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

Formats the value using the given formatter.

impl Default for LowRankMatrixRecovery<Untrained>

fn default() -> Self

Returns the “default value” for a type.

impl Fit<ArrayBase<OwnedRepr<f64>, Dim<[usize; 2]>>, ()> for LowRankMatrixRecovery<Untrained>

type Fitted = LowRankMatrixRecovery<TrainedLowRankMatrixRecovery>

The fitted model type

fn fit(self, x: &Array2<f64>, _y: &()) -> Result<Self::Fitted>

Fit the model to the provided data with validation

fn fit_with_validation( self, x: &X, y: &Y, _x_val: Option<&X>, _y_val: Option<&Y>, ) -> Result<(Self::Fitted, FitMetrics), SklearsError>

where Self: Sized,

Fit with custom validation and early stopping

impl Transform<ArrayBase<OwnedRepr<f64>, Dim<[usize; 2]>>> for LowRankMatrixRecovery<TrainedLowRankMatrixRecovery>

fn transform(&self, x: &Array2<f64>) -> Result<Array2<f64>>

Transform the input data