Struct RANSACRegressor 

pub struct RANSACRegressor<F, E> {
    pub estimator: E,
    pub min_samples: Option<usize>,
    pub residual_threshold: Option<F>,
    pub max_trials: usize,
    pub random_state: Option<u64>,
}

RANSAC robust regression meta-estimator.

Wraps a base regressor (e.g., LinearRegression) and repeatedly fits it on random subsets to find a model robust to outliers.

Type Parameters

• F: The floating-point type (f32 or f64).
• E: The base estimator type.

Fields

estimator: E

The base estimator.

min_samples: Option<usize>

Minimum number of samples for fitting.

residual_threshold: Option<F>

Residual threshold: points with absolute residual below this are considered inliers. If None, uses the MAD of the target.

max_trials: usize

Maximum number of random trials.

random_state: Option<u64>

Optional random seed for reproducibility.

Implementations

impl<F: Float, E> RANSACRegressor<F, E>

pub fn new(estimator: E) -> Self

Create a new RANSACRegressor with the given base estimator.

Defaults: min_samples = None (auto: n_features + 1), residual_threshold = None (auto: MAD), max_trials = 100, random_state = None.

pub fn with_min_samples(self, min_samples: usize) -> Self

Set the minimum number of samples for fitting.

pub fn with_residual_threshold(self, threshold: F) -> Self

Set the residual threshold for inlier detection.

pub fn with_max_trials(self, max_trials: usize) -> Self

Set the maximum number of random trials.

pub fn with_random_state(self, seed: u64) -> Self

Set the random seed for reproducibility.

Trait Implementations

impl<F: Clone, E: Clone> Clone for RANSACRegressor<F, E>

fn clone(&self) -> RANSACRegressor<F, E>

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl<F: Debug, E: Debug> Debug for RANSACRegressor<F, E>

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

Formats the value using the given formatter.

impl<F, E, Ef> Fit<ArrayBase<OwnedRepr<F>, Dim<[usize; 2]>>, ArrayBase<OwnedRepr<F>, Dim<[usize; 1]>>> for RANSACRegressor<F, E>

where F: Float + Send + Sync + ScalarOperand + FromPrimitive + 'static, E: Fit<Array2<F>, Array1<F>, Fitted = Ef, Error = FerroError> + Clone, Ef: Predict<Array2<F>, Output = Array1<F>, Error = FerroError> + Clone,

fn fit( &self, x: &Array2<F>, y: &Array1<F>, ) -> Result<FittedRANSACRegressor<E::Fitted>, FerroError>

Fit the RANSAC model by repeatedly sampling and fitting.

Errors

Returns FerroError::ShapeMismatch if x and y have different sample counts. Returns FerroError::ConvergenceFailure if no valid model is found after max_trials iterations.

type Fitted = FittedRANSACRegressor<Ef>

The fitted model type returned by fit.

type Error = FerroError

The error type returned by fit.