Struct SparseRandomProjection 

pub struct SparseRandomProjection<F> { /* private fields */ }

Sparse random projection transformer.

Projects data into a lower-dimensional space using a sparse random matrix with entries {-1, 0, +1} drawn with probabilities {d/2, 1 - d, d/2}, scaled by sqrt(1 / (d * n_components)).

The default density d = 1 / sqrt(n_features) is used when not specified.

Examples

use ferrolearn_preprocess::random_projection::SparseRandomProjection;
use ferrolearn_core::traits::{Fit, Transform};
use ndarray::Array2;

let x = Array2::<f64>::ones((10, 100));
let proj = SparseRandomProjection::<f64>::new(5).random_state(42);
let fitted = proj.fit(&x, &()).unwrap();
let out = fitted.transform(&x).unwrap();
assert_eq!(out.shape(), &[10, 5]);

Implementations

impl<F: Float + Send + Sync + 'static> SparseRandomProjection<F>

pub fn new(n_components: usize) -> Self

Create a new sparse random projection with an explicit n_components number of output dimensions (NComponents::Fixed).

pub fn new_auto() -> Self

Create a new sparse random projection with n_components='auto' (sklearn default): resolved at fit time from the Johnson-Lindenstrauss bound for the configured eps (random_projection.py:388-391).

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

Set the density of non-zero entries.

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

Set the distortion-rate eps used by the Auto JL bound.

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

Set the random seed for reproducibility.

Trait Implementations

impl<F: Clone> Clone for SparseRandomProjection<F>

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

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl<F: Debug> Debug for SparseRandomProjection<F>

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

Formats the value using the given formatter.

impl<F: Float + Send + Sync + 'static> Fit<ArrayBase<OwnedRepr<F>, Dim<[usize; 2]>>, ()> for SparseRandomProjection<F>

fn fit( &self, x: &Array2<F>, _y: &(), ) -> Result<FittedSparseRandomProjection<F>, FerroError>

Fit the projection by generating a sparse random matrix.

When n_components is NComponents::Auto the target dimension is resolved from the Johnson-Lindenstrauss bound for eps (random_projection.py:388-391).

Errors

Returns FerroError::InvalidParameter if the resolved n_components is 0 or density is not in (0, 1]. Returns FerroError::InsufficientSamples if x has zero rows.

type Fitted = FittedSparseRandomProjection<F>

The fitted model type returned by fit.

type Error = FerroError

The error type returned by fit.

impl<F: Float + Send + Sync + 'static> FitTransform<ArrayBase<OwnedRepr<F>, Dim<[usize; 2]>>> for SparseRandomProjection<F>

fn fit_transform(&self, x: &Array2<F>) -> Result<Array2<F>, FerroError>

Fit and transform in one step.

type FitError = FerroError

The error type for the combined fit-transform operation.

impl<F: Float + Send + Sync + 'static> PipelineTransformer<F> for SparseRandomProjection<F>

fn fit_pipeline( &self, x: &Array2<F>, _y: &Array1<F>, ) -> Result<Box<dyn FittedPipelineTransformer<F>>, FerroError>

Fit this transformer on the given data.

impl<F: Float + Send + Sync + 'static> Transform<ArrayBase<OwnedRepr<F>, Dim<[usize; 2]>>> for SparseRandomProjection<F>

fn transform(&self, _x: &Array2<F>) -> Result<Array2<F>, FerroError>

Always returns an error — the projection must be fitted first.

type Output = ArrayBase<OwnedRepr<F>, Dim<[usize; 2]>>

The transformed output type.

type Error = FerroError

The error type returned by transform.