Struct PCA 

pub struct PCA { /* private fields */ }

Principal Component Analysis (PCA), mirroring sklearn.decomposition.PCA.

Centers the data, computes the covariance matrix, and projects onto the top eigenvectors found via Jacobi eigenvalue decomposition.

Examples

use datarust::matrix::Matrix;
use datarust::decomposition::{PCA, PCAComponents};
use datarust::Transformer;

let x = Matrix::new(vec![vec![1.0, 2.0], vec![3.0, 4.0], vec![5.0, 6.0]]).unwrap();
let mut pca = PCA::new(PCAComponents::Count(1));
let projected = pca.fit_transform(&x).unwrap();
// projects 2D data onto 1 principal component

Implementations

impl PCA

pub fn new(n_components: PCAComponents) -> Self

Creates a new PCA with the given component selection.

pub fn whiten(self, b: bool) -> Self

Sets whether to whiten the projected components.

pub fn solver(self, s: PCASolver) -> Self

Selects the decomposition backend used by fit.

PCASolver::Auto (the default) currently uses the exact eigensolver; select PCASolver::Randomized explicitly for the randomized backend.

pub fn components(&self) -> &[Vec<f64>]

Returns the principal axes (one row per component).

pub fn explained_variance(&self) -> &[f64]

Returns the variance explained by each kept component.

pub fn explained_variance_ratio(&self) -> &[f64]

Returns the fraction of total variance explained by each kept component.

pub fn noise_variance(&self) -> f64

Estimated noise variance, computed as the mean of discarded eigenvalues. Returns 0 when all components are kept.

pub fn mean(&self) -> &[f64]

Returns the per-feature mean estimated during fit.

pub fn n_components(&self) -> usize

Returns the number of components kept after fit.

Trait Implementations

impl Clone for PCA

fn clone(&self) -> PCA

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Debug for PCA

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

Formats the value using the given formatter.

impl Default for PCA

Default: keep 95% of variance, no whitening, auto solver.

fn default() -> Self

Returns the “default value” for a type.

impl<'de> Deserialize<'de> for PCA

fn deserialize<__D>(__deserializer: __D) -> Result<Self, __D::Error>

where __D: Deserializer<'de>,

Deserialize this value from the given Serde deserializer.

impl Estimator for PCA

impl FeatureNames for PCA

fn feature_names_out(&self, _input_features: Option<&[String]>) -> Vec<String>

Returns the output feature names given the optional input feature names.

impl Serialize for PCA

fn serialize<__S>(&self, __serializer: __S) -> Result<__S::Ok, __S::Error>

where __S: Serializer,

Serialize this value into the given Serde serializer.

impl Transformer for PCA

fn name(&self) -> &'static str

Name of the transformer, used for diagnostics.

fn fit(&mut self, x: &Matrix) -> Result<()>

Fit the transformer on training data.

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

Transform data using fitted parameters.

fn inverse_transform(&self, projected: &Matrix) -> Result<Matrix>

Reverse the transformation, recovering an approximation of the original input. Not all transformers support this; the default implementation returns an error.

fn is_fitted(&self) -> bool

Whether the transformer has been fitted.

fn fit_transform(&mut self, x: &Matrix) -> Result<Matrix>

Convenience: fit then transform.

fn fit_with_target(&mut self, x: &Matrix, _y: &[f64]) -> Result<()>

Fit the transformer with supervised target values when they are available. Unsupervised transformers ignore y by default. Supervised feature selectors override this method, allowing them to live safely in a SupervisedPipeline.