Struct KernelPCA 

pub struct KernelPCA<K: Kernel> { /* private fields */ }

Kernel-PCA estimator generic over any kernel that implements the crate’s Kernel trait. Typical ones are crate::RbfKernel, crate::LinearKernel, and crate::PolynomialKernel, but crate::SymbolicKernel (built with crate::KernelBuilder) also slots in.

KernelPCA is stateless — the “fitted model” is FittedKernelPCA, returned by KernelPCA::fit or KernelPCA::fit_transform.

Implementations

impl<K: Kernel> KernelPCA<K>

pub fn new(kernel: K, config: KernelPcaConfig) -> KernelPcaResult<Self>

Build a new Kernel-PCA estimator from a kernel and a config.

Errors

• KernelPcaError::InvalidInput when the config requests n_components == 0.

pub fn kernel(&self) -> &K

Access the underlying kernel (useful for diagnostics).

pub fn config(&self) -> &KernelPcaConfig

Access the configuration.

impl<K: Kernel + Clone + 'static> KernelPCA<K>

pub fn build(kernel: K, config: KernelPcaConfig) -> KernelPcaResult<Self>

Preferred constructor when the kernel is Clone + 'static — identical signature to KernelPCA::new but re-exposed here so that auto-derefs pick this up for the common case.

pub fn fit(&self, training: &[Vec<f64>]) -> KernelPcaResult<FittedKernelPCA<K>>

Fit the model on training — compute the Gram matrix, double center it, eigendecompose, and cache the top-n_components components for later projection.

Errors

• KernelPcaError::InvalidInput for empty or ragged training sets.
• KernelPcaError::EigendecompositionFailed if the underlying solver fails.
• KernelPcaError::InsufficientComponents if the kernel matrix does not have enough positive eigenvalues.

pub fn fit_transform( &self, training: &[Vec<f64>], ) -> KernelPcaResult<(FittedKernelPCA<K>, Array2<f64>)>

Convenience: fit on training and immediately project those same points.

Trait Implementations

impl<K: Clone + Kernel> Clone for KernelPCA<K>

fn clone(&self) -> KernelPCA<K>

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl<K: Debug + Kernel> Debug for KernelPCA<K>

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

Formats the value using the given formatter.

impl<K> Kernel for KernelPCA<K>

where K: Kernel,

fn compute(&self, _x: &[f64], _y: &[f64]) -> Result<f64>

Compute kernel value between two inputs.

fn name(&self) -> &str

Get kernel name for identification.

fn is_psd(&self) -> bool

Check if kernel is positive semi-definite.

fn compute_matrix(&self, inputs: &[Vec<f64>]) -> Result<Vec<Vec<f64>>>

Compute kernel matrix for a set of inputs.