Struct ShapeletClassifierFit 

#[non_exhaustive]
pub struct ShapeletClassifierFit {
    pub transform: ShapeletTransformFit,
    pub classifier: ClassifFit,
    pub config: ShapeletClassifierConfig,
    pub classes: Vec<usize>,
}

A fitted shapelet-transform classifier: the discovered shapelet transform plus the inner classifier trained on the n×K distance features.

Stores the (already z-normalized) shapelets and the fitted inner ClassifFit so that predict reuses the identical shapelets, normalization, and FPCA rotation as the fit.

Fields (Non-exhaustive)

Non-exhaustive structs could have additional fields added in future. Therefore, non-exhaustive structs cannot be constructed in external crates using the traditional Struct { .. } syntax; cannot be matched against without a wildcard ..; and struct update syntax will not work.

transform: ShapeletTransformFit

The fitted shapelet transform (discovered shapelets + training features).

classifier: ClassifFit

The inner classifier fitted on the n×K shapelet-distance features.

config: ShapeletClassifierConfig

The configuration used to produce this fit.

classes: Vec<usize>

Distinct original labels in sorted order; index = remapped class produced by the inner classifier, value = the caller’s original label. Used to map inner predictions back to the caller’s label space.

Implementations

impl ShapeletClassifierFit

pub fn shapelets(&self) -> &ShapeletSet

The fitted shapelet set (already z-normalized, ordered by quality).

pub fn transform(&self) -> &ShapeletTransformFit

The fitted shapelet transform.

pub fn classifier(&self) -> &ClassifFit

The inner classifier fitted on the shapelet-distance features.

pub fn train_accuracy(&self) -> f64

Training-set accuracy of the inner classifier on the shapelet-distance features.

This is not a generalization estimate. The shapelets were selected to separate the training classes; evaluate on a held-out split via predict instead.

pub fn predict(&self, new_data: &FdMatrix) -> Result<Vec<usize>, FdarError>

Predict class labels for out-of-sample curves.

Transforms new_data through the stored shapelets (identical sequences, stored z-normalization) into an n_new×K distance-feature matrix, then classifies each row with the stored inner ClassifFit by reusing the FpcPredictor projection path (project features → FPC scores → predict_from_scores). Predictions are mapped back to the caller’s original label space.

Errors

• FdarError::InvalidDimension (propagated from the transform) if any series in new_data is shorter than a discovered shapelet.
• FdarError::InvalidParameter if the stored shapelet set is empty.

Trait Implementations

impl Clone for ShapeletClassifierFit

fn clone(&self) -> ShapeletClassifierFit

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Debug for ShapeletClassifierFit

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

Formats the value using the given formatter.

impl PartialEq for ShapeletClassifierFit

fn eq(&self, other: &ShapeletClassifierFit) -> bool

Equality operator ==.

fn ne(&self, other: &Rhs) -> bool

Inequality operator !=.

impl StructuralPartialEq for ShapeletClassifierFit