Struct FeatureNormalizer 

pub struct FeatureNormalizer { /* private fields */ }

Z-score (standardization) normalizer: (x - mean) / std.

Must be fit on training data before calling transform.

Example

use fin_primitives::ml_features::FeatureNormalizer;

let mut norm = FeatureNormalizer::new();
norm.fit(&[1.0, 2.0, 3.0, 4.0, 5.0]).unwrap();
let z = norm.transform(3.0).unwrap();
assert!((z - 0.0).abs() < 1e-10);

Implementations

impl FeatureNormalizer

pub fn new() -> Self

Constructs an unfitted normalizer.

pub fn fit(&mut self, data: &[f64]) -> Result<(), FinError>

Fit the normalizer to a dataset.

Errors

FinError::InsufficientData if fewer than 2 observations. FinError::InvalidInput if std dev is zero (constant series).

pub fn transform(&self, value: f64) -> Result<f64, FinError>

Transform a single value to z-score.

Errors

FinError::InvalidInput if the normalizer has not been fit.

pub fn inverse_transform(&self, z: f64) -> Result<f64, FinError>

Inverse transform a z-score back to original scale.

Errors

FinError::InvalidInput if the normalizer has not been fit.

pub fn transform_slice(&self, values: &[f64]) -> Result<Vec<f64>, FinError>

Transform a slice of values in place.

Errors

FinError::InvalidInput if the normalizer has not been fit.

Trait Implementations

impl Clone for FeatureNormalizer

fn clone(&self) -> Self

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Debug for FeatureNormalizer

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

Formats the value using the given formatter.

impl Default for FeatureNormalizer

fn default() -> Self

Returns the “default value” for a type.