Struct RobustScaler 

pub struct RobustScaler { /* private fields */ }

Scales features using statistics robust to outliers.

Uses the median and interquartile range (IQR = Q3 - Q1) instead of mean and standard deviation, making it robust to outliers.

The transformation is: X_scaled = (X - median) / IQR

Examples

use aprender::preprocessing::RobustScaler;
use aprender::traits::Transformer;
use aprender::primitives::Matrix;

let data = Matrix::from_vec(5, 1, vec![1.0, 2.0, 3.0, 4.0, 100.0])
    .expect("valid matrix dimensions");

let mut scaler = RobustScaler::new();
let scaled = scaler.fit_transform(&data).expect("fit_transform should succeed");
// Median=3.0, IQR=Q3-Q1. Outlier (100) does not distort scaling.

Implementations

impl RobustScaler

pub fn new() -> RobustScaler

Creates a new RobustScaler with default settings.

pub fn with_centering(self, centering: bool) -> RobustScaler

Sets whether to center the data by subtracting the median.

pub fn with_scaling(self, scaling: bool) -> RobustScaler

Sets whether to scale the data by dividing by IQR.

pub fn median(&self) -> &[f32]

Returns the median of each feature.

pub fn iqr(&self) -> &[f32]

Returns the IQR of each feature.

pub fn is_fitted(&self) -> bool

Returns true if the scaler has been fitted.

Trait Implementations

impl Clone for RobustScaler

fn clone(&self) -> RobustScaler

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Debug for RobustScaler

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

Formats the value using the given formatter.

impl Default for RobustScaler

fn default() -> RobustScaler

Returns the “default value” for a type.

impl<'de> Deserialize<'de> for RobustScaler

fn deserialize<__D>( __deserializer: __D, ) -> Result<RobustScaler, <__D as Deserializer<'de>>::Error>

where __D: Deserializer<'de>,

Deserialize this value from the given Serde deserializer.

impl Serialize for RobustScaler

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

where __S: Serializer,

Serialize this value into the given Serde serializer.

impl Transformer for RobustScaler

fn fit(&mut self, x: &Matrix<f32>) -> Result<(), AprenderError>

Fits the transformer to data.

fn transform(&self, x: &Matrix<f32>) -> Result<Matrix<f32>, AprenderError>

Transforms data using fitted parameters.

fn fit_transform( &mut self, x: &Matrix<f32>, ) -> Result<Matrix<f32>, AprenderError>

Fits and transforms in one step.