Struct linfa_preprocessing::linear_scaling::LinearScaler

pub struct LinearScaler<F: Float> { /* fields omitted */ }

Linear Scaler: learns scaling parameters, according to the specified method, from a dataset, producing a fitted linear scaler that can be used to scale different datasets using the same parameters.

Example

use linfa::traits::{Fit, Transformer};
use linfa_preprocessing::linear_scaling::LinearScaler;

// Load dataset
let dataset = linfa_datasets::diabetes();
// Learn scaling parameters
let scaler = LinearScaler::standard().fit(&dataset).unwrap();
// scale dataset according to parameters
let dataset = scaler.transform(dataset);

Implementations

impl<F: Float> LinearScaler<F>

pub fn new(method: ScalingMethod<F>) -> Self

Initializes the scaler with the specified method.

pub fn method(self, method: ScalingMethod<F>) -> Self

Setter for the scaler method

pub fn standard() -> Self

Initializes a Standard scaler

pub fn standard_no_mean() -> Self

Initializes a Standard scaler that does not subract the mean to the features

pub fn standard_no_std() -> Self

Initializes a Stadard scaler that does not scale the features by the inverse of the standard deviation

pub fn min_max() -> Self

Initializes a MinMax scaler with range [0,1]

pub fn min_max_range(min: F, max: F) -> Self

Initializes a MinMax scaler with the specified minimum and maximum values for the range.

If min is bigger than max then fitting will return an error on any input.

pub fn max_abs() -> Self

Initializes a MaxAbs scaler

Trait Implementations

impl<F: Float, D: Data<Elem = F>, T: AsTargets> Fit<ArrayBase<D, Dim<[usize; 2]>>, T, Error> for LinearScaler<F>

type Object = FittedLinearScaler<F>

fn fit(&self, x: &DatasetBase<ArrayBase<D, Ix2>, T>) -> Result<Self::Object>

Fits the input dataset accordng to the scaler method. Will return an error if the dataset does not contain any samples or (in the case of MinMax scaling) if the specified range is not valid.