Enum FeatureTransform 

pub enum FeatureTransform {
    StandardScaler {
        mean: f64,
        std: f64,
    },
    MinMaxScaler {
        min: f64,
        max: f64,
    },
    Log1p,
    Sqrt,
    Clip {
        lo: f64,
        hi: f64,
    },
    OneHotEncode {
        categories: Vec<String>,
    },
    Binarize {
        threshold: f64,
    },
    PolynomialFeatures {
        degree: u32,
    },
    ImputeMean {
        mean: f64,
    },
    ImputeMode {
        mode: String,
    },
}

A single reversible or irreversible transform applied to a FeatureValue.

Variants

StandardScaler

Z-score normalisation: (x - mean) / std. Output is 0.0 when std ≈ 0.

Fields

mean: f64

std: f64

MinMaxScaler

Min-max normalisation: (x - min) / (max - min). Output is 0.0 when range ≈ 0.

Fields

min: f64

max: f64

Log1p

Signed log1p: ln(1 + |x|) × sign(x).

Sqrt

Signed square-root: √|x| × sign(x).

Clip

Clamp the value to [lo, hi].

Fields

lo: f64

hi: f64

OneHotEncode

One-hot encode a Categorical value — produces one f64 per category. Unknown categories produce an all-zero vector.

Fields

categories: Vec<String>

Binarize

Threshold binarisation: 1.0 if x > threshold else 0.0.

Fields

threshold: f64

PolynomialFeatures

Polynomial expansion to degree: [x, x², …, xᵈ].

Fields

degree: u32

ImputeMean

Replace Missing with a pre-computed numeric mean.

Fields

mean: f64

ImputeMode

Replace a Missing categorical with the pre-computed mode string.

Fields

mode: String

Trait Implementations

impl Clone for FeatureTransform

fn clone(&self) -> FeatureTransform

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Debug for FeatureTransform

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

Formats the value using the given formatter.

impl PartialEq for FeatureTransform

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

Tests for self and other values to be equal, and is used by ==.

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

Tests for !=. The default implementation is almost always sufficient, and should not be overridden without very good reason.

impl StructuralPartialEq for FeatureTransform