Skip to main content

RobustScaler

#include <Skigen/Preprocessing>

template <typename Scalar = double>
class Skigen::RobustScaler(with_centering=true, with_scaling=true, quantile_range={25, 75})

Scale features using statistics that are robust to outliers.

This Scaler removes the median and scales the data according to the quantile range (defaults to IQR: Interquartile Range). The IQR is the range between the 1st quartile (25th quantile) and the 3rd quartile (75th quantile).

Centering and scaling happen independently on each feature.

Mirrors sklearn.preprocessing.RobustScaler.

Parameters:

  • with_centering : bool, default=true If true, center the data before scaling by subtracting the median (bool, default true).

  • with_scaling : bool, default=true If true, scale the data to interquartile range (bool, default true).

  • quantile_range : std::pair< Scalar, Scalar >, default={25, 75} Quantile range for computing scale_ (std::pair<Scalar,Scalar>, default {25, 75}).

Attributes:

  • with_centering : bool Whether centering is enabled.

  • with_scaling : bool Whether scaling is enabled.

  • center : RowVectorType Per-feature median (1 × n_features).

  • scale : RowVectorType Per-feature interquartile range (1 × n_features).

Methods​

fit(X)​

Compute median and quantile range for later scaling.

Parameters:

  • X : MatrixType Training data of shape (n_samples, n_features).

Returns:

  • result : RobustScaler Reference to the fitted transformer (*this).

transform(X)​

Center and scale the data.

Parameters:

  • X : MatrixType Data matrix of shape (n_samples, n_features).

Returns:

  • result : MatrixType Transformed data of same shape.

Throws:

  • std::runtime_error — if the model has not been fitted.

inverse_transform(X)​

Scale back the data to the original representation.

Parameters:

  • X : MatrixType Transformed data of shape (n_samples, n_features).

Returns:

  • result : MatrixType Un-transformed data of same shape.

Throws:

  • std::runtime_error — if the model has not been fitted.

transform_inplace(X)​

Transform features in-place using robust statistics.

Parameters:

  • X : Eigen::Ref< MatrixType > Data matrix of shape (n_samples, n_features), modified in place.

Throws:

  • std::runtime_error — if the model has not been fitted.

inverse_transform_inplace(X)​

Inverse-transform features in-place to original scale.

Parameters:

  • X : Eigen::Ref< MatrixType > Scaled data matrix, modified in place.

Throws:

  • std::runtime_error — if the model has not been fitted.

Example​

// RobustScaler uses median and IQR — resistant to outliers
Skigen::RobustScaler<double> robust;
Eigen::MatrixXd Z_robust = robust.fit_transform(X);

std::cout << "Robust-scaled:\n" << Z_robust << "\n\n";
std::cout << "Center (median): " << robust.center() << "\n";
std::cout << "Scale  (IQR):    " << robust.scale()  << "\n\n";