standard_scale — n4m.transform.scaling.standard_scale

Namespace: n4m.transform.scaling · Fully-qualified: n4m.transform.scaling.standard_scale · Catalog id: preprocessing.scaling.standard_scale

API surface

C ABI: no standalone exported symbol is declared for this method.

Python: catalog binding is not currently an AST-verified public n4m re-export. See the implementation source below.

R: no current source-verified entry point was found for this catalog method.

MATLAB / Octave: no current source-verified entry point was found for this catalog method.

Explanations

Bibliographic source

Population standardization uses the arithmetic mean and population variance. The public StandardScaler reference documents the same ddof=0 convention: https://scikit-learn.org/stable/modules/generated/sklearn.preprocessing.StandardScaler.html.

Mathematical principle

For each training column, the native fitter stores its mean and population standard deviation. Transform applies (x-mean)/scale, with centering and scaling independently controlled by with_mean and with_std. A zero-variance column receives scale one. Held-out rows reuse the learned vectors; they never update training statistics.

Appropriate uses

Standardizing heterogeneous numerical predictors before a fitted encoder or estimator, with a separate scaler fitted on each training fold.

Limits and validation

Fitting on validation or Test rows leaks their distribution into training. Standardization is sensitive to outliers and does not establish normality. Population variance differs from sample variance; ddof=1 references require an explicit convention adjustment before parity comparisons.

Implementation

The native transformer preprocessing.scaling.standard_scale is exposed as n4m.roles.StandardScale through the shared n4m_estimator_* lifecycle. cpp/include/n4m/estimator.h defines the shared ABI. cpp/src/core/estimator/fitted_core.cpp owns fitting, transformation and portable learned means/scales; bindings marshal inputs without numerical logic.

Sources and provenance

https://github.com/GBeurier/nirs4all-methods/blob/main/cpp/src/core/estimator/fitted_core.cpp; https://github.com/GBeurier/nirs4all-methods/blob/main/catalog/methods/preprocessing.scaling.standard_scale.yaml

Catalog note

Native transformer through the shared n4m_estimator_* role API, with no dedicated C exports. Train-only population standardization (ddof=0; zero variance scale=1), with portable native learned means and scales.

See also: methods index.