pp_kbins_disc — Per-feature integer k-bin discretization¶
Group: Preprocessing · C ABI: n4m_transform_kbins_discretizer_*
Description¶
Per-column integer binning using uniform or quantile edges.
Parameters¶
Name |
Type |
Default |
|---|---|---|
|
|
|
|
|
|
API and bindings¶
C ABI (ABI 2): n4m_transform_kbins_discretizer_create · n4m_transform_kbins_discretizer_destroy · n4m_transform_kbins_discretizer_fit · n4m_transform_kbins_discretizer_is_fitted · n4m_transform_kbins_discretizer_transform. Use the linked public header for the exact signature, configuration, and result handles.
Python (verified public re-export):
from n4m.transform.resampling import IntegerKBinsDiscretizer
Source signature: IntegerKBinsDiscretizer(n_bins: int = 5, strategy: str | int = 0).
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¶
No single canonical paper: this is uniform or empirical-quantile scalar quantization. Scikit-learn documents the corresponding estimator at https://scikit-learn.org/stable/modules/generated/sklearn.preprocessing.KBinsDiscretizer.html.
Mathematical principle¶
Fit computes n_bins + 1 edges independently for every column, either equally spaced between its extrema or at empirical quantiles. Transform replaces each value by the integer bin index selected by those stored edges.
Appropriate uses¶
Robust coarse encoding for threshold models or exploratory analyses where continuous amplitude resolution is unnecessary.
Limits and validation¶
Quantization discards within-bin information; quantile edges can collapse with ties and all edges are training-data dependent. It is rarely appropriate before models that exploit smooth spectral geometry.
Implementation¶
n4m.transform.resampling.IntegerKBinsDiscretizer calls n4m_transform_kbins_discretizer_*; fitting and integer encoding are in resampling/kbins_discretizer.c.
The ABI-2 implementation is the n4m_transform_kbins_discretizer_* lifecycle in libn4m.
Sources and provenance¶
https://scikit-learn.org/stable/modules/generated/sklearn.preprocessing.KBinsDiscretizer.html; https://github.com/GBeurier/nirs4all-methods/blob/main/cpp/src/core/preprocessing/resampling/kbins_discretizer.c
See also: methods index · interactive dashboard