moments — n4m.lowlevel.moments.moments¶
Namespace: n4m.lowlevel.moments · Fully-qualified: n4m.lowlevel.moments.moments · Catalog id: utilities.moments
API surface¶
C ABI (ABI 2): n4m_lowlevel_moments_compute · n4m_lowlevel_moments_subset_compute · n4m_lowlevel_moments_subtract. Use the linked public header for the exact signature, configuration, and result handles.
Python (verified public re-export): from n4m.lowlevel.moments import moments
Signature: moments(X, y, *, row_indices = None)
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.
Parameters¶
Name |
Type |
Default |
|---|---|---|
|
|
|
|
|
|
|
|
|
Explanations¶
Bibliographic source¶
No single canonical paper defines the ABI utility surface; it implements standard first- and second-moment sufficient statistics used by linear-model updates.
Mathematical principle¶
Accumulate quantities such as sample count, feature sums, cross-products, and response cross-products so compatible linear fits can be computed or updated without retaining every original row.
Appropriate uses¶
Auditable aggregation and replay of moment-compatible linear calibration calculations.
Limits and validation¶
Moments do not retain row identities, nonlinear transformations, or arbitrary fold structure. Merging statistics is only valid for identically defined feature spaces and preprocessing; it cannot recover information discarded upstream.
Implementation¶
n4m.lowlevel.moments; C ABI symbols in the n4m_lowlevel_moments_* family.
Sources and provenance¶
https://github.com/GBeurier/nirs4all-methods/blob/main/bindings/python/src/n4m/lowlevel/moments.py; https://github.com/GBeurier/nirs4all-methods/blob/main/cpp/include/n4m/lowlevel.h
Catalog note¶
Row-additive sufficient-statistics substrate returning raw and centered X’X, X’Y and Y’Y moments. Fold subtraction is exact for train-from-heldout moment screens.
Timing benchmark: benchmarks/cross_binding/bench_moment_sweep_timing.py
See also: methods index.