aug_poly_drift — Random polynomial baseline drift¶
Group: Augmentation · C ABI: n4m_augmentation_poly_drift_*
Description¶
Add random polynomial baseline drift.
Parameters¶
Name |
Type |
Default |
|---|---|---|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
API and bindings¶
C ABI (ABI 2): n4m_augmentation_poly_drift_apply · n4m_augmentation_poly_drift_create · n4m_augmentation_poly_drift_destroy. Use the linked public header for the exact signature, configuration, and result handles.
Python (verified public re-export):
from n4m.augmentation.drift import PolynomialBaselineDrift
Source signature: PolynomialBaselineDrift(degree: int = 2, coeff_min = None, coeff_max = None, rng: Optional[PCG64] = None, seed: 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 canonical publication specifies these random coefficient ranges. The method is a controlled extension of offset/slope spectral augmentation.
Mathematical principle¶
Channels are mapped to \(t_j\in[-1,1]\). For every order \(k=0,\ldots,d\), a coefficient \(c_{ik}\) is sampled uniformly between coeff_min[k] and coeff_max[k], and \(X'_{ij}=X_{ij}+\sum_{k=0}^{d}c_{ik}t_j^k\). degree sets curvature order; the coefficient arrays must contain degree + 1 bounds.
Appropriate uses¶
Generating smooth, low-frequency baseline shapes beyond an affine tilt.
Limits and validation¶
Polynomial drift is a numerical nuisance model rather than a physical scatter model. High degrees can oscillate and extrapolation meaning changes with the number or ordering of channels.
Implementation¶
Python role API n4m.augmentation.drift.PolynomialBaselineDrift; ABI 2 family n4m_augmentation_poly_drift_{create,apply,destroy}.
The ABI-2 implementation is the n4m_augmentation_poly_drift_* lifecycle in libn4m.
Sources and provenance¶
https://github.com/GBeurier/nirs4all-methods/blob/main/cpp/src/core/augmentation/drift/poly_drift.h
See also: methods index · interactive dashboard