`pp_haar` — Single-level Haar discrete wavelet transform¶

Group: Augmentation · C ABI: n4m_transform_haar_*

Description¶

Single-level Haar DWT coefficient transform.

Parameters¶

No constructor parameters.

API and bindings¶

C ABI (ABI 2): n4m_transform_haar_create · n4m_transform_haar_destroy · n4m_transform_haar_output_cols · n4m_transform_haar_transform. Use the linked public header for the exact signature, configuration, and result handles.

Python (verified public re-export):

from n4m.transform.wavelet import Haar

Source signature: Haar().

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¶

Haar (1910), Zur Theorie der orthogonalen Funktionensysteme, Mathematische Annalen 69, 331–371, https://doi.org/10.1007/BF01456326.

Mathematical principle¶

Adjacent sample pairs are projected onto the Haar scaling and wavelet filters, producing approximation \((x_{2j}+x_{2j+1})/\sqrt2\) and detail \((x_{2j}-x_{2j+1})/\sqrt2\) coefficients under the configured endpoint convention.

Appropriate uses¶

Compact separation of coarse spectral shape and one-channel-scale detail.

Limits and validation¶

Haar is discontinuous and shift-sensitive; it can represent smooth bands less efficiently than longer wavelets. Odd-length boundary behaviour must be considered.

Implementation¶

n4m.transform.wavelet.Haar wraps n4m_transform_haar_*; the single-level native transform is in preprocessing/wavelets/haar.c.

The ABI-2 implementation is the n4m_transform_haar_* lifecycle in libn4m.

Sources and provenance¶

https://doi.org/10.1007/BF01456326; https://github.com/GBeurier/nirs4all-methods/blob/main/cpp/src/core/preprocessing/wavelets/haar.c

See also: methods index · interactive dashboard

pp_haar — Single-level Haar discrete wavelet transform¶