nk::stats::NormalizeResult Struct Reference

Column-wise normalisation (Y = normalize(A, method)). More...

#include <descriptive.hpp>

Public Attributes
Value	n
	Normalized array ((A - c) ./ s).
Value	c
	Centering values C.
Value	s
	Scaling values S.

Detailed Description

Column-wise normalisation (Y = normalize(A, method)).

Supported method:

• "zscore" (default): (A - mean) / std (population std, normFlag=0)
• "center": A - mean
• "scale": A / std
• "range": (A - min) / (max - min)
• "norm": A / sqrt(sum(A²)) (unit ℓ² norm per column)
• "medianiqr": (A - median) / iqr (robust)

Parameters

A	Input matrix.
method	Normalisation method name (see list above).
param	Optional method parameter (Value::Empty → default): for "range" a [lo hi] vector; for "norm" the exponent p (Inf allowed); for "scale" a divisor string ('std'/'first'/'iqr'/'mad') or numeric; for "center" 'mean'/'median' or numeric.
mr	Memory resource (nullptr → process default).

Returns

{n, c, s} — n the normalised matrix (same shape as A), c the centering value C, s the scaling value S (MATLAB [N,C,S]=normalize): one value per operating slice, 1×1 for a vector / 1×W for a matrix, with n == (A - c) ./ s.

See also

rescale, zscore

Result of array normalization ([N, C, S] = normalize(A, ...)).

Definition at line 577 of file descriptive.hpp.

Member Data Documentation

c

Value nk::stats::NormalizeResult::c

Centering values C.

Definition at line 579 of file descriptive.hpp.

n

Value nk::stats::NormalizeResult::n

Normalized array ((A - c) ./ s).

Definition at line 578 of file descriptive.hpp.

s

Value nk::stats::NormalizeResult::s

Scaling values S.

Definition at line 580 of file descriptive.hpp.

---

The documentation for this struct was generated from the following file:

• descriptive.hpp