cpd::Normalization Struct Reference

The results of normalizing data to a unit cube (or whatever dimensionality). More...

#include <normalization.hpp>

Public Member Functions
	Normalization (const Matrix &fixed, const Matrix &moving, bool linked=true)
	Creates a new normalization for the provided matrices. More...

Public Attributes
Vector	fixed_mean
	The average of the fixed points, that was subtracted from those data. More...
Matrix	fixed
	The fixed points. More...
double	fixed_scale
	The scaling factor for the fixed points. More...
Vector	moving_mean
	The average of the moving points, that was subtracted from those data. More...
Matrix	moving
	The moving points. More...
double	moving_scale
	The scaling factor for the moving points. More...

Detailed Description

The results of normalizing data to a unit cube (or whatever dimensionality).

Constructor & Destructor Documentation

cpd::Normalization::Normalization	(	const Matrix &	fixed,
		const Matrix &	moving,
		bool	linked = true
	)

Creates a new normalization for the provided matrices.

If linked = true, apply the same scaling to both sets of points. This is recommended if you are working with data that should not be scaled, e.g. LiDAR data. If linked = false, each point set is scaled seperately.

Myronenko's original implementation only had linked = false logic.

Member Data Documentation

Matrix cpd::Normalization::fixed

The fixed points.

Vector cpd::Normalization::fixed_mean

The average of the fixed points, that was subtracted from those data.

double cpd::Normalization::fixed_scale

The scaling factor for the fixed points.

Matrix cpd::Normalization::moving

The moving points.

Vector cpd::Normalization::moving_mean

The average of the moving points, that was subtracted from those data.

double cpd::Normalization::moving_scale

The scaling factor for the moving points.

---

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

• /home/travis/build/gadomski/cpd/include/cpd/normalization.hpp