nem: Negative Eigenvalue Magnitude

In maotai: Tools for Matrix Algebra, Optimization and Inference

Description

Negative Eigenvalue Magnitude (NEM) is a measure of distortion for the data whether they are lying in Euclidean manner or not. When the value is exactly 0, it means the data is Euclidean. On the other hand, when NEM is far away from 0, it means not Euclidean. The concept of NEM is closely related to the definiteness of a Gram matrix.

Usage

nem(data)

Arguments

data	an (n\times p) matrix whose rows are observations.

Value

a nonnegative NEM value.

References

\insertRef

pekalska_noneuclidean_2006maotai

Examples

## use simple example of iris dataset 
data(iris) 
mydat = as.matrix(iris[,1:4])

## calculate NEM
nem(mydat)

maotai documentation built on Oct. 25, 2021, 9:06 a.m.