Mahalanobis

Calculate the Mahalanobis distances from an array of multivariate data. Useful for calculating "outlierness" of data points across dimensions in certain situations.

Installation

npm install mahalanobis

Usage

mahalanobis(points) returns an object with two methods: .distance(point) to get the Mahalanobis distance of one point vs. the distribution, and .all() to return an array of Mahalanobis distances for all the input points.

The input array should be an array of rows, like:

[
  [1,2,3],
  [1,5,6],
  [7,9,10],
  [9,0,-5]
]

var mahalanobis = require("mahalanobis");

var data = [
  [1, 2, 3],
  [1, 5, 6],
  [7, 3, 4],
  [2, 3, 0],
  [9, 0, -5]
];

var m = mahalanobis(data);

data.forEach(function(point, i) {
  console.log("The distance for row " + i + " is " + m.distance(point));
});

/*
The distance for row 0 is 1.78834390789133
The distance for row 1 is 1.3487167047236224
The distance for row 2 is 1.5829207125424334
The distance for row 3 is 1.367039530625441
The distance for row 4 is 1.6150400171571428
*/

You can also use the .all() method to directly return an array of distances for all the points in the input array.

var mahalanobis = require("mahalanobis");

var data = [
  [1, 2, 3],
  [1, 5, 6],
  [7, 3, 4],
  [2, 3, 0],
  [9, 0, -5]
];

var distances = mahalanobis(data).all();

distances.forEach(function(distance, i) {
  console.log("The distance for row " + i + " is " + distance);
});

/*
The distance for row 0 is 1.78834390789133
The distance for row 1 is 1.3487167047236224
The distance for row 2 is 1.5829207125424334
The distance for row 3 is 1.367039530625441
The distance for row 4 is 1.6150400171571428
*/