array-correl

1.1.1 • Public • Published

[![NPM package][npm-img]][npm-url] [![Build Size][build-size-img]][build-size-url]

array-correl

Generates array of correlated pairs of numbers with desired mean and deviation. It also inspects an array of such correlated pairs and returns the Pearson correlation coefficient, the deviation, mean and extent (min and max)

Installation

npm install array-correl

Generate

The generate method generates an array of length count with correlated sets of numbers in the { x: xValue, y: yValue } format. Optionally, the following parameters can be specified:

  • Target correlation coefficient ranging from 0 to 1,
  • Target mean,
  • Target deviation

If the correlation coefficient is not specified, it defaults to 0.7; if mean is not specified, it defaults to 0; if deviation is not specified, it defaults to 1.

API

array-correl.generate(count, [correlation, [mean, [deviation]]]);

Errors

If no count argument is passed, a ReferenceError is thrown. If any passed argument is not a number (or cannot be coerced to a number), a TypeError is thrown.

Inspect

The inspect method analyzes the supplied array and returns several statistics measures. The format of the input array must be one of the following:

  • Each array element contains an object with the following properties: { x: xValue, y: yValue }
  • Each array element contains a two-element array of the following structure: [ xValue, yValue ]

The function returns the following parameters in a JS object:

  • The Pearson correlation coefficient of the x and y data series
  • The deviation of each of the the x and y data series
  • The mean of each of the x and y data series
  • The extent (min and max) of each of the x and y data series

API

array-correl.inspect(array);

Errors

If no array argument is passed, a ReferenceError is thrown. If the passed array doesn't contain either two-element arrays or objects with x and y properties, a TypeError is thrown.

Example:

// Import methods and version from array-correl from node_modules
const { generate, inspect, version } = require('array-correl');

// Import methods and version from array-correl locally
// const { generate, inspect, version } = require('../src/index.js');

// Log version
console.log(`array-correl version: ${version}`);

// Define inputs to the generate method
const count = 10000;
const correlation = 0.8;
const mean = 100;
const deviation = 1;

// Log inputs
console.log('\nInputs to generate:');
console.log(`  count: ${count}`);
console.log(`  correlation: ${correlation}`);
console.log(`  mean: ${mean}`);
console.log(`  deviation: ${deviation}`);

// Create an array with values of correlated pairs of numbers
const array = generate(count, correlation, mean, deviation);

console.log('\nFirst 10 correlated pairs:');
// Log the first 10 array elements
for (let i = 0; i < 10; i++) {
  console.log(`  ${JSON.stringify(array[i])}`);
}

// Inspect the array
const result = inspect(array);
console.log(`\nOutput of inspect: \n${JSON.stringify(result, null, 2)}\n`);

/*
Example output:
  array-correl version: 1.0.0

  Inputs to generate:
    count: 10000
    correlation: 0.8
    mean: 100
    deviation: 1

  First 10 correlated pairs:
    {"x":99.71068933565104,"y":99.86127640311815}
    {"x":100.83597240809056,"y":100.074669667149}
    {"x":102.0130498785201,"y":101.38201755974566}
    {"x":100.65061261210614,"y":99.88629765791357}
    {"x":98.30631250480867,"y":98.49421518032264}
    {"x":99.48427439034523,"y":99.81419151256067}
    {"x":98.90338120221003,"y":98.38761587336445}
    {"x":101.06519781042032,"y":101.50229105410696}
    {"x":98.7248721438115,"y":98.06148966509333}
    {"x":99.40834127473076,"y":98.90308575665651}

  Output of inspect:
  {
    "r": 0.802558017048035,
    "xDeviation": 1.0004768422899535,
    "yDeviation": 0.9984232776109369,
    "xExtent": [
      95.99318668690016,
      103.92406651194449
    ],
    "yExtent": [
      96.15152222172428,
      103.84111296300848
    ],
    "xMean": 100.00871360178544,
    "yMean": 100.00635903137874
  }
*/

Test

Run the test suite like so:

npm run test

Package Sidebar

Install

npm i array-correl

Weekly Downloads

4

Version

1.1.1

License

MIT

Unpacked Size

21.1 kB

Total Files

9

Last publish

Collaborators

  • boeric