mse-ts

Mean Squared Error

Mean Squared Error estimation function + type definitions.

Installing

$ npm install mse-ts

Importing the package

Using import

import mse from 'mse-ts';

Using require

const mse = require('mse-ts');

Usage

Calculating MSE - Example 1

import mse from 'mse-ts';

const y_true = [3, -0.5, 2, 7];
const y_pred = [2.5, 0.0, 2, 8];

const meanSquaredError = mse(y_true, y_pred);
console.log(meanSquaredError);

output

0.375

Calculating MSE - Example 2

import mse from 'mse-ts';


const y_true = [
  188, 100, 114, 171, 171, 173, 230, 149,
  191,  81,  61,  62, 127, 217,  62,  81,
  178, 159, 245,  18,   9,  86, 201, 166,
  122, 210,   4, 182,  15,  18, 135, 203,
  222, 134, 154,  21,  71, 217,  48, 153,
  113, 234, 207, 119,  51,  61, 149, 222,
  186,  38, 158,  79, 185,   1, 118, 222,
  22, 137, 110, 206,  94, 120, 163, 241
];
const y_pred = [
  188, 100, 114, 171, 171, 173, 230, 149,
  191,  81,  61,  62, 123, 217,  62,  81,
  178, 159, 245,  18,   9,  86, 201, 166,
  122, 210,   4, 200,  15,  18, 135, 203,
  222, 134, 154,  21,  71, 217,  48, 153,
  113, 234, 207, 119,  51,  61, 149, 222,
  186,  38, 158,  79, 185,   1, 118, 222,
  22, 137, 110, 206,  94, 120, 163, 241
];
const meanSquaredError = mse(y_true, y_pred);

if (meanSquaredError <> 0) {
    console.log('data sets are different by ' + meanSquaredError);
}

output

'data sets are different by 5.3125'

Steps

You may provide a custom step value

import mse from 'mse-ts';

const y_true = [3, -0.5, 2, 7];
const y_pred = [2.5, 0.0, 2, 8];

const meanSquaredError = mse(y_true, y_pred, { step: 2 });
console.log(meanSquaredError);

output

0.0625

Caveats

• The length of y_true should always be higher than or equal to y_pred. Non-compliance will result in an yPred at index i is undefined error
• Passing in empty arrays will return NaN

More info

Find out more about the applications of MSE over on Wikipedia: https://en.wikipedia.org/wiki/Mean_squared_error

license

MIT

Changelog

[1.0.2] - 2021-02-21

Fixed

• Package name under install instructions in README.md