Library that provide you some statistics which are useful and with the best performance execution.

Getting Started

npm install --save statistcal-js

import the library :

/* how to import and use it */
const statistical = require('statistical-js');

use statistical methods :

const res = statistical.methods.sum([1, 2, 3]);

or use machine learning methods :

const p = new statistical.ml.Perceptron({
    learningRate: 0.3,
    bias: 0.5
});

Performance

To provide you with the best performance when using statistical-js, this library has implemented a variable cache system to save the last calculations. To see how manage it if it necessary, you can go to perfomance section !

Actually supported

Click there see the complete documentation.

• Simple statistics

  • Min
  • Max
  • Sum
  • Median
  • Mode
  • Mean
  • Variance
  • Standard deviation
  • Quantile
  • Summary (regroup all descriptives statistics above)
  • Percentile
  • Factorial
  • Geometric mean
  • Harmonic mean
  • InterQuartileRange
  • Sample variance (Non biased variance)
  • Sample Standard deviation (Non biased std deviation)

• Advanced statistics

  • Covariance
  • Binomial
  • Bernoulli
  • Poisson
  • Chi Squared Goodness Of Fit
  • T-test - one sample (t-value)
  • T-test - two sample (t-value)
  • Linear regression (slope, intersect)

• Machine learning

  • Perceptron (Single layer)

As soon as possible, more features will be available.

Test

To run test without coverage.

npm run test

If you want to generate coverage in ./coverage.

npm run test:coverage

Lint

The command to lint code and fix it at the same time

npm run lint

build

To build dev version ./lib/statistical.js

npm run dev

If you want to generate the production version ./lib/statistical.min.js

npm run build

License

This project is licensed under the MIT License - see the LICENSE.md file for details