Numbers - an advanced mathematics toolkit for JavaScript and Node.js.
Numbers provides a comprehensive set of mathematical tools that currently are not offered in JavaScript. These tools include:
- Basic calculations
- Calculus
- Matrix Operations
- Prime Numbers
- Statistics
- More...
A few things to note before using: JavaScript, like many languages, does not necessarily manage floating points as well as we'd all like it to. For example, if adding decimals, the addition tool won't return the exact value. This is an unfortunate error. Precautions have been made to account for this. After including numbers, you can set an error bound. Anything in this will be considered an "acceptable outcome."
The primary uses cases are client side operations which the DOM will recognize (e.g. 1.1px == 1px). It can be used for data analysis, calculations, etc. on the server as well.
Numbers is pretty straightforward to use.
With node, simply require it:
var numbers = require('numbers');
For example, if we wanted to estimate the integral of sin(x) from -2 to 4, we could:
Use Riemann integrals (with 200 subdivisions)
numbers.calculus.Riemann(Math.sin, -2, 4, 200);
Or use adaptive simpson quadrature (with epsilon .0001)
numbers.calculus.adaptiveSimpson(Math.sin, -2, 4, .0001);
User-defined functions can be used too:
var myFunc = function(x) {
return 2*Math.pow(x,2) + 1;
}
numbers.calculus.Riemann(myFunc, -2, 4, 200);
numbers.calculus.adaptiveSimpson(myFunc, -2, 4, .0001);
Now say we wanted to run some matrix calculations:
We can add two matrices
var array1 = [0, 1, 2];
var array2 = [3, 4, 5];
numbers.matrix.addition(array1, array2);
We can transpose a matrix
numbers.matrix.transpose(array);
When working with vectors, treat them like single row matrices:
var vector1 = [[1, 0, 0]];
Numbers also includes some basic prime number analysis. We can check if a number is prime:
// basic check
numbers.prime.simple(number);
// Miller-Rabin primality test
numbers.prime.millerRabin(number);
The statistics tools include mean, median, mode, standard deviation, random sample generator, correlation, confidence intervals, t-test, chi-square, and more.
numbers.statistic.mean(array);
numbers.statistic.median(array);
numbers.statistic.mode(array);
numbers.statistic.standardDev(array);
numbers.statistic.randomSample(lower, upper, n);
numbers.statistic.correlation(array1, array2);
For further documentation, check out numbers.github.io
To execute, run:
npm test
Note: Make sure to install the plugins by running npm install
.
With gulp:
npm run g-test
To perform a code quality check using jshint, run
npm run lint
or, with gulp:
npm run g-lint
Running without gulp will error on warnings.
To format all the tests and lib files using jsbeautifier, run
npm run format
or, with gulp:
npm run g-format
To update the public JavaScript, run
npm run build
This will compile the entire library into a single file accessible at src/numbers.js. It will also minify the file into public/numbers.min.js.
With gulp:
npm run g-build
Numbers.js is also available on Bower via
$ bower install numbers.js
- Steve Kaliski - @stevekaliski
- David Byrd - @thebyrd
- Ethan Resnick - @studip101
- Dakota St. Laurent - @SaintDako
- Kartik Talwar - @KartikTalwar
In no particular order:
- Ethan aka
altercation
- Hrishikesh Paranjape aka
hrishikeshparanjape
- Greg Leppert aka
leppert
- Lars-Magnus Skog aka
ralphtheninja
- Tim Wood aka
codearachnid
- Miles McCrocklin aka
milroc
- Nate Kohari aka
nkohari
- Eric LaForce aka
elaforc
- btmills aka
btmills
- swair shah aka
swairshah
- Jason Hutchinson aka
Zikes
- Philip I. Thomas aka
philipithomas
- Brandon Benvie aka
Benvie
- Larry Battle aka
LarryBattle