math-script

Navigation: Docs/Reference | Demo Playground | Github Source

Math-Script

A math library for JS with an emphasis on matrix operations, non linear solvers, visualizations, and overload support!.

var M = require('math-script');
 
M.execute(function(M){
    var X = M.range(0,0.1,10); // Create an input matrix, [100,1] matrix 
    // Your typical ugly JS math library 
    var Y = X.pow(3).multiply(-1.3).add(X.pow(2).multiply(13)).add(X.multiply(-5)).add(30);
    // What math-script supports
    Y = -1.3*X**3 + 13*X**2 + -5*X + 30; // not normally possible in JS
    console.log(Y.print());
});

Core features:

• Matrix operations operator support in execute environment *, +, **, -
• Feature rich [M,N] matrix operation support, see reference
• SVD support / Nonlinear solvers

Install

NPM Installation:

npm install math-script 

Browser minimal:

See dist folder for files

<script type="text/javascript" src="math-script.js"></script>

Math-Script with Plotting

<script type="text/javascript" src="plotly.min1.19.2.js"></script> 
<script type="text/javascript" src="math-script-full.js"></script>

Library Application Sample

Linear Regression Example:

This example uses linear least squares to identify a best fitting polynomial to arbitrary data. This example demonstrates the matrix and plotting functionality of libary.

var X = Matrixs.range(-1.2, 0.01, 1.2); // Creates a matrixs with a range of values 
var Y = X.pow(3).subtract(X).addNoise(0.9); // Y = X^3 - X + noise 
 
// Create a column matrix for a least square solution [x^3,x^2,x,1] 
var A = X.pow(3).catH(X.pow(2)).catH(X).catH(Matrixs.ones(X.length(),1)); 
 
var x = A.lsq(Y); // Least squares solution 
var yFit = A.multiply(x); // Ax = Y 
plotyLayout.showlegend= false;
plotyLayout.title= 'Least Square Example'; 
Plots.create([X,Y],{type:'scatter'}); 
Plots.add([X,yFit]);

Basic Matrix Examples

Maths.js supports numerous matrixs opperations. Below is small sampling and see here for the matrixs reference.

Matrix Creation:

var A = Matrixs.make([[1,2,3],[4,5,6],[7,8,5]); //creates a 3x3 matrix 
A.print(); //prints matrix 
//Results 
//1.000		2.000		3.000
//4.000		5.000		6.000
//7.000		8.000		5.000

Matrix Opperations (Addition, Multiplicaiton ...ect): One can add or perform any basic matrix opperation with scalars or other matricies. These commands can be sequentially chained.

var A = Matrixs.make([[1,2,3],[4,5,6],[7,8,5]); //creates a 3x3 matrix 
A.add(1).add([[1,2,3],[0,0,0],[0,0,0]]).print(); // adds scalar of 1 to matrix, then adds a 3x3 matrix, then prints result.
//Results
//3.000		5.000		7.000
//5.000		6.000		7.000
//8.000		9.000		6.000

Matrix Inversion: Maths.js can invert arbitrary sized(e.g. 100x100) singular and non singular(with SVD) matricies.

var A = new Matrixs([[1,2,3],[4,5,6],[7,8,9]]); //creates a 3x3 matrix
A.invert().print(); //Prints after inverison. Print is set to 3 decimals 
//Results: 
//-0.639        -0.167		0.306
//-0.056        0.000		0.056
//0.528         0.167      -0.194

Basic Plot Example

Simple Line and Scatter Plot :

Maths.js can easily create plots for visualization. Please see here for the plot reference. . The backbone of this library is plotly whos backbone is D3 meaning the plots are SVG based.

var X = Matrixs.range(100); // Creates a 0-99 value array
var Y = X.addNoise(0.9); // Adds 90% relative noise 
Plots.create([Y],{type:'scatter'}); //Creates scatter plot 
Plots.add([X]); // Adds orange best fit line 

License: MIT

https://opensource.org/licenses/MIT