MatlabJS

A lightweight JavaScript library for MATLAB/Octave users

This library gives you some essential array and matrix functions written in JavaScript for easing your scientific computing projects or code conversion from Matlab or Octave.

Documentation

To see how it works, go to https://VivekTRamamoorthy.github.io/MatlabJS. The library is loaded on the webpage and you may start testing it in the console. The website also contains a cheatsheet.

Examples

    tic()
    a=linspace(0,1,100) 
    disp(a) ;
    b=eye(4); 
    c=rand(4); 
    d=mul(b,c) ;
    disp(d)
    toc();

Using in your projects

Method 1: CDN (Loads the latest update)

• Include the following script tag to your html. This will include all updates and bug fixes.

<script src="https://cdn.jsdelivr.net/gh/VivekTRamamoorthy/MatlabJS/Matlab.js"></script>

This would only give the standard functions. To use plotting tools and linsolve function, additionally includeplotlib.js andndarray.js files.

Method 2: Including file using script tag

• Download theMatlab.jsfile to your project folder.
• Include the file in a script tag:

<script src="Matlab.js"></script>

Node:

Install using npm install matlabjs

Use in your projects like below

const MatlabJS = require("matlabjs")
let A = MatlabJS.zeros(10,10) // [[0,0,0,...],  [0,0,0,...], [0,0,0,...], .... ]

or

const {add, linspace} = require("matlabjs")
let A = add(linspace(0,1),200) // [ 200, 201.010101... , 202.020202... , ...]

A note of caution

• Operator overloading is not yet permitted in JavaScript and doesn't seem like it will be anytime soon. So it is not possible to writec=A*b for matrices in Javascript. Instead, one has to resort to the syntactically inconvenient universal functions, for example,c=mul(A,b).
• When using the script tag method, all functions are loaded into the global scope usingvars and users may overwrite them if needed like in Matlab. If you prefer a namespace import, use it as a module.
• The code may not be optimised for efficiency. For intensive computations, users can manually optimise the code for their use case if performance is an issue.
• This project is in its initial stage. Additional functionalities may be included as time progresses.

Contributing

You can contribute to this repository by following these steps:

• Fork the repository and clone it locally, and create a new branch.
• Include your contribution in theMatlab.js file. For example:

    var add2 = function(a){
     return a+2;
     }

• Write testing scripts for the new function intests/run_tests.js file by includingtest("your_function_name(args)","expectedoutput").
  • An example would betest("add2(4)",6)
• Test on the browser
  • Serve and openindex.html and click theRun tests button on the web page displayed. This will callrun_tests.js and print results in a popup.
• Test on Node
  • To initiate the testing suite, executenpm test from the terminal.
• If these tests pass, you may commit and raise a pull request to themain branch.

List of functions

Use left or right arrow keys to scroll if the the table is not fully visible.

Matlab function	MatlabJS equivalent	Example usage	Description
tic	tic()		Starts recording time
toc	toc()	tic(); t=toc();	Prints elapsed time since start
clc	clc()	clc()	Clears console
linspace	linspace	A=linspace(0,1) // [0 .1010.. 1] B=linspace(10,20,3) // [10,15,20]	Produces linearly spaced array.
logspace	logspace	A=logspace(1,1000,4) // [1 10 100 1000] B=logspace(1,25,3) // [1,5,25]	Produces logarithmically spaced arrays.
disp,display	disp,display	A=linspace(0,1) B=linspace(10,20,3) disp(A) display(B)	Displays matrices and arrays in console
isfield	isfield(struc,fieldname)	struc={x:10,y:100} isfield(struc,'x') // true isfield(struc,'a') // false	Checks the presence of fieldnames in a structure
size	size(A)	A=[[1,2,3],[3,4,5]] size(A) // [2,3]	Dimensions of a matrix or array
length	A.length length(A)	A=[1,2,3,3,4,5] A.length // 6 length(A) // 6	Length of an array
find	find	A=[1,2,0,0,4,5] find(A) // [1, 2, 5, 6]	Find nonzero elements
sort	sort()	A=[3,2,1,5,7]; [sortedA,indices]=sort(A); disp(sortedA); // [1,2,3,5,7] disp(indices); // [3,2,1,4,5]	Sorts numbers ascending
sum	sum	A=[1,2,3] sum(A) // 6 B=[[1,2,3],[4,5,6],[7,8,9]] disp(sum(B,1)) // column sum [12,15,18] disp(sum(B,2)) // row sum [[6],[15],[24]]	Sum of an array Column sum of a matrix Row sum of a matrix
abs	abs	A=[1,-2,3] abs(A) // [1,2,3] B=[[1,-2,3],[-4,5,6],[-7,8,-9]] disp(abs(B)) //[1,2,3],[4,5,6],[7,8,9]	Absolute value of num, array or matrix
sqrt	sqrt	A=[1,4,2] disp(sqrt(A)) // [1,2,1.414] A=rand(4) disp(A) disp(sqrt(A))	Square root of a number, array or matrix
setdiff	setdiff	A=[4,3,1,5] B=[5,3,7,8] setdiff(A,B) // [1,4]	Set difference (sorted)
min	min	A=[1,3,-5,9] disp(min(A)) // -5 B=[[1,2,3],[4,5,6],[7,8,9]] disp(B) disp(min(B,1)) // elemwise min disp(min(B,4)) // disp(min(B,[],1)) // column min disp(min(B,[],2)) // row min	Minimum of an array or matrix
max	max	Similar to min	Maximum of an array or matrix
a:b:c	range(a,b,c)	range(2,0.5,4) // [2,2.5,3,3.5,4]	Array from a to c in steps of b
triu	triu(Matrix,k)	disp(triu(rand(4))) disp(triu(rand(4),1))	Upper trianular matrix
[A, B]	concatRows(A,B)	A=ones(3,3) disp(A) B=rand(3,3) disp(B) C=concatRows(A,B) disp(C) // 3 x 6 matrix	Concatenate rows of two matrices
[A; B]	concatCols(A,B)	A=ones(3,3) disp(A) B=rand(3,3) disp(B) C=concatCols(A,B) disp(C) // 6 x 3 matrix	Concatenate columns of two matrices
A'	transpose(A)	A=[[1,2,3],[4,5,6]] transpose(A) // [[1,4],[2,5],[3,6]]	Transposes a matrix
ones	ones	disp(ones(3)) // 3x3 matrix of 1s disp(ones(3,2)) // 3x2 matrix of 1s disp(ones(3,1)) // column of 1s	Matrix of ones
eye	eye	disp(eye(3)) // 3x3 identity matrix disp(eye(4)) // 4x4 matrix of 1s disp(eye(10)) // column of 1s	Generates identity matrices
zeros	zeros	disp(zeros(3)) // 3x3 matrix of 0s disp(zeros(3,2)) // 3x2 matrix of 0s disp(zeros(3,1)) // column of 0s	Generates zero matrices
rand	rand	disp(rand()) // random no in [0,1] disp(rand(3)) // 3x3 random disp(rand(3,2)) // 3x2 random disp(rand(3,1)) // column of random	Generates matrix with values uniformly random in [0,1]
randi	randi(N,rows,cols)	disp(randi(5)) // random num in {1,2...5} disp(randi(5,3)) disp(randi(5,3,2)) // 3x2 random in {1,2...5}	Generate random integer matrices with values in[0,1,2,...,N]
diag	diag(D)	disp(diag([5,3,2])) // returns: // [ [5, 0, 0], // [0, 3, 0], // [0, 0, 2] ]	Diagonal matrix from an array
reshape	reshape	reshape([1,2,3,4,5,6],2,3) // [1,2,3; 4,5,6]	Reshape a vector or a matrix
Getting values A(rowrange,colrange) A(1:3,1:3) A(10,10) A(end,end-1)	get(A,rowrange,colrange) get(A,[1,2,3],[1,2,3]) get(A,10,10) get(A,0,-1)	A=rand(10,10);disp(A) B=get(A,[1,2,3],[2,5,7]) disp(B) B=get(A,':',[1,2,3]) disp(B) // gets all rows & first 3 cols	Get values of a submatrix
Setting values in matrices A(rowrange,colrange)=B A(1:3,1:3)=B A(end-5:end,:)=2 a(arrayrange)=b a(1:3)=[10 20 30] a(1)=10 a(end-2)=8	set(A,rowrange,colrange,B) set(A,[1,2,3],[1,2,3],B) set(A,range(-5,0),':',2) set(a,arrayrange,b) set(a,[1,2,3],[10,20,30]) set(a,1,10) set(a,-2,8)	Matrix example: A=rand(5,5) set(A,range(1,3),range(1,3),0) disp(A) // sets first 3 rows and cols to 0 set( A, range(1,3), range(1,3), randi(2,3,3) ) disp(A) // sets first 3 rows and cols // to random in {1,2} Array example: A=[1,2,3,4,5,6] set(A,2,10)// A(2)=10 set(A,0,100) // A(end)=20 set(A,-1,20) // A(end-1)=20 disp(A) // [1, 10, 3, 4, 20, 100]	Set values to a submatrix
repmat	repmat(mat,rows,cols)	A=rand(2,3) B=repmat(A,4,5) disp(B)	Repeat matrix
kron	kron(X,Y)	A=[[1,2,3],[2,3,4]]; Y=[[1],[1],[1]]; display(kron(A,Y))	Kronecker tensor product
union	union(X,Y)	A=[1,2,3,4]; B=[5,3,10]; display(union(A,B)) //[1,2,3,4,5,10]	Union of two sets
unique	unique(A)	A=[10,2,3,3,4]; display(unique(A)) //[2,3,4,10]	Unique items of a set
sparse(I,J,K)	sparse(I,J,K,nRows,nCols)	A=sparse([1,2],[1,2],[10,10],10,10); disp(A)	Initiate a sparse matrix I - row indices J - column indices K - element values nRows - no of rows nCols - no of cols
B=A	B=copy(A)	A=rand(4) disp(A) B=A B[0][0]=20 disp(A) // note: A changes // when B is changed C=copy(A); C[0][0]=100 disp(A) disp(C)	For array and matrices B=A; will not actually copy A but only creates a reference Use B=copy(A) instead.
C=A+B	C=add(A,B)	disp(add(3,4)) disp(add(ones(4,1),100)) disp(add(100,rand(1,4))) disp(add(ones(4),100)) disp(add(100,rand(4))) disp(add(ones(4),rand(4)))	Universal add for number + number array + number array + array number + matrix
C=A-B	C=sub(A,B)	disp(sub(3,4)) disp(sub(ones(4,1),100)) disp(sub(100,rand(1,4))) disp(sub(ones(4),100)) disp(sub(100,rand(4))) disp(sub(ones(4),rand(4)))	Universal subtract for number - number array - number array - array number - matrix
C=A*B	C=mul(A,B)	disp(mul(3,4)) disp(mul(ones(4,1),100)) disp(mul(100,rand(1,4))) disp(mul(ones(4),100)) disp(mul(100,rand(4))) disp(mul(ones(4),rand(4))) disp(mul(eye(4),rand(4))) disp(mul(rand(5,4),rand(4,3))) disp(mul(rand(5),rand(5,1))) disp(mul(rand(1,10),rand(10,1)))	Universal multiply for number * number array * number array * array matrix(n by k) * matrix(k by m)
C=A.*B	C=dotmul(A,B)	disp(dotmul(3,4)) A=ones(4) B=rand(4) disp(dotmul(rand(10,1),rand(10,1))) disp(dotmul(A,B)) disp(dotmul(eye(4),B))	Elementwise multiply for number .* number array .* array matrix .* matrix
C=A/B	C=div(A,B)	disp(div(3,4)) disp(div(ones(4,1),100)) disp(div(100,rand(1,4))) disp(div(ones(4),100)) disp(div(100,rand(4))) disp(div(ones(4),rand(4)))	Universal divide for number / number array / number array / array number / matrix
C=A./B	C=dotdiv(A,B)	A=rand(1,4) B=mul(100,ones(1,4)) disp(dotdiv(A,B)) C=add(rand(10),1) disp(dotdiv(rand(10),C)) disp(dotdiv(eye(4),rand(4)))	Elementwise divide for number ./ number array ./ array matrix ./ matrix
C=A^B	C=pow(A,B)	disp(pow(3,4)) disp(pow(ones(4,1),100)) disp(pow(100,rand(1,4))) disp(pow(ones(4),100)) disp(pow(100,rand(4))) disp(pow(ones(4),rand(4))) disp(pow(eye(4),rand(4)))	Universal power for number ^ number array ^ number array ^ array number ^ matrix matrix ^ number matrix ^ matrix
A(:)	colon(A)	disp(A=rand(4)) disp(colon(A))	List all columns as vector
x=A\b	x=linsolve(A,b)	A=[[2,3,4],[1,1,1],[1,0,1]] b=[[9],[3],[2]] x=linsolve(A,b) disp(x) disp(mul(A,x))	Linear solve (or) mldivide solve a sys of linear equations Uses ndarrayjs Include ndarray.js file in your project.
all	all	all([[true,true],[true,false]]) // false all([true,true],[true,true]]) // true	If all elements are true, returns true Multi-dimensional
any	any	any([[false,false],[false,false]]) // false any([false,false],[true,false]]) // true	If any element is true, returns true Multi-dimensional
Not available	map(function, arg1,arg2,arg3...)	map(x=>x>3, [1,2,3,4,5]) // [false,false,false,true,true] map(x=>x>3, [[1,2,3],[4,5,6]]) // [[false,false,false], [true,true,true]] map((a,b)=>a>b, [[1,2,3],[4,5,6]],3) // [[false,false,false], [true,true,true]] map((a,b)=>a+b, [[1,2,3],[4,5,6]],3,[[1,2,3],[4,5,6]] ) // [[5,7,9],[11,13,15]]	Multi dimensional map: Applies function across all elements, If any arg is not an array, uses the value instead Considers two arguments at a time until the last
exp	exp(number or complex or array or complex array	exp(1) // 2.718.. a=exp(new cx(0,Math.PI/2)) disp(a)// 0+1i i.e. a.re=0 a.im=1	Multi dimensional universal exponention function

License

MIT License 2.0

Vivek Thaminni Ramamoorthy