Find the min/max element of an array in JavaScript - javascript

How can I easily obtain the min or max element of a JavaScript array?
Example pseudocode:
let array = [100, 0, 50]
array.min() //=> 0
array.max() //=> 100

How about augmenting the built-in Array object to use Math.max/Math.min instead:
Array.prototype.max = function() {
return Math.max.apply(null, this);
};
Array.prototype.min = function() {
return Math.min.apply(null, this);
};
let p = [35,2,65,7,8,9,12,121,33,99];
console.log(`Max value is: ${p.max()}` +
`\nMin value is: ${p.min()}`);
Here is a JSFiddle.
Augmenting the built-ins can cause collisions with other libraries (some see), so you may be more comfortable with just apply'ing Math.xxx() to your array directly:
var min = Math.min.apply(null, arr),
max = Math.max.apply(null, arr);
Alternately, assuming your browser supports ECMAScript 6, you can use spread syntax which functions similarly to the apply method:
var min = Math.min( ...arr ),
max = Math.max( ...arr );

var max_of_array = Math.max.apply(Math, array);
For a full discussion see:
http://aaroncrane.co.uk/2008/11/javascript_max_api/

Using spread operator (ES6)
Math.max(...array) // The same with "min" => Math.min(...array)
const array = [10, 2, 33, 4, 5];
console.log(
Math.max(...array)
)

For big arrays (~10⁷ elements), Math.min and Math.max both produces the following error in Node.js.
RangeError: Maximum call stack size exceeded
A more robust solution is to not add every element to the call stack, but to instead pass an array:
function arrayMin(arr) {
return arr.reduce(function (p, v) {
return ( p < v ? p : v );
});
}
function arrayMax(arr) {
return arr.reduce(function (p, v) {
return ( p > v ? p : v );
});
}
If you are concerned about speed, the following code is ~3 times faster then Math.max.apply is on my computer. See https://jsben.ch/JPOyL.
function arrayMin(arr) {
var len = arr.length, min = Infinity;
while (len--) {
if (arr[len] < min) {
min = arr[len];
}
}
return min;
};
function arrayMax(arr) {
var len = arr.length, max = -Infinity;
while (len--) {
if (arr[len] > max) {
max = arr[len];
}
}
return max;
};
If your arrays contains strings instead of numbers, you also need to coerce them into numbers. The below code does that, but it slows the code down ~10 times on my machine. See https://jsben.ch/uPipD.
function arrayMin(arr) {
var len = arr.length, min = Infinity;
while (len--) {
if (Number(arr[len]) < min) {
min = Number(arr[len]);
}
}
return min;
};
function arrayMax(arr) {
var len = arr.length, max = -Infinity;
while (len--) {
if (Number(arr[len]) > max) {
max = Number(arr[len]);
}
}
return max;
};

tl;dr
// For regular arrays:
var max = Math.max(...arrayOfNumbers);
// For arrays with tens of thousands of items:
let max = testArray[0];
for (let i = 1; i < testArrayLength; ++i) {
if (testArray[i] > max) {
max = testArray[i];
}
}
MDN solution
The official MDN docs on Math.max() already covers this issue:
The following function uses Function.prototype.apply() to find the maximum element in a numeric array. getMaxOfArray([1, 2, 3]) is equivalent to Math.max(1, 2, 3), but you can use getMaxOfArray() on programmatically constructed arrays of any size.
function getMaxOfArray(numArray) {
return Math.max.apply(null, numArray);
}
Or with the new spread operator, getting the maximum of an array becomes a lot easier.
var arr = [1, 2, 3];
var max = Math.max(...arr);
Maximum size of an array
According to MDN the apply and spread solutions had a limitation of 65536 that came from the limit of the maximum number of arguments:
But beware: in using apply this way, you run the risk of exceeding the JavaScript engine's argument length limit. The consequences of applying a function with too many arguments (think more than tens of thousands of arguments) vary across engines (JavaScriptCore has hard-coded argument limit of 65536), because the limit (indeed even the nature of any excessively-large-stack behavior) is unspecified. Some engines will throw an exception. More perniciously, others will arbitrarily limit the number of arguments actually passed to the applied function. To illustrate this latter case: if such an engine had a limit of four arguments (actual limits are of course significantly higher), it would be as if the arguments 5, 6, 2, 3 had been passed to apply in the examples above, rather than the full array.
They even provide a hybrid solution which doesn't really have good performance compared to other solutions. See performance test below for more.
In 2019 the actual limit is the maximum size of the call stack. For modern Chromium based desktop browsers this means that when it comes to finding min/max with apply or spread, practically the maximum size for numbers only arrays is ~120000. Above this, there will be a stack overflow and the following error will be thrown:
RangeError: Maximum call stack size exceeded
With the script below (based on this blog post), by catching that error you can calculate the limit for your specific environment.
Warning! Running this script takes time and depending on the performance of your system it might slow or crash your browser/system!
let testArray = Array.from({length: 10000}, () => Math.floor(Math.random() * 2000000));
for (i = 10000; i < 1000000; ++i) {
testArray.push(Math.floor(Math.random() * 2000000));
try {
Math.max.apply(null, testArray);
} catch (e) {
console.log(i);
break;
}
}
Performance on large arrays
Based on the test in EscapeNetscape's comment I created some benchmarks that tests 5 different methods on a random number only array with 100000 items.
In 2019, the results show that the standard loop (which BTW doesn't have the size limitation) is the fastest everywhere. apply and spread comes closely after it, then much later MDN's hybrid solution then reduce as the slowest.
Almost all tests gave the same results, except for one where spread somewhy ended up being the slowest.
If you step up your array to have 1 million items, things start to break and you are left with the standard loop as a fast solution and reduce as a slower.
JSPerf benchmark
JSBen benchmark
JSBench.me benchmark
Benchmark source code
var testArrayLength = 100000
var testArray = Array.from({length: testArrayLength}, () => Math.floor(Math.random() * 2000000));
// ES6 spread
Math.min(...testArray);
Math.max(...testArray);
// reduce
testArray.reduce(function(a, b) {
return Math.max(a, b);
});
testArray.reduce(function(a, b) {
return Math.min(a, b);
});
// apply
Math.min.apply(Math, testArray);
Math.max.apply(Math, testArray);
// standard loop
let max = testArray[0];
for (let i = 1; i < testArrayLength; ++i) {
if (testArray[i] > max) {
max = testArray[i];
}
}
let min = testArray[0];
for (let i = 1; i < testArrayLength; ++i) {
if (testArray[i] < min) {
min = testArray[i];
}
}
// MDN hibrid soltuion
// Source: https://developer.mozilla.org/en-US/docs/Web/JavaScript/Reference/Global_Objects/Function/apply#Using_apply_and_built-in_functions
function minOfArray(arr) {
var min = Infinity;
var QUANTUM = 32768;
for (var i = 0, len = arr.length; i < len; i += QUANTUM) {
var submin = Math.min.apply(null, arr.slice(i, Math.min(i + QUANTUM, len)));
min = Math.min(submin, min);
}
return min;
}
minOfArray(testArray);
function maxOfArray(arr) {
var max = -Infinity;
var QUANTUM = 32768;
for (var i = 0, len = arr.length; i < len; i += QUANTUM) {
var submax = Math.max.apply(null, arr.slice(i, Math.max(i + QUANTUM, len)));
max = Math.max(submax, max);
}
return max;
}
maxOfArray(testArray);

If you're paranoid like me about using Math.max.apply (which could cause errors when given large arrays according to MDN), try this:
function arrayMax(array) {
return array.reduce(function(a, b) {
return Math.max(a, b);
});
}
function arrayMin(array) {
return array.reduce(function(a, b) {
return Math.min(a, b);
});
}
Or, in ES6:
function arrayMax(array) {
return array.reduce((a, b) => Math.max(a, b));
}
function arrayMin(array) {
return array.reduce((a, b) => Math.min(a, b));
}
The anonymous functions are unfortunately necessary (instead of using Math.max.bind(Math) because reduce doesn't just pass a and b to its function, but also i and a reference to the array itself, so we have to ensure we don't try to call max on those as well.

Alternative Methods
The Math.min and Math.max are great methods to get the minimum and maximum item out of a collection of items, however it's important to be aware of some cavities that can comes with it.
Using them with an array that contains large number of items (more than ~10⁷ items, depends on the user's browser) most likely will crash and give the following error message:
const arr = Array.from(Array(1000000).keys());
Math.min(arr);
Math.max(arr);
Uncaught RangeError: Maximum call stack size exceeded
UPDATE
Latest browsers might return NaN instead. That might be a better way to handle errors, however it doesn't solve the problem just yet.
Instead, consider using something like so:
function maxValue(arr) {
return arr.reduce((max, val) => max > val ? max : val)
}
Or with better run-time:
function maxValue(arr) {
let max = arr[0];
for (let val of arr) {
if (val > max) {
max = val;
}
}
return max;
}
Or to get both Min and Max:
function getMinMax(arr) {
return arr.reduce(({min, max}, v) => ({
min: min < v ? min : v,
max: max > v ? max : v,
}), { min: arr[0], max: arr[0] });
}
Or with even better run-time*:
function getMinMax(arr) {
let min = arr[0];
let max = arr[0];
let i = arr.length;
while (i--) {
min = arr[i] < min ? arr[i] : min;
max = arr[i] > max ? arr[i] : max;
}
return { min, max };
}
* Tested with 1,000,000 items:
Just for a reference, the 1st function run-time (on my machine) was 15.84ms vs 2nd function with only 4.32ms.

Two ways are shorter and easy:
let arr = [2, 6, 1, 0]
Way 1:
let max = Math.max.apply(null, arr)
Way 2:
let max = arr.reduce(function(a, b) {
return Math.max(a, b);
});

.apply is often used when the intention is to invoke a variadic function with a list of argument values, e.g.
The Math.max([value1[,value2, ...]]) function returns the largest of zero or more numbers.
Math.max(10, 20); // 20
Math.max(-10, -20); // -10
Math.max(-10, 20); // 20
The Math.max() method doesn't allow you to pass in an array. If you have a list of values of which you need to get the largest, you would normally call this function using Function.prototype.apply(), e.g.
Math.max.apply(null, [10, 20]); // 20
Math.max.apply(null, [-10, -20]); // -10
Math.max.apply(null, [-10, 20]); // 20
However, as of the ECMAScript 6 you can use the spread operator:
The spread operator allows an expression to be expanded in places where multiple arguments (for function calls) or multiple elements (for array literals) are expected.
Using the spread operator, the above can be rewritten as such:
Math.max(...[10, 20]); // 20
Math.max(...[-10, -20]); // -10
Math.max(...[-10, 20]); // 20
When calling a function using the variadic operator, you can even add additional values, e.g.
Math.max(...[10, 20], 50); // 50
Math.max(...[-10, -20], 50); // 50
Bonus:
Spread operator enables you to use the array literal syntax to create new arrays in situations where in ES5 you would need to fall back to imperative code, using a combination of push, splice, etc.
let foo = ['b', 'c'];
let bar = ['a', ...foo, 'd', 'e']; // ['a', 'b', 'c', 'd', 'e']

You do it by extending the Array type:
Array.max = function( array ){
return Math.max.apply( Math, array );
};
Array.min = function( array ){
return Math.min.apply( Math, array );
};
Boosted from here (by John Resig)

A simple solution to find the minimum value over an Array of elements is to use the Array prototype function reduce:
A = [4,3,-9,-2,2,1];
A.reduce((min, val) => val < min ? val : min, A[0]); // returns -9
or using JavaScript's built-in Math.Min() function (thanks #Tenflex):
A.reduce((min,val) => Math.min(min,val), A[0]);
This sets min to A[0], and then checks for A[1]...A[n] whether it is strictly less than the current min. If A[i] < min then min is updated to A[i]. When all array elements has been processed, min is returned as the result.
EDIT: Include position of minimum value:
A = [4,3,-9,-2,2,1];
A.reduce((min, val) => val < min._min ? {_min: val, _idx: min._curr, _curr: min._curr + 1} : {_min: min._min, _idx: min._idx, _curr: min._curr + 1}, {_min: A[0], _idx: 0, _curr: 0}); // returns { _min: -9, _idx: 2, _curr: 6 }

For a concise, modern solution, one can perform a reduce operation over the array, keeping track of the current minimum and maximum values, so the array is only iterated over once (which is optimal). Destructuring assignment is used here for succinctness.
let array = [100, 0, 50];
let [min, max] = array.reduce(([prevMin,prevMax], curr)=>
[Math.min(prevMin, curr), Math.max(prevMax, curr)], [Infinity, -Infinity]);
console.log("Min:", min);
console.log("Max:", max);
To only find either the minimum or maximum, we can use perform a reduce operation in much the same way, but we only need to keep track of the previous optimal value. This method is better than using apply as it will not cause errors when the array is too large for the stack.
const arr = [-1, 9, 3, -6, 35];
//Only find minimum
const min = arr.reduce((a,b)=>Math.min(a,b), Infinity);
console.log("Min:", min);//-6
//Only find maximum
const max = arr.reduce((a,b)=>Math.max(a,b), -Infinity);
console.log("Max:", max);//35

Others have already given some solutions in which they augment Array.prototype. All I want in this answer is to clarify whether it should be Math.min.apply( Math, array ) or Math.min.apply( null, array ). So what context should be used, Math or null?
When passing null as a context to apply, then the context will default to the global object (the window object in the case of browsers). Passing the Math object as the context would be the correct solution, but it won't hurt passing null either. Here's an example when null might cause trouble, when decorating the Math.max function:
// decorate Math.max
(function (oldMax) {
Math.max = function () {
this.foo(); // call Math.foo, or at least that's what we want
return oldMax.apply(this, arguments);
};
})(Math.max);
Math.foo = function () {
print("foo");
};
Array.prototype.max = function() {
return Math.max.apply(null, this); // <-- passing null as the context
};
var max = [1, 2, 3].max();
print(max);
The above will throw an exception because this.foo will be evaluated as window.foo, which is undefined. If we replace null with Math, things will work as expected and the string "foo" will be printed to the screen (I tested this using Mozilla Rhino).
You can pretty much assume that nobody has decorated Math.max so, passing null will work without problems.

One more way to do it:
var arrayMax = Function.prototype.apply.bind(Math.max, null);
Usage:
var max = arrayMax([2, 5, 1]);

I am surprised not one mentiond the reduce function.
var arr = [1, 10, 5, 11, 2]
var b = arr.reduce(function(previous,current){
return previous > current ? previous:current
});
b => 11
arr => [1, 10, 5, 11, 2]

https://developer.mozilla.org/ru/docs/Web/JavaScript/Reference/Global_Objects/Math/max
function getMaxOfArray(numArray) {
return Math.max.apply(null, numArray);
}
var arr = [100, 0, 50];
console.log(getMaxOfArray(arr))
this worked for me.

This may suit your purposes.
Array.prototype.min = function(comparer) {
if (this.length === 0) return null;
if (this.length === 1) return this[0];
comparer = (comparer || Math.min);
var v = this[0];
for (var i = 1; i < this.length; i++) {
v = comparer(this[i], v);
}
return v;
}
Array.prototype.max = function(comparer) {
if (this.length === 0) return null;
if (this.length === 1) return this[0];
comparer = (comparer || Math.max);
var v = this[0];
for (var i = 1; i < this.length; i++) {
v = comparer(this[i], v);
}
return v;
}

let array = [267, 306, 108]
let longest = Math.max(...array);

I thought I'd share my simple and easy to understand solution.
For the min:
var arr = [3, 4, 12, 1, 0, 5];
var min = arr[0];
for (var k = 1; k < arr.length; k++) {
if (arr[k] < min) {
min = arr[k];
}
}
console.log("Min is: " + min);
And for the max:
var arr = [3, 4, 12, 1, 0, 5];
var max = arr[0];
for (var k = 1; k < arr.length; k++) {
if (arr[k] > max) {
max = arr[k];
}
}
console.log("Max is: " + max);

For big arrays (~10⁷ elements), Math.min and Math.max procuces a RangeError (Maximum call stack size exceeded) in node.js.
For big arrays, a quick & dirty solution is:
Array.prototype.min = function() {
var r = this[0];
this.forEach(function(v,i,a){if (v<r) r=v;});
return r;
};

For an array containing objects instead of numbers:
arr = [
{ name: 'a', value: 5 },
{ name: 'b', value: 3 },
{ name: 'c', value: 4 }
]
You can use reduce to get the element with the smallest value (min)
arr.reduce((a, b) => a.value < b.value ? a : b)
// { name: 'b', value: 3 }
or the largest value (max)
arr.reduce((a, b) => a.value > b.value ? a : b)
// { name: 'a', value: 5 }

Aside using the math function max and min, another function to use is the built in function of sort(): here we go
const nums = [12, 67, 58, 30].sort((x, y) =>
x - y)
let min_val = nums[0]
let max_val = nums[nums.length -1]

I had the same problem, I needed to obtain the minimum and maximum values of an array and, to my surprise, there were no built-in functions for arrays. After reading a lot, I decided to test the "top 3" solutions myself:
discrete solution: a FOR loop to check every element of the array against the current max and/or min value;
APPLY solution: sending the array to the Math.max and/or Math.min internal functions using apply(null,array);
REDUCE solution: recursing a check against every element of the array using reduce(function).
The test code was this:
function GetMaxDISCRETE(A)
{ var MaxX=A[0];
for (var X=0;X<A.length;X++)
if (MaxX<A[X])
MaxX=A[X];
return MaxX;
}
function GetMaxAPPLY(A)
{ return Math.max.apply(null,A);
}
function GetMaxREDUCE(A)
{ return A.reduce(function(p,c)
{ return p>c?p:c;
});
}
The array A was filled with 100,000 random integer numbers, each function was executed 10,000 times on Mozilla Firefox 28.0 on an intel Pentium 4 2.99GHz desktop with Windows Vista. The times are in seconds, retrieved by performance.now() function. The results were these, with 3 fractional digits and standard deviation:
Discrete solution: mean=0.161s, sd=0.078
APPLY solution: mean=3.571s, sd=0.487
REDUCE solution: mean=0.350s, sd=0.044
The REDUCE solution was 117% slower than the discrete solution. The APPLY solution was the worse, 2,118% slower than the discrete solution. Besides, as Peter observed, it doesn't work for large arrays (about more than 1,000,000 elements).
Also, to complete the tests, I tested this extended discrete code:
var MaxX=A[0],MinX=A[0];
for (var X=0;X<A.length;X++)
{ if (MaxX<A[X])
MaxX=A[X];
if (MinX>A[X])
MinX=A[X];
}
The timing: mean=0.218s, sd=0.094
So, it is 35% slower than the simple discrete solution, but it retrieves both the maximum and the minimum values at once (any other solution would take at least twice that to retrieve them). Once the OP needed both values, the discrete solution would be the best choice (even as two separate functions, one for calculating maximum and another for calculating minimum, they would outperform the second best, the REDUCE solution).

Iterate through, keeping track as you go.
var min = null;
var max = null;
for (var i = 0, len = arr.length; i < len; ++i)
{
var elem = arr[i];
if (min === null || min > elem) min = elem;
if (max === null || max < elem) max = elem;
}
alert( "min = " + min + ", max = " + max );
This will leave min/max null if there are no elements in the array. Will set min and max in one pass if the array has any elements.
You could also extend Array with a range method using the above to allow reuse and improve on readability. See a working fiddle at http://jsfiddle.net/9C9fU/
Array.prototype.range = function() {
var min = null,
max = null,
i, len;
for (i = 0, len = this.length; i < len; ++i)
{
var elem = this[i];
if (min === null || min > elem) min = elem;
if (max === null || max < elem) max = elem;
}
return { min: min, max: max }
};
Used as
var arr = [3, 9, 22, -7, 44, 18, 7, 9, 15];
var range = arr.range();
console.log(range.min);
console.log(range.max);

You can use the following function anywhere in your project:
function getMin(array){
return Math.min.apply(Math,array);
}
function getMax(array){
return Math.max.apply(Math,array);
}
And then you can call the functions passing the array:
var myArray = [1,2,3,4,5,6,7];
var maximo = getMax(myArray); //return the highest number

The following code works for me :
var valueList = [10,4,17,9,3];
var maxValue = valueList.reduce(function(a, b) { return Math.max(a, b); });
var minValue = valueList.reduce(function(a, b) { return Math.min(a, b); });

array.sort((a, b) => b - a)[0];
Gives you the maximum value in an array of numbers.
array.sort((a, b) => a - b)[0];
Gives you the minimum value in an array of numbers.
let array = [0,20,45,85,41,5,7,85,90,111];
let maximum = array.sort((a, b) => b - a)[0];
let minimum = array.sort((a, b) => a - b)[0];
console.log(minimum, maximum)

let arr=[20,8,29,76,7,21,9]
Math.max.apply( Math, arr ); // 76

Simple stuff, really.
var arr = [10,20,30,40];
arr.max = function() { return Math.max.apply(Math, this); }; //attach max funct
arr.min = function() { return Math.min.apply(Math, this); }; //attach min funct
alert("min: " + arr.min() + " max: " + arr.max());

Here's one way to get the max value from an array of objects. Create a copy (with slice), then sort the copy in descending order and grab the first item.
var myArray = [
{"ID": 1, "Cost": 200},
{"ID": 2, "Cost": 1000},
{"ID": 3, "Cost": 50},
{"ID": 4, "Cost": 500}
]
maxsort = myArray.slice(0).sort(function(a, b) { return b.ID - a.ID })[0].ID;

Related

min max values 2D array javascript

I've found lots of similar posts, but none yet that fully satisfy the question: How can I get the min & max values from the following 2D array in javascript?
[[1,2,3],[4,5,6],[7,8,9]]
i.e., return 1 and 9.
this question isn't quite what I'm after (as the user wants to ignore col 0), and here asks only for a 1D array.
The accepted answer here asks only for the first value of each set.
Can anyone point me to the correct (accepted) method? Many thanks!
How about flattening the array, then using Math.max.apply and Math.min.apply:
var arr = [[1,2,3],[4,5,6],[7,8,9]].reduce(function (p, c) {
return p.concat(c);
});
var max = Math.max.apply(null, arr); // 9
var min = Math.min.apply(null, arr); // 1
Try to use traditional way of finding min and max from an array,
var x = [[1, 2, 3],[4, 5, 6],[7, 8, 9]];
var min,max;
x.forEach(function(itm) {
itm.forEach(function(itmInner) {
min = (min == undefined || itmInner<min) ? itmInner : min;
max = (max == undefined || itmInner>max) ? itmInner : max;
});
});
console.log(max,min); // 9, 1
DEMO
And you can see a performance test below,
Performance Test
Use Array.prototype.reduce in order to flatten the array and Array.prototype.sort or Math.Min\Math.Max in order to find the max\min values.
var arr = [[9,2,3],[4,5,6],[7,8,1]];
var flattenedArr = arr.reduce(function(arr1, arr2) { return arr1.concat(arr2)});
var sorted = flattenedArr.sort();
var min = sorted[0];
var max = sorted[sorted.length - 1];
try this
var array = String([[1,2,3],[4,5,6],[7,8,9]]).split(",").map( function(value){ return parseInt(value); } );
var max_of_array = Math.max.apply(Math, array);
var min_of_array = Math.min.apply(Math, array);
function minMaxIn2D(arr) {
var min,max;
for(var i in arr) {
for(var j in arr[i]){
min = min - arr[i][j] <= 0 ? min : arr[i][j] ;
max = max - arr[i][j] >= 0 ? max: arr[i][j];
}
}
console.log(min, max)
}

How to find the biggest number in array

How can I find the biggest number from the numbers I've entered in array?
things i've tried already:
var largest = Math.max.apply(Math, number[i]);
var smallest = Math.min.apply(Math, number[i]);
You're very close:
Math.max.apply(Math, number);
You need to pass all of the elements of the array to the max() metho; [i] doesn't make sense here.
Math.max.apply is an option like said in other answers, but I think reduce is more common among languages:
var number=[1,2,3,2,1,1];
number.reduce(function(a,b){return a>b ? a : b;});
// returns 3
Math.max.apply(Math.max, number);
// returns 3
var max = number[0], i = 0;
for (i = 0; i < n; ++i) {
if (number[i] > max) {
max = number[i];
}
}
or I recommend underscore.js
var max = _.max(number);
You're close, try this:
Math.max.apply(Math.max, number);
Ie, the second argument converts an array to the list of arguments.
(Actually since Math.max never needs this, you don't need the first argument of apply; so Math.max.apply(null, number) or Math.max.apply('aaaaahhh', number) works too.)
Using - Array.prototype.reduce() is cool!
[267, 306, 108].reduce((acc,val)=> (acc>val)?acc:val)
where acc = accumulator and val = current value;
function compare(a, b) {
if (a > b) return 1
else if (a < b) return -1
else return 0
}
var arr = [ 1, 2, 15, 12.5, 16, 3.35 ]
arr.sort(compare)
alert( arr[2] ) // your n number
// i used sort() but because by default it sorting in lexicographical order i
passed it throuth custom comparisom , it works fine for me

Return index of greatest value in an array

I have this:
var arr = [0, 21, 22, 7];
What's the best way to return the index of the highest value into another variable?
This is probably the best way, since it’s reliable and works on old browsers:
function indexOfMax(arr) {
if (arr.length === 0) {
return -1;
}
var max = arr[0];
var maxIndex = 0;
for (var i = 1; i < arr.length; i++) {
if (arr[i] > max) {
maxIndex = i;
max = arr[i];
}
}
return maxIndex;
}
There’s also this one-liner:
let i = arr.indexOf(Math.max(...arr));
It performs twice as many comparisons as necessary and will throw a RangeError on large arrays, though. I’d stick to the function.
In one line and probably faster then arr.indexOf(Math.max.apply(Math, arr)):
var a = [0, 21, 22, 7];
var indexOfMaxValue = a.reduce((iMax, x, i, arr) => x > arr[iMax] ? i : iMax, 0);
document.write("indexOfMaxValue = " + indexOfMaxValue); // prints "indexOfMaxValue = 2"
Where:
iMax - the best index so far (the index of the max element so far, on the first iteration iMax = 0 because the second argument to reduce() is 0, we can't omit the second argument to reduce() in our case)
x - the currently tested element from the array
i - the currently tested index
arr - our array ([0, 21, 22, 7])
About the reduce() method (from "JavaScript: The Definitive Guide" by David Flanagan):
reduce() takes two arguments. The first is the function that performs the reduction operation. The task of this reduction function is to somehow combine or reduce two values into a single value, and to return that reduced value.
Functions used with reduce() are different than the functions used with forEach() and map(). The familiar value, index, and array values are passed as the second, third, and fourth arguments. The first argument is the accumulated result of the reduction so far. On the first call to the function, this first argument is the initial value you passed as the
second argument to reduce(). On subsequent calls, it is the value returned by the previous invocation of the function.
When you invoke reduce() with no initial value, it uses the first element of the array as the initial value. This means that the first call to the reduction function will have the first and second array elements as its
first and second arguments.
Another solution of max using reduce:
[1,2,5,0,4].reduce((a,b,i) => a[0] < b ? [b,i] : a, [Number.MIN_VALUE,-1])
//[5,2]
This returns [5e-324, -1] if the array is empty. If you want just the index, put [1] after.
Min via (Change to > and MAX_VALUE):
[1,2,5,0,4].reduce((a,b,i) => a[0] > b ? [b,i] : a, [Number.MAX_VALUE,-1])
//[0, 3]
To complete the work of #VFDan, I benchmarked the 3 methods: the accepted one (custom loop), reduce, and find(max(arr)) on an array of 10000 floats.
Results on chromimum 85 linux (higher is better):
custom loop: 100%
reduce: 94.36%
indexOf(max): 70%
Results on firefox 80 linux (higher is better):
custom loop: 100%
reduce: 96.39%
indexOf(max): 31.16%
Conclusion:
If you need your code to run fast, don't use indexOf(max).
reduce is ok but use the custom loop if you need the best performances.
You can run this benchmark on other browser using this link:
https://jsben.ch/wkd4c
If you are utilizing underscore, you can use this nice short one-liner:
_.indexOf(arr, _.max(arr))
It will first find the value of the largest item in the array, in this case 22. Then it will return the index of where 22 is within the array, in this case 2.
Unless I'm mistaken, I'd say it's to write your own function.
function findIndexOfGreatest(array) {
var greatest;
var indexOfGreatest;
for (var i = 0; i < array.length; i++) {
if (!greatest || array[i] > greatest) {
greatest = array[i];
indexOfGreatest = i;
}
}
return indexOfGreatest;
}
var arr=[0,6,7,7,7];
var largest=[0];
//find the largest num;
for(var i=0;i<arr.length;i++){
var comp=(arr[i]-largest[0])>0;
if(comp){
largest =[];
largest.push(arr[i]);
}
}
alert(largest )//7
//find the index of 'arr'
var arrIndex=[];
for(var i=0;i<arr.length;i++){
var comp=arr[i]-largest[0]==0;
if(comp){
arrIndex.push(i);
}
}
alert(arrIndex);//[2,3,4]
EDIT: Years ago I gave an answer to this that was gross, too specific, and too complicated. So I'm editing it. I favor the functional answers above for their neat factor but not their readability; but if I were more familiar with javascript then I might like them for that, too.
Pseudo code:
Track index that contains largest value. Assume index 0 is largest initially. Compare against current index. Update index with largest value if necessary.
Code:
var mountains = [3, 1, 5, 9, 4];
function largestIndex(array){
var counter = 1;
var max = 0;
for(counter; counter < array.length; counter++){
if(array[max] < array[counter]){
max = counter;
}
}
return max;
}
console.log("index with largest value is: " +largestIndex(mountains));
// index with largest value is: 3
function findIndicesOf(haystack, needle)
{
var indices = [];
var j = 0;
for (var i = 0; i < haystack.length; ++i) {
if (haystack[i] == needle)
indices[j++] = i;
}
return indices;
}
pass array to haystack and Math.max(...array) to needle. This will give all max elements of the array, and it is more extensible (for example, you also need to find min values)
If you create a copy of the array and sort it descending, the first element of the copy will be the largest. Than you can find its index in the original array.
var sorted = [...arr].sort((a,b) => b - a)
arr.indexOf(sorted[0])
Time complexity is O(n) for the copy, O(n*log(n)) for sorting and O(n) for the indexOf.
If you need to do it faster, Ry's answer is O(n).
A minor modification revised from the "reduce" version of #traxium 's solution taking the empty array into consideration:
function indexOfMaxElement(array) {
return array.reduce((iMax, x, i, arr) =>
arr[iMax] === undefined ? i :
x > arr[iMax] ? i : iMax
, -1 // return -1 if empty
);
}
A stable version of this function looks like this:
// not defined for empty array
function max_index(elements) {
var i = 1;
var mi = 0;
while (i < elements.length) {
if (!(elements[i] < elements[mi]))
mi = i;
i += 1;
}
return mi;
}
To find the index of the greatest value in an array, copy the original array into the new array and then sort the original array in decreasing order to get the output [22, 21, 7, 0]; now find the value 22 index in the copyNumbers array using this code copyNumbers.indexOf(numbers[0]);
<script>
const numbers = [0, 21, 22, 7];
const copyNumbers = [];
copyNumbers.push(...numbers);
numbers.sort(function(a, b){
return b - a
});
const index = copyNumbers.indexOf(numbers[0]);
console.log(index);
</script>
Make this
const max = arr.reduce((m, n) => Math.max(m, n)), then the indexes of the max
get index with findIndex
var index = arr.findIndex(i => i === max)

Max value in array, based on a function

I have an array of objects. I want to find the "maximum" of this array based on a function that returns whichever object is bigger when given 2 objects.
function comparison(first, second) {
// ... arbitrary comparison based on properties...
return first; // or second
}
var a = [obj1, obj2, obj3];
var maxObj = ????(comparison);
What do I fill in here? What's elegant and short?
Something like this should be quicker than sort (depending on the data):
/*
values: array of values to test.
fn: function that takes two arguements and returns true if the first is bigger.
*/
var maximum = function(values, fn) {
var currentValue, maxValue = values.pop();
while(values.length)
maxValue = fn(maxValue, currentValue = values.pop()) ? maxValue : currentValue;
return maxValue;
}
Examples: http://jsfiddle.net/SaBJ4/2/
Even better, use Array.reduce:
var a = ['abc', 'defg', 'highlkasd', 'ac', 'asdh'];
a.reduce(function(a, b) { return a.length > b.length ? a : b; }); // highlkasd
What's wrong with the obvious approach?
for(var i = 0, max; i < a.length; ++i)
max = typeof max == 'undefined' ? a[i] : comparison(a[i], max);
Wrap that up however you like.
Or you can take advantage of the fact that a = []; x = a[0] leaves you with undefined in x and do it RobG's way:
for(var i = 1, max = a[0]; i < a.length; ++i)
max = comparison(a[i], max);
That nicely avoids a bunch of typeof operators and comparisons that you really don't need.
[obj,obj,obj].sort(comparison)
// aka
var sorted = [obj,obj,obj].sort(function(a,b){
// return 1/0/-1
});
Then either pop the top or bottom element off (however you're sorting) to get the "max" object.
Array.Sort

Most efficient way to create a zero filled JavaScript array?

What is the most efficient way to create an arbitrary length zero filled array in JavaScript?
ES6 introduces Array.prototype.fill. It can be used like this:
new Array(len).fill(0);
Not sure if it's fast, but I like it because it's short and self-describing.
It's still not in IE (check compatibility), but there's a polyfill available.
Although this is an old thread, I wanted to add my 2 cents to it. Not sure how slow/fast this is, but it's a quick one liner. Here is what I do:
If I want to pre-fill with a number:
Array.apply(null, Array(5)).map(Number.prototype.valueOf,0);
// [0, 0, 0, 0, 0]
If I want to pre-fill with a string:
Array.apply(null, Array(3)).map(String.prototype.valueOf,"hi")
// ["hi", "hi", "hi"]
Other answers have suggested:
new Array(5+1).join('0').split('')
// ["0", "0", "0", "0", "0"]
but if you want 0 (the number) and not "0" (zero inside a string), you can do:
new Array(5+1).join('0').split('').map(parseFloat)
// [0, 0, 0, 0, 0]
In short
Fastest solution:
let a = new Array(n); for (let i=0; i<n; ++i) a[i] = 0;
Shortest (handy) solution (3x slower for small arrays, slightly slower for big (slowest on Firefox))
Array(n).fill(0)
Details
Today 2020.06.09 I perform tests on macOS High Sierra 10.13.6 on browsers Chrome 83.0, Firefox 77.0, and Safari 13.1. I test chosen solutions for two test cases
small array - with 10 elements - you can perform test HERE
big arrays - with 1M elements - you can perform test HERE
Conclusions
solution based on new Array(n)+for (N) is fastest solution for small arrays and big arrays (except Chrome but still very fast there) and it is recommended as fast cross-browser solution
solution based on new Float32Array(n) (I) returns non typical array (e.g. you cannot call push(..) on it) so I not compare its results with other solutions - however this solution is about 10-20x faster than other solutions for big arrays on all browsers
solutions based on for (L,M,N,O) are fast for small arrays
solutions based on fill (B,C) are fast on Chrome and Safari but surprisingly slowest on Firefox for big arrays. They are medium fast for small arrays
solution based on Array.apply (P) throws error for big arrays
function P(n) {
return Array.apply(null, Array(n)).map(Number.prototype.valueOf,0);
}
try {
P(1000000);
} catch(e) {
console.error(e.message);
}
Code and example
Below code presents solutions used in measurements
function A(n) {
return [...new Array(n)].fill(0);
}
function B(n) {
return new Array(n).fill(0);
}
function C(n) {
return Array(n).fill(0);
}
function D(n) {
return Array.from({length: n}, () => 0);
}
function E(n) {
return [...new Array(n)].map(x => 0);
}
// arrays with type
function F(n) {
return Array.from(new Int32Array(n));
}
function G(n) {
return Array.from(new Float32Array(n));
}
function H(n) {
return Array.from(new Float64Array(n)); // needs 2x more memory than float32
}
function I(n) {
return new Float32Array(n); // this is not typical array
}
function J(n) {
return [].slice.apply(new Float32Array(n));
}
// Based on for
function K(n) {
let a = [];
a.length = n;
let i = 0;
while (i < n) {
a[i] = 0;
i++;
}
return a;
}
function L(n) {
let a=[]; for(let i=0; i<n; i++) a[i]=0;
return a;
}
function M(n) {
let a=[]; for(let i=0; i<n; i++) a.push(0);
return a;
}
function N(n) {
let a = new Array(n); for (let i=0; i<n; ++i) a[i] = 0;
return a;
}
function O(n) {
let a = new Array(n); for (let i=n; i--;) a[i] = 0;
return a;
}
// other
function P(n) {
return Array.apply(null, Array(n)).map(Number.prototype.valueOf,0);
}
function Q(n) {
return "0".repeat( n ).split("").map( parseFloat );
}
function R(n) {
return new Array(n+1).join('0').split('').map(parseFloat)
}
// ---------
// TEST
// ---------
[A,B,C,D,E,F,G,H,I,J,K,L,M,N,O,P,Q,R].forEach(f => {
let a = f(10);
console.log(`${f.name} length=${a.length}, arr[0]=${a[0]}, arr[9]=${a[9]}`)
});
This snippets only present used codes
Example results for Chrome:
Elegant way to fill an array with precomputed values
Here is another way to do it using ES6 that nobody has mentioned so far:
> Array.from(Array(3), () => 0)
< [0, 0, 0]
It works by passing a map function as the second parameter of Array.from.
In the example above, the first parameter allocates an array of 3 positions filled with the value undefined and then the lambda function maps each one of them to the value 0.
Although Array(len).fill(0) is shorter, it doesn't work if you need to fill the array by doing some computation first (I know the question didn't ask for it, but a lot of people end up here looking for this).
For instance, if you need an array with 10 random numbers:
> Array.from(Array(10), () => Math.floor(10 * Math.random()))
< [3, 6, 8, 1, 9, 3, 0, 6, 7, 1]
It's more concise (and elegant) than the equivalent:
const numbers = Array(10);
for (let i = 0; i < numbers.length; i++) {
numbers[i] = Math.round(10 * Math.random());
}
This method can also be used to generate sequences of numbers by taking advantage of the index parameter provided in the callback:
> Array.from(Array(10), (d, i) => i)
< [0, 1, 2, 3, 4, 5, 6, 7, 8, 9]
Bonus answer: fill an array using String repeat()
Since this answer is getting a good deal of attention, I also wanted to show this cool trick. Although not as useful as my main answer, will introduce the still not very known, but very useful String repeat() method. Here's the trick:
> "?".repeat(10).split("").map(() => Math.floor(10 * Math.random()))
< [5, 6, 3, 5, 0, 8, 2, 7, 4, 1]
Cool, huh? repeat() is a very useful method to create a string that is the repetition of the original string a certain number of times. After that, split() creates an array for us, which is then map()ped to the values we want. Breaking it down in steps:
> "?".repeat(10)
< "??????????"
> "?".repeat(10).split("")
< ["?", "?", "?", "?", "?", "?", "?", "?", "?", "?"]
> "?".repeat(10).split("").map(() => Math.floor(10 * Math.random()))
< [5, 6, 3, 5, 0, 8, 2, 7, 4, 1]
The already mentioned ES 6 fill method takes care of this nicely. Most modern desktop browsers already support the required Array prototype methods as of today (Chromium, FF, Edge and Safari) [1]. You can look up details on MDN. A simple usage example is
a = new Array(10).fill(0);
Given the current browser support you should be cautious to use this unless you are sure your audience uses modern Desktop browsers.
Note added August 2013, updated February 2015: The answer below from 2009 relates to JavaScript's generic Array type. It doesn't relate to the newer typed arrays defined in ES2015 [and available now in many browsers], like Int32Array and such. Also note that ES2015 adds a fill method to both Arrays and typed arrays, which is likely to be the most efficient way to fill them...
Also, it can make a big difference to some implementations how you create the array. Chrome's V8 engine, in particular, tries to use a highly-efficient, contiguous-memory array if it thinks it can, shifting to the object-based array only when necessary.
With most languages, it would be pre-allocate, then zero-fill, like this:
function newFilledArray(len, val) {
var rv = new Array(len);
while (--len >= 0) {
rv[len] = val;
}
return rv;
}
But, JavaScript arrays aren't really arrays, they're key/value maps just like all other JavaScript objects, so there's no "pre-allocate" to do (setting the length doesn't allocate that many slots to fill), nor is there any reason to believe that the benefit of counting down to zero (which is just to make the comparison in the loop fast) isn't outweighed by adding the keys in reverse order when the implementation may well have optimized their handling of the keys related to arrays on the theory you'll generally do them in order.
In fact, Matthew Crumley pointed out that counting down is markedly slower on Firefox than counting up, a result I can confirm — it's the array part of it (looping down to zero is still faster than looping up to a limit in a var). Apparently adding the elements to the array in reverse order is a slow op on Firefox. In fact, the results vary quite a bit by JavaScript implementation (which isn't all that surprising). Here's a quick and dirty test page (below) for browser implementations (very dirty, doesn't yield during tests, so provides minimal feedback and will run afoul of script time limits). I recommend refreshing between tests; FF (at least) slows down on repeated tests if you don't.
The fairly complicated version that uses Array#concat is faster than a straight init on FF as of somewhere between 1,000 and 2,000 element arrays. On Chrome's V8 engine, though, straight init wins out every time...
Here's a test:
const tests = [
{
name: "downpre",
total: 0,
desc: "Count down, pre-decrement",
func: makeWithCountDownPre
},
{
name: "downpost",
total: 0,
desc: "Count down, post-decrement",
func: makeWithCountDownPost
},
{
name: "up",
total: 0,
desc: "Count up (normal)",
func: makeWithCountUp
},
{
name: "downandup",
total: 0,
desc: "Count down (for loop) and up (for filling)",
func: makeWithCountDownArrayUp
},
{
name: "concat",
total: 0,
desc: "Concat",
func: makeWithConcat
}
];
const q = sel => document.querySelector(sel);
let markup = "";
for (const {name, desc} of tests) {
markup += `
<div><input type="checkbox" id="chk_${name}" checked>
<label for="chk_${name}">${desc}</label></div>`;
}
q("#checkboxes").innerHTML = markup;
q("#btnTest").addEventListener("click", btnTestClick);
function btnTestClick() {
// Clear log
q("#log").innerHTML = "Testing...";
// Show running
q("#btnTest").disabled = true;
// Run after a pause while the browser updates display
setTimeout(btnTestClickPart2, 0);
}
function btnTestClickPart2() {
try {
runTests();
} catch (e) {
log(`Exception: ${e.message}`);
}
// Re-enable the button
q("#btnTest").disabled = false;
}
function getNumField(name) {
const val = q("#" + name).value.trim();
const num = /^\d+$/.test(val) ? parseInt(val) : NaN;
if (isNaN(num) || num <= 0) {
throw new Error(`Invalid ${name} value ${JSON.stringify(val)}`);
}
return num;
}
function runTests() {
try {
// Clear log
q("#log").innerHTML = "";
const runCount = getNumField("loops");
const length = getNumField("length");
// Do it (we run runCount + 1 times, first time is a warm up)
for (let counter = 0; counter <= runCount; ++counter) {
for (const test of tests) {
if (q("#chk_" + test.name).checked) {
const start = Date.now();
const a = test.func(length);
const time = Date.now() - start;
if (counter == 0) {
// Don't count (warm up), but do check the algorithm works
const invalid = validateResult(a, length);
if (invalid) {
log(`<span class=error>FAILURE</span> with test ${test.name}: ${invalid}`);
return;
}
} else {
// Count this one
log(`#${counter}: ${test.desc}: ${time}ms`);
test.total += time;
}
}
}
}
for (const test of tests) {
if (q("#chk_" + test.name).checked) {
test.avg = test.total / runCount;
if (typeof lowest != "number" || lowest > test.avg) {
lowest = test.avg;
}
}
}
let results =
"<p>Results:" +
"<br>Length: " + length +
"<br>Loops: " + runCount +
"</p>";
for (const test of tests) {
if (q("#chk_" + test.name).checked) {
results +=
`<p ${lowest == test.avg ? " class=winner" : ""}>${test.desc}, average time: ${test.avg}ms</p>`;
}
}
results += "<hr>";
q("#log").insertAdjacentHTML("afterbegin", results);
} catch (e) {
log(e.message);
return;
}
}
function validateResult(a, length) {
if (a.length != length) {
return "Length is wrong";
}
for (let n = length - 1; n >= 0; --n) {
if (a[n] != 0) {
return "Index " + n + " is not zero";
}
}
return undefined;
}
function makeWithCountDownPre(len) {
const a = new Array(len);
while (--len >= 0) {
a[len] = 0;
}
return a;
}
function makeWithCountDownPost(len) {
const a = new Array(len);
while (len-- > 0) {
a[len] = 0;
}
return a;
}
function makeWithCountUp(len) {
const a = new Array(len);
for (let i = 0; i < len; ++i) {
a[i] = 0;
}
return a;
}
function makeWithCountDownArrayUp(len) {
const a = new Array(len);
let i = 0;
while (--len >= 0) {
a[i++] = 0;
}
return a;
}
function makeWithConcat(len) {
if (len == 0) {
return [];
}
let a = [0];
let currlen = 1;
while (currlen < len) {
const rem = len - currlen;
if (rem < currlen) {
a = a.concat(a.slice(0, rem));
} else {
a = a.concat(a);
}
currlen = a.length;
}
return a;
}
function log(msg) {
const p = document.createElement("p");
p.textContent = msg;
q("#log").appendChild(p);
}
body {
font-family: sans-serif;
}
#log p {
margin: 0;
padding: 0;
}
.error {
color: red;
}
.winner {
color: green;
}
<div>
<label for='txtLength'>Length:</label><input type='text' id='length' value='1000'>
<br><label for='txtLoops'>Loops:</label><input type='text' id='loops' value='100000'>
<div id='checkboxes'></div>
<br><input type='button' id='btnTest' value='Test'>
<hr>
<div id='log'></div>
</div>
If you use ES6, you can use Array.from() like this:
Array.from({ length: 3 }, () => 0);
//[0, 0, 0]
Has the same result as
Array.from({ length: 3 }).map(() => 0)
//[0, 0, 0]
Because
Array.from({ length: 3 })
//[undefined, undefined, undefined]
By default Uint8Array, Uint16Array and Uint32Array classes keep zeros as its values, so you don't need any complex filling techniques, just do:
var ary = new Uint8Array(10);
all elements of array ary will be zeros by default.
function makeArrayOf(value, length) {
var arr = [], i = length;
while (i--) {
arr[i] = value;
}
return arr;
}
makeArrayOf(0, 5); // [0, 0, 0, 0, 0]
makeArrayOf('x', 3); // ['x', 'x', 'x']
Note that while is usually more efficient than for-in, forEach, etc.
using object notation
var x = [];
zero filled? like...
var x = [0,0,0,0,0,0];
filled with 'undefined'...
var x = new Array(7);
obj notation with zeros
var x = [];
for (var i = 0; i < 10; i++) x[i] = 0;
As a side note, if you modify Array's prototype, both
var x = new Array();
and
var y = [];
will have those prototype modifications
At any rate, I wouldn't be overly concerned with the efficiency or speed of this operation, there are plenty of other things that you will likely be doing that are far more wasteful and expensive than instanciating an array of arbitrary length containing zeros.
I've tested all combinations of pre-allocating/not pre-allocating, counting up/down, and for/while loops in IE 6/7/8, Firefox 3.5, Chrome, and Opera.
The functions below was consistently the fastest or extremely close in Firefox, Chrome, and IE8, and not much slower than the fastest in Opera and IE 6. It's also the simplest and clearest in my opinion. I've found several browsers where the while loop version is slightly faster, so I'm including it too for reference.
function newFilledArray(length, val) {
var array = [];
for (var i = 0; i < length; i++) {
array[i] = val;
}
return array;
}
or
function newFilledArray(length, val) {
var array = [];
var i = 0;
while (i < length) {
array[i++] = val;
}
return array;
}
ES6 solution:
[...new Array(5)].map(x => 0); // [0, 0, 0, 0, 0]
const arr = Array.from({ length: 10 }).fill(0);
console.log(arr)
If you need to create many zero filled arrays of different lengths during the execution of your code, the fastest way I've found to achieve this is to create a zero array once, using one of the methods mentioned on this topic, of a length which you know will never be exceeded, and then slice that array as necessary.
For example (using the function from the chosen answer above to initialize the array), create a zero filled array of length maxLength, as a variable visible to the code that needs zero arrays:
var zero = newFilledArray(maxLength, 0);
Now slice this array everytime you need a zero filled array of length requiredLength < maxLength:
zero.slice(0, requiredLength);
I was creating zero filled arrays thousands of times during execution of my code, this speeded up the process tremendously.
function zeroFilledArray(size) {
return new Array(size + 1).join('0').split('');
}
Using lodash or underscore
_.range(0, length - 1, 0);
Or if you have an array existing and you want an array of the same length
array.map(_.constant(0));
I have nothing against:
Array.apply(null, Array(5)).map(Number.prototype.valueOf,0);
new Array(5+1).join('0').split('').map(parseFloat);
suggested by Zertosh, but in a new ES6 array extensions allow you to do this natively with fill method. Now IE edge, Chrome and FF supports it, but check the compatibility table
new Array(3).fill(0) will give you [0, 0, 0]. You can fill the array with any value like new Array(5).fill('abc') (even objects and other arrays).
On top of that you can modify previous arrays with fill:
arr = [1, 2, 3, 4, 5, 6]
arr.fill(9, 3, 5) # what to fill, start, end
which gives you: [1, 2, 3, 9, 9, 6]
To create an all new Array
new Array(arrayLength).fill(0);
To add some values at the end of an existing Array
[...existingArray, ...new Array(numberOfElementsToAdd).fill(0)]
Example
//**To create an all new Array**
console.log(new Array(5).fill(0));
//**To add some values at the end of an existing Array**
let existingArray = [1,2,3]
console.log([...existingArray, ...new Array(5).fill(0)]);
The way I usually do it (and is amazing fast) is using Uint8Array. For example, creating a zero filled vector of 1M elements:
var zeroFilled = [].slice.apply(new Uint8Array(1000000))
I'm a Linux user and always have worked for me, but once a friend using a Mac had some non-zero elements. I thought his machine was malfunctioning, but still here's the safest way we found to fix it:
var zeroFilled = [].slice.apply(new Uint8Array(new Array(1000000))
Edited
Chrome 25.0.1364.160
Frederik Gottlieb - 6.43
Sam Barnum - 4.83
Eli - 3.68
Joshua 2.91
Mathew Crumley - 2.67
bduran - 2.55
Allen Rice - 2.11
kangax - 0.68
Tj. Crowder - 0.67
zertosh - ERROR
Firefox 20.0
Allen Rice - 1.85
Joshua - 1.82
Mathew Crumley - 1.79
bduran - 1.37
Frederik Gottlieb - 0.67
Sam Barnum - 0.63
Eli - 0.59
kagax - 0.13
Tj. Crowder - 0.13
zertosh - ERROR
Missing the most important test (at least for me): the Node.js one. I suspect it close to Chrome benchmark.
As of ECMAScript2016, there is one clear choice for large arrays.
Since this answer still shows up near the top on google searches, here's an answer for 2017.
Here's a current jsbench with a few dozen popular methods, including many proposed up to now on this question. If you find a better method please add, fork and share.
I want to note that there is no true most efficient way to create an arbitrary length zero filled array. You can optimize for speed, or for clarity and maintainability - either can be considered the more efficient choice depending on the needs of the project.
When optimizing for speed, you want to: create the array using literal syntax; set the length, initialize iterating variable, and iterate through the array using a while loop. Here's an example.
const arr = [];
arr.length = 120000;
let i = 0;
while (i < 120000) {
arr[i] = 0;
i++;
}
Another possible implementation would be:
(arr = []).length = n;
let i = 0;
while (i < n) {
arr[i] = 0;
i++;
}
But I strongly discourage using this second implantation in practice as it's less clear and doesn't allow you to maintain block scoping on your array variable.
These are significantly faster than filling with a for loop, and about 90% faster than the standard method of
const arr = Array(n).fill(0);
But this fill method is still the most efficient choice for smaller arrays due to it's clarity, conciseness and maintainability. The performance difference likely won't kill you unless you're making a lot of arrays with lengths on the order of thousands or more.
A few other important notes. Most style guides recommend you no longer use varwithout a very special reason when using ES6 or later. Use const for variables that won't be redefined and let for variables that will. The MDN and Airbnb's Style Guide are great places to go for more information on best practices. The questions wasn't about syntax, but it's important that people new to JS know about these new standards when searching through these reams of old and new answers.
Didn't see this method in answers, so here it is:
"0".repeat( 200 ).split("").map( parseFloat )
In result you will get zero-valued array of length 200:
[ 0, 0, 0, 0, ... 0 ]
I'm not sure about the performance of this code, but it shouldn't be an issue if you use it for relatively small arrays.
What about new Array(51).join('0').split('')?
let filled = [];
filled.length = 10;
filled.fill(0);
console.log(filled);
This concat version is much faster in my tests on Chrome (2013-03-21). About 200ms for 10,000,000 elements vs 675 for straight init.
function filledArray(len, value) {
if (len <= 0) return [];
var result = [value];
while (result.length < len/2) {
result = result.concat(result);
}
return result.concat(result.slice(0, len-result.length));
}
Bonus: if you want to fill your array with Strings, this is a concise way to do it (not quite as fast as concat though):
function filledArrayString(len, value) {
return new Array(len+1).join(value).split('');
}
I was testing out the great answer by T.J. Crowder, and came up with a recursive merge based on the concat solution that outperforms any in his tests in Chrome (i didn't test other browsers).
function makeRec(len, acc) {
if (acc == null) acc = [];
if (len <= 1) return acc;
var b = makeRec(len >> 1, [0]);
b = b.concat(b);
if (len & 1) b = b.concat([0]);
return b;
},
call the method with makeRec(29).
It might be worth pointing out, that Array.prototype.fill had been added as part of the ECMAScript 6 (Harmony) proposal. I would rather go with the polyfill written below, before considering other options mentioned on the thread.
if (!Array.prototype.fill) {
Array.prototype.fill = function(value) {
// Steps 1-2.
if (this == null) {
throw new TypeError('this is null or not defined');
}
var O = Object(this);
// Steps 3-5.
var len = O.length >>> 0;
// Steps 6-7.
var start = arguments[1];
var relativeStart = start >> 0;
// Step 8.
var k = relativeStart < 0 ?
Math.max(len + relativeStart, 0) :
Math.min(relativeStart, len);
// Steps 9-10.
var end = arguments[2];
var relativeEnd = end === undefined ?
len : end >> 0;
// Step 11.
var final = relativeEnd < 0 ?
Math.max(len + relativeEnd, 0) :
Math.min(relativeEnd, len);
// Step 12.
while (k < final) {
O[k] = value;
k++;
}
// Step 13.
return O;
};
}
Shortest for loop code
a=i=[];for(;i<100;)a[i++]=0;
edit:
for(a=i=[];i<100;)a[i++]=0;
or
for(a=[],i=100;i--;)a[i]=0;
Safe var version
var a=[],i=0;for(;i<100;)a[i++]=0;
edit:
for(var i=100,a=[];i--;)a[i]=0;
My fastest function would be:
function newFilledArray(len, val) {
var a = [];
while(len--){
a.push(val);
}
return a;
}
var st = (new Date()).getTime();
newFilledArray(1000000, 0)
console.log((new Date()).getTime() - st); // returned 63, 65, 62 milliseconds
Using the native push and shift to add items to the array is much faster (about 10 times) than declaring the array scope and referencing each item to set it's value.
fyi: I consistently get faster times with the first loop, which is counting down, when running this in firebug (firefox extension).
var a = [];
var len = 1000000;
var st = (new Date()).getTime();
while(len){
a.push(0);
len -= 1;
}
console.log((new Date()).getTime() - st); // returned 863, 894, 875 milliseconds
st = (new Date()).getTime();
len = 1000000;
a = [];
for(var i = 0; i < len; i++){
a.push(0);
}
console.log((new Date()).getTime() - st); // returned 1155, 1179, 1163 milliseconds
I'm interested to know what T.J. Crowder makes of that ? :-)
I knew I had this proto'd somewhere :)
Array.prototype.init = function(x,n)
{
if(typeof(n)=='undefined') { n = this.length; }
while (n--) { this[n] = x; }
return this;
}
var a = (new Array(5)).init(0);
var b = [].init(0,4);
Edit: tests
In response to Joshua and others methods I ran my own benchmarking, and I'm seeing completely different results to those reported.
Here's what I tested:
//my original method
Array.prototype.init = function(x,n)
{
if(typeof(n)=='undefined') { n = this.length; }
while (n--) { this[n] = x; }
return this;
}
//now using push which I had previously thought to be slower than direct assignment
Array.prototype.init2 = function(x,n)
{
if(typeof(n)=='undefined') { n = this.length; }
while (n--) { this.push(x); }
return this;
}
//joshua's method
function newFilledArray(len, val) {
var a = [];
while(len--){
a.push(val);
}
return a;
}
//test m1 and m2 with short arrays many times 10K * 10
var a = new Date();
for(var i=0; i<10000; i++)
{
var t1 = [].init(0,10);
}
var A = new Date();
var b = new Date();
for(var i=0; i<10000; i++)
{
var t2 = [].init2(0,10);
}
var B = new Date();
//test m1 and m2 with long array created once 100K
var c = new Date();
var t3 = [].init(0,100000);
var C = new Date();
var d = new Date();
var t4 = [].init2(0,100000);
var D = new Date();
//test m3 with short array many times 10K * 10
var e = new Date();
for(var i=0; i<10000; i++)
{
var t5 = newFilledArray(10,0);
}
var E = new Date();
//test m3 with long array created once 100K
var f = new Date();
var t6 = newFilledArray(100000, 0)
var F = new Date();
Results:
IE7 deltas:
dA=156
dB=359
dC=125
dD=375
dE=468
dF=412
FF3.5 deltas:
dA=6
dB=13
dC=63
dD=8
dE=12
dF=8
So by my reckoning push is indeed slower generally but performs better with longer arrays in FF but worse in IE which just sucks in general (quel surprise).
Anonymous function:
(function(n) { while(n-- && this.push(0)); return this; }).call([], 5);
// => [0, 0, 0, 0, 0]
A bit shorter with for-loop:
(function(n) { for(;n--;this.push(0)); return this; }).call([], 5);
// => [0, 0, 0, 0, 0]
Works with any Object, just change what's inside this.push().
You can even save the function:
function fill(size, content) {
for(;size--;this.push(content));
return this;
}
Call it using:
var helloArray = fill.call([], 5, 'hello');
// => ['hello', 'hello', 'hello', 'hello', 'hello']
Adding elements to an already existing array:
var helloWorldArray = fill.call(helloArray, 5, 'world');
// => ['hello', 'hello', 'hello', 'hello', 'hello', 'world', 'world', 'world', 'world', 'world']
Performance: http://jsperf.com/zero-filled-array-creation/25

Categories