CODEWITHSUNDEEP
javascriptjquery
Mark Pieszak - Trilon.io
Mark Pieszak - Trilon.io

Reputation: 66971

$.each() vs for() loop - and performance

These are mainly just some things I've been wondering, maybe someone can give me a little more insight on them, i'll share what i've noticed so far as well!

First thing i've been wondering... is there any difference good or reason to use:

$('element').each(function (i, el) { });

-- versus --

$.each($('element'), function (i, el) { });

Looking at the jQuery docs I can't see any rhyme or reason for one or the other (maybe you know an instance or additional things one can do over the other.

But more importantly I'm concerned with speed here

// As opposed to $.each() looping through a jQuery object
// -- 8x faster 
for (var i = 0, $('.whatever').length; i < len; i++) {
    $('.whatever')[i] // do stuff
}

If you check out this jsFiddle DEMO here, you'll see the difference in speed is basically equivalent with either of them, but more importantly I feel like I should always be using for() loops...

I was just unit testing (looping through each of 5 different scenario functions, 50,000 times), simply looping through a bunch of list items, and setting a data-newAttr, nothing special.

QUESTION :: I guess my biggest question is, why not always use for loops while iterating through an object?? Is there even a point to using $.each()? Do you always use for() loops even when going through jQuery objects?

jsFiddle DEMO here

Function type:                  Execution Time:
_testArea.each() + $(this)               1947   <-- using $(this) slows it down tremendously
$.each()         + $(this)               1940
_testArea.each() + el(plain JS)           458   <-- using the Element speeds things up
$.each()         + el(plain JS)           452
for() loop       + plainJS[0] iteration   236   <-- over 8x faster

Just my 2cents. :)

Upvotes: 75

Views: 77597

Answers (6)

Yeti
Yeti

Reputation: 2845

jQuery.each vs for-loop

Advantages jQuery.each:

  • Fits well in jQuery code (chaining and style).
  • No worries about scope (references to the iterator and object will be persistent).
  • Can be used universally (for all kinds of objects and iterating over object keys).

Advantages for-loop:

  • High performance (for games/animations/large datasets).
  • Full control over iterator (skip items, splice items from list, etc).
  • The for-loop will always work, for there is no dependency on jQuery.
  • Same familiar syntax as most other similar languages.

Example code

This is example code of how I prefer to iterate over a list.

var list = ['a', 'b', 'c', 'd', 'e', 'f'];

jQuery.each:

$.each(list, function(i, v)
{
    // code...
});

For-loop without closure:

for(var i=0,v,n=list.length;i<n;i+=1)
{
    v = list[i];
    // code...
}

For-loop with closure:

for(var i=0,n=list.length;i<n;i+=1)
{
    (function(i, v)
    {
        // code...
    })(i, list[i]);
}

Note: I suggest you just use the standard for-loop, and only use a closure when necessary. However, when your code looks more like jQuery than Javascript anyway, it could be easier to just use $.each. In case of performance issues, you could always look into it afterwards.

Upvotes: 9

ryanneufeld
ryanneufeld

Reputation: 157

Actually there is a big difference between $.each() and $().each().

They do slightly different things depending on what you're passing in.

http://api.jquery.com/each/ vs http://api.jquery.com/jquery.each/

jquery.each is a generic iterator, where $().each() is specific to a jquery collection.

Also see: http://jsperf.com/each-vs-each-vs-for-in/9

Upvotes: 2

op1ekun
op1ekun

Reputation: 1928

I ran some simple performance test while ago http://jsperf.com/forloops3. Seems that sticking to plain, old for loop (where it's possible) is the way to go :)

Upvotes: 5

Peter-Paul van Gemerden
Peter-Paul van Gemerden

Reputation: 7011

One thing that .each() allows you to do that can't be done with a for loop is chaining.

$('.rows').each(function(i, el) {
    // do something with ALL the rows
}).filter('.even').each(function(i, el) {
    // do something with the even rows
});

I played around with your JSFiddle to see how chaining would influence performance in cases where you have to loop through subsets of the original set of matched elements.

The result wasn't all that unexpected, although I think the overhead of end() was exaggerated here because of the combination of few elements and many loops. Other than that: plain JS loops are still slightly faster, but whether that weighs up to the added readability of .each() (and chaining) is debatable.

Upvotes: 49

JayC
JayC

Reputation: 7141

One thing you do get with .each() is automatic local scoping (because you are invoking an anonymous function for every object), which in turn means if you are creating even more anonymous functions/closures/event handlers/whatever on every iteration, you never have to worry about your handlers sharing a variable. That is, JavaScript doesn't act like other languages when it comes to local scopes, but because you can declare a variable anywhere, it can fool you sometimes.

In other words, this is wrong:

var idx,el;
for (idx = 0; idx <someObjectArray.length; idx++){
   el = someObjectArray[idx]
   el.someEventHandler(function(){  
       alert( "this is element " + idx);
   }); 
}

Whenever any of those objects invoke their "someEvent" after this loop (please note that this is made up), the alert is always going to say whatever was last assigned to idx, which should be (as of the time invoked) someObjectArray.length;

To make sure you save the proper index, you have to declare a local scope, create a variable and assign to that variable for use.

var idx,el;
for (idx = 0; idx <someObjectArray.length; idx++){
   el = someObjectArray[idx];
   (function(){
       var localidx = idx;
       el.someEventHandler(function(){  
           alert( "this is element " + localidx);
       }); 
   })();
}

As you can see, that's as ugly as hell, but it should work. Each event handler gets its own copy of localidx

Now compare that to .each()

$(someObjectArray).each(function (idx, el) { 
   el.someEventHandler(function(){  
       alert( "this is element " + idx);
   }); 
});

A lot simpler, isn't it?

Upvotes: 22

James Black
James Black

Reputation: 41858

When I went to your link here are two numbers I got:

$.each() + el(plain JS) 401
for() loop + plainJS[0] iteration   377

If the difference is that small, then go with the one that is most readable, but, if you have very high time requirements, then you may just need to go with what ends up being the fastest.

I would suggest you write your program to use three different methods, the two above, and then use the foreach found in newer versions of javascript, and for those browsers that don't support it you can add it as a prototype.

https://developer.mozilla.org/en-US/docs/JavaScript/Reference/Global_Objects/Array/forEach

You know what your requirements are, and what your program will do, so just write your own tests and ensure it meets the requirements across at the browsers you will be supporting.

For your first question, I would go with $('element').each as it is much easier to read, but that just my opinion.

Upvotes: 4

Related Questions