Javascript prototype operator performance: saves memory, but is it faster?

I read here (Douglas Crockford) using prototype operator to add methods to Javascript classes saves also memory.

Then I read in this John Resig's article "Instantiating a function with a bunch of prototype properties is very, very, fast", but is he talking about using prototype in the standard way, or is he talking about his specific example in his article?

For example, is creating this object:

function Class1()
{
   this.showMsg = function(string) { alert(string); }
}
var c = new Class1();
c.showMsg();

slower than creating this object, then?

function Class1() {}
Class1.prototype.showMsg = function(string) { alert(string); }
var c = new Class1();
c.showMsg();

P.S.

I know prototype is used to create inheritance and singleton object etc. But this question does not have anyhting to do with these subjects.


EDIT: to whom it might be interested also in performance comparison between a JS object and a JS static objet can read this answer below. Static object are definitely faster, obviously they can be usued only when you don't need more than one instance of the object.

Answers:

Answer

It was an interesting question, so I ran some very simple tests (I should have restarted my browsers to clear out the memory, but I didn't; take this for what it's worth). It looks like at least on Safari and Firefox, prototype runs significantly faster [edit: not 20x as stated earlier]. I'm sure a real-world test with fully-featured objects would be a better comparison. The code I ran was this (I ran the tests several times, separately):

var X,Y, x,y, i, intNow;

X = function() {};
X.prototype.message = function(s) { var mymessage = s + "";}
X.prototype.addition = function(i,j) { return (i *2 + j * 2) / 2; }

Y = function() {
    this.message = function(s) { var mymessage = s + "";}
    this.addition = function(i,j) { return (i *2 + j * 2) / 2; }
};


intNow = (new Date()).getTime();
for (i = 0; i < 1000000; i++) {
    y = new Y();
    y.message('hi');
    y.addition(i,2)
}
console.log((new Date()).getTime() - intNow); //FF=5206ms; Safari=1554

intNow = (new Date()).getTime();
for (i = 0; i < 1000000; i++) {
    x = new X();
    x.message('hi');
    x.addition(i,2)
}
console.log((new Date()).getTime() - intNow);//FF=3894ms;Safari=606

It's a real shame, because I really hate using prototype. I like my object code to be self-encapsulated, and not allowed to drift. I guess when speed matters, though, I don't have a choice. Darn.

[Edit] Many thanks to @Kevin who pointed out my previous code was wrong, giving a huge boost to the reported speed of the prototype method. After fixing, prototype is still around significantly faster, but the difference is not as enormous.

Answer

I would guess that it depends on the type of object you want to create. I ran a similar test as Andrew, but with a static object, and the static object won hands down. Here's the test:

var X,Y,Z,x,y,z;

X = function() {};
X.prototype.message = function(s) { var mymessage = s + "";}
X.prototype.addition = function(i,j) { return (i *2 + j * 2) / 2; }

Y = function() {
    this.message = function(s) { var mymessage = s + "";}
    this.addition = function(i,j) { return (i *2 + j * 2) / 2; }
};

Z = {
 message: function(s) { var mymessage = s + "";}
 ,addition: function(i,j) { return (i *2 + j * 2) / 2; }
}

function TestPerformance()
{
  var closureStartDateTime = new Date();
  for (var i = 0; i < 100000; i++)
  {
 y = new Y();
    y.message('hi');
    y.addition(i,2);
  }
  var closureEndDateTime = new Date();

  var prototypeStartDateTime = new Date();
  for (var i = 0; i < 100000; i++)
  {
    x = new X();
    x.message('hi');
    x.addition(i,2);
  }
  var prototypeEndDateTime = new Date();

  var staticObjectStartDateTime = new Date();
  for (var i = 0; i < 100000; i++)
  {
 z = Z; // obviously you don't really need this
    z.message('hi');
    z.addition(i,2);
  }
  var staticObjectEndDateTime = new Date();
  var closureTime = closureEndDateTime.getTime() - closureStartDateTime.getTime();
  var prototypeTime = prototypeEndDateTime.getTime() - prototypeStartDateTime.getTime();
  var staticTime = staticObjectEndDateTime.getTime() - staticObjectStartDateTime.getTime();
  console.log("Closure time: " + closureTime + ", prototype time: " + prototypeTime + ", static object time: " + staticTime);
}

TestPerformance();

This test is a modification of code I found at:

http://blogs.msdn.com/b/kristoffer/archive/2007/02/13/javascript-prototype-versus-closure-execution-speed.aspx

Results:

IE6: closure time: 1062, prototype time: 766, static object time: 406

IE8: closure time: 781, prototype time: 406, static object time: 188

FF: closure time: 233, prototype time: 141, static object time: 94

Safari: closure time: 152, prototype time: 12, static object time: 6

Chrome: closure time: 13, prototype time: 8, static object time: 3

The lesson learned is that if you DON'T have a need to instantiate many different objects from the same class, then creating it as a static object wins hands down. So think carefully about what kind of class you really need.

Answer

So I decided to test this as well. I tested creation time, execution time, and memory use. I used Nodejs v0.8.12 and the mocha test framework running on a Mac Book Pro booted into Windows 7. The 'fast' results are using prototypes and the 'slow' ones are using module pattern. I created 1 million of each type of object and then accessed the 4 methods in each object. Here are the results:

c:\ABoxAbove>mocha test/test_andrew.js

Fast Allocation took:170 msec
·Fast Access took:826 msec
state[0] = First0
Free Memory:5006495744

·Slow Allocation took:999 msec
·Slow Access took:599 msec
state[0] = First0
Free Memory:4639649792

Mem diff:358248k
Mem overhead per obj:366.845952bytes

? 4 tests complete (2.6 seconds)

The code is as follows:

var assert = require("assert"), os = require('os');

function Fast (){}
Fast.prototype = {
    state:"",
    getState:function (){return this.state;},
    setState:function (_state){this.state = _state;},
    name:"",
    getName:function (){return this.name;},
    setName:function (_name){this.name = _name;}
};

function Slow (){
    var state, name;
    return{
        getState:function (){return this.state;},
        setState:function (_state){this.state = _state;},
        getName:function (){return this.name;},
        setName:function (_name){this.name = _name;}
    };
}
describe('test supposed fast prototype', function(){
    var count = 1000000, i, objs = [count], state = "First", name="Test";
    var ts, diff, mem;
    it ('should allocate a bunch of objects quickly', function (done){
        ts = Date.now ();
        for (i = 0; i < count; ++i){objs[i] = new Fast ();}
        diff = Date.now () - ts;
        console.log ("Fast Allocation took:%d msec", diff);
        done ();
    });
    it ('should access a bunch of objects quickly', function (done){
        ts = Date.now ();
        for (i = 0; i < count; ++i){
            objs[i].setState (state + i);
            assert (objs[i].getState () === state + i, "States should be equal");
            objs[i].setName (name + i);
            assert (objs[i].getName () === name + i, "Names should be equal");
        }
        diff = Date.now() - ts;
        console.log ("Fast Access took:%d msec", diff);
        console.log ("state[0] = " + objs[0].getState ());
        mem = os.freemem();
        console.log ("Free Memory:" + mem + "\n");
        done ();
    });
    it ('should allocate a bunch of objects slowly', function (done){
        ts = Date.now ();
        for (i = 0; i < count; ++i){objs[i] = Slow ();}
        diff = Date.now() - ts;
        console.log ("Slow Allocation took:%d msec", diff);
        done ();
    });
    it ('should access a bunch of objects slowly', function (done){
        ts = Date.now ();
        for (i = 0; i < count; ++i){
            objs[i].setState (state + i);
            assert (objs[i].getState () === state + i, "States should be equal");
            objs[i].setName (name + i);
            assert (objs[i].getName () === name + i, "Names should be equal");
        }
        diff = Date.now() - ts;
        console.log ("Slow Access took:%d msec", diff);
        console.log ("state[0] = " + objs[0].getState ());
        var mem2 = os.freemem();
        console.log ("Free Memory:" + mem2 + "\n");
        console.log ("Mem diff:" + (mem - mem2) / 1024 + "k");
        console.log ("Mem overhead per obj:" + (mem - mem2) / count + 'bytes');
        done ();
    });
});

Conclusion: This backs up what others in this post have found. If you are constantly creating objects then the prototype mechanism is clearly faster. If your code spends most of its time accessing objects then the module pattern is faster. If you are sensitive about memory use, the prototype mechanism uses ~360 bytes less per object.

Answer

Intuitively, it seems that it would be more memory-efficient and faster to create functions on the prototype: the function's only created once, not each time a new instance is created.

However, there will be a slight performance difference when it's time to access the function. When c.showMsg is referenced, the JavaScript runtime first checks for the property on c. If it's not found, c's prototype is then checked.

So, creating the property on the instance would result in slightly faster access time - but this might only be an issue for a very deep prototype hierarchy.

Answer

We need to separate object construction and usage.

When declaring a function on a prototype, it is shared between all instances. When declaring a function in a constructor, this is recreated every time new instance is made. Given that, we need to benchmark construction and usage separately to have better results. That is what I did and want to share the results with you. This benchmark does not test for speed of construction.

function ThisFunc() {
    this.value = 0;
    this.increment = function(){
        this.value++;
    }
}

function ProtFunc() {
    this.value = 0;
}

ProtFunc.prototype.increment = function (){
    this.value++;
}

function ClosFunc() {
    var value = 0;

    return {
        increment:function(){
            value++;
        }
    };
}

var thisInstance = new ThisFunc;

var iterations = 1000000;
var intNow = (new Date()).getTime();
for (i = 0; i < iterations; i++) {
    thisInstance.increment();
}
console.log(`ThisFunc: ${(new Date()).getTime() - intNow}`); // 27ms node v4.6.0

var protInstance = new ProtFunc;
intNow = (new Date()).getTime();
for (i = 0; i < iterations; i++) {
    protInstance.increment();
}
console.log(`ProtFunc: ${(new Date()).getTime() - intNow}`); // 4ms node v4.6.0

var closInstance = ClosFunc();
intNow = (new Date()).getTime();
for (i = 0; i < iterations; i++) {
    closInstance.increment();
}
console.log(`ClosFunc: ${(new Date()).getTime() - intNow}`); // 7ms node v4.6.0

From these results we can see that the prototype version is the fastest (4ms), but the closure version is very close (7ms). You may still need to benchmark for your particular case.

So:

  • We can use prototype version when we need to have every bit of performance or share functions between instances.
  • We can use other versions when what we want is the features they provide. (private state encapsulation, readability etc.)

PS: I used Andrew's answer as a reference. Used the same loops and notation.

Answer

High Resolution Browser Performance API Tests

None of the tests here are taking advantage of the performance API for high resolution testing so I wrote one that will show current fastest results for many different scenarios including 2 that are faster than any of the other answers on most runs.

Fasted in each category (10,000 iterations)

  • Property access only (~0.5ms): { __proto__: Type }
  • Looping object creation with property access (<3ms): Object.create(Type)

The code uses ES6 without babel transpilation to ensure accuracy. It works in current chrome. Run the test below to see the breakdown.

function profile () {
  function test ( name
                , define
                , construct
                , { index = 0
                  , count = 10000
                  , ordinals = [ 0, 1 ]
                  , constructPrior = false
                  } = {}
                ) {
    performance.clearMarks()
    performance.clearMeasures()
    const symbols = { type: Symbol('type') }
    const marks = (
      { __proto__: null
      , start: `${name}_start`
      , define: `${name}_define`
      , construct: `${name}_construct`
      , end: `${name}_end`
      }
    )

    performance.mark(marks.start)
    let Type = define()
    performance.mark(marks.define)

    let obj = constructPrior ? construct(Type) : null
    do {
      if(!constructPrior)
        obj = construct(Type)
      if(index === 0)
        performance.mark(marks.construct)

      const measureOrdinal = ordinals.includes(index)
      if(measureOrdinal)
          performance.mark(`${name}_ordinal_${index}_pre`)

      obj.message('hi')
      obj.addition(index, 2)

      if(measureOrdinal)
        performance.mark(`${name}_ordinal_${index}_post`)
    } while (++index < count)
    performance.mark(marks.end)

    const measureMarks = Object.assign (
      { [`${name}_define`]: [ marks.start, marks.define ]
      , [`${name}_construct`]: [ marks.define, marks.construct ]
      , [`${name}_loop`]: [ marks.construct, marks.end ]
      , [`${name}_total`]: [ marks.start, marks.end ]
      }
    , ordinals.reduce((reduction, i) => Object.assign(reduction, { [`${name}_ordinal_${i}`]: [ `${name}_ordinal_${i}_pre`, `${name}_ordinal_${i}_post` ] }), {})
    )

    Object.keys(measureMarks).forEach((key) => performance.measure(key, ...measureMarks[key]))

    const measures = performance.getEntriesByType('measure').map(x => Object.assign(x, { endTime: x.startTime + x.duration }))
    measures.sort((a, b) => a.endTime - b.endTime)
    const durations = measures.reduce((reduction, measure) => Object.assign(reduction, { [measure.name]: measure.duration }), {})

    return (
      { [symbols.type]: 'profile'
      , profile: name
      , duration: durations[`${name}_total`]
      , durations
      , measures
      }
    )
  }

  const refs = (
    { __proto__: null
    , message: function(s) { var mymessage = s + '' }
    , addition: function(i, j) { return (i *2 + j * 2) / 2 }
    }
  )

  const testArgs = [
    [ 'constructor'
    , function define() {
        return function Type () {
          this.message = refs.message
          this.addition = refs.addition
        }
      }
    , function construct(Type) {
        return new Type()
      }
    ]
  , [ 'prototype'
    , function define() {
        function Type () {
        }
        Type.prototype.message = refs.message
        Type.prototype.addition = refs.addition
        return Type
      }
    , function construct(Type) {
        return new Type()
      }
    ]
  , [ 'Object.create'
    , function define() {
        return (
          { __proto__: null
          , message: refs.message
          , addition: refs.addition
          }
        )
      }
    , function construct(Type) {
        return Object.create(Type)
      }
    ]
  , [ 'proto'
    , function define() {
        return (
          { __proto__: null
          , message: refs.message
          , addition: refs.addition
          }
        )
      }
    , function construct(Type) {
        return { __proto__: Type }
      }
    ]
  ]

  return testArgs.reduce(
    (reduction, [ name, ...args ]) => (
      Object.assign( reduction
      , { [name]: (
            { normal: test(name, ...args, { constructPrior: true })
            , reconstruct: test(`${name}_reconstruct`, ...args, { constructPrior: false })
            }
          )
        }
      )
    )
  , {})
}

let profiled = profile()
const breakdown = Object.keys(profiled).reduce((reduction, name) => [ ...reduction, ...Object.keys(profiled[name]).reduce((r, type) => [ ...r, { profile: `${name}_${type}`, duration: profiled[name][type].duration } ], []) ], [])
breakdown.sort((a, b) => a.duration - b.duration)
try {
  const Pre = props => React.createElement('pre', { children: JSON.stringify(props.children, null, 2) })
  
  ReactDOM.render(React.createElement(Pre, { children: { breakdown, profiled } }), document.getElementById('profile'))
} catch(err) {
    console.error(err)
}
<script src="https://cdnjs.cloudflare.com/ajax/libs/react/15.1.0/react.min.js"></script>
<script src="https://cdnjs.cloudflare.com/ajax/libs/react/15.1.0/react-dom.min.js"></script>

<div id="profile"></div>

Answer

I'm sure that as far as instantiating the object goes, it's way faster and also consumes less memory, no doubts about that, but I would think that the javascript engine needs to loop through all the properties of the object to determine if the property/method invoked is part of that object and if not, then go check for the prototype. I am not 100% sure about this but I'm assuming that's how it works and if so, then in SOME cases where your object has a LOT of methods added to it, instantiated only once and used heavily, then it could possibly be a little slower, but that's just a supposition I haven't tested anything.

But in the end, I would still agree that as a general rules, using prototype will be faster.

Answer

So, creating the property on the instance would result in slightly faster access time - but this might only be an issue for a very deep prototype hierarchy.

Actually the result is different then we could expect - access time to prototyped methods is faster then accessing to the methods attached exactly to the object (FF tested).

Answer

I ran my own tests.

The first conclusion is, that static access is actually slower than real prototyping. Interestingly, the Version 23 of this test has a flawed prototyping (Variable X) in it, which just returns the completely overridden prototype object over and over again and when I was creating my test, this prototyping was still slower than my "real prototype" test.

Anyway, to the answer: Unless my test is flawed, it shows that real prototyping is fastest. It beats or is at least equal to the static object when ignoring instantiation. this-assignments on instantiation and private variables are both much slower. I wouldn't have guessed private variables would be this slow.

It might be of interest that I extended the prototype Object with jQuery.extend in between and it was about the same speed as the direct assignment. The extend was outside the test itself, of course. At least this is a way to circumvent writing annoying ".prototype."-Parts all the time.

Tags

Recent Questions

Top Questions

Home Tags Terms of Service Privacy Policy DMCA Contact Us

©2020 All rights reserved.