var a = 1;
var b = Number(1);
var c = new Number(1);

I was wondering what is the difference between these three statements. I understand that first and second statements are same, as if(a===b) gives true, but the third one will create a object of type number.

What I want to know is how these methods are different, and any advantages one will give over the other?

var a = 1;
var b = Number(1);
var c = new Number(1);

I was wondering what is the difference between these three statements. I understand that first and second statements are same, as if(a===b) gives true, but the third one will create a object of type number.

What I want to know is how these methods are different, and any advantages one will give over the other?

• javascript

• If you could provide some context for your question it might help get you some better answers. – Pointy Commented Sep 10, 2012 at 15:51

3 Answers 3

A value like 1 is a primitive, not an object. JavaScript will generally promote numbers to Number objects when necessary. There's rarely a reason to explicitly construct one, and there's certainly no particular "advantage". There's also no reason for something like Number(1), though the Number constructor is one of several ways of coercing a value to be a number.

In short, non: the new String() and new Number() constructors are to be ignored if you want to save yourself a world of trouble.
The first two methods you present here assign a numeric constant to the variable, the third way -as you say- creates an object. The value of that object will be 1, but you can change that value without loosing any specific methods you set to the object.

There aren't very many advantages to storing numbers or strings in objects. AFAIK, the only thing you "gain" is a very, very, very slight performance difference over the constants when invoking certain methods, like toExponential etc...
In my view, that isn't worth the trouble of creating objects for all numbers you're bound to use. I think of it as one of the bad parts of JS, meant to make the language look familiar to Java Applet developers.

The second, without the new keyword, allows you to sort-of-type-cast: Number(document.getElementById('formElem').value) === 123; and has its uses (mainly with Date objects, in my experience). But then again, converting to a number can be achieved using the + operator, too: +document.getElementById('formElem').value) === 123

On the whole, just stay well clear of these primitive constructors. The only reason they're still there is because they're objects, and therefore have prototypes. Now THAT is an advantage:

Number.prototype.addOneToString = function()
{
    return (1+this).toString();
};
String.prototype.UpperFirst = function()
{
    return this.charAt(0).toUpperCase() + this.slice(1);
}
var foo = new Number(3);
foo.addOneToString();//returns "4"
foo = new String('foo');
foo.UpperFirst();//Foo

Since JS wraps constant operands in an instance of its object counterpart, when the statement requires it, you can apply prototype methods (either native or self-made ones) to any constant. (Thanks to Pointy for this, and +1)

(3).addOneToString();//"4"
'foo'.UpperFirst();//Foo

So just regard them as legacy quirks, that are still there because of their prototypes.

In ES12 and after you can use the below ways to declare large numbers.

const oneMillion = 1_000_000;

var oneMillion1 = 1_000_000;

you can use separators _, for bitter readability