If you're seeing this message, it means we're having trouble loading external resources on our website.

If you're behind a web filter, please make sure that the domains *.kastatic.org and *.kasandbox.org are unblocked.

# Static functions vs. instance methods

Before we get to Algorithm #3 (accelerate towards the mouse), we need to cover one more rather important aspect of working with vectors and the `PVector` object: the difference between using static functions and instance methods.
Forgetting about vectors for a moment, take a look at the following code:
``````var x = 0;
var y = 5;
x = x + y;``````
Pretty simple, right? `x` has the value of 0, we add `y` to it, and now `x` is equal to 5. We could write the corresponding code pretty easily based on what we’ve learned about `PVector`.
``````var v = new PVector(0,0);
var u = new PVector(4,5);
The vector v has the value of (0,0), we add u to it, and now v is equal to (4,5). Easy, right?
Let’s take a look at another example of some simple math:
``````var x = 0;
var y = 5;
var z = x + y;``````
`x` has the value of 0, we add `y` to it, and store the result in a new variable `z`. The value of `x` does not change in this example, and neither does y! This may seem like a trivial point, and one that is quite intuitive when it comes to mathematical operations with numbers. However, it’s not so obvious with mathematical operations in `PVector`. Let’s try to write the code based on what we know so far.
``````var v = new PVector(0,0);
var u = new PVector(4,5);
var w = v.add(u); // Don’t be fooled; this is incorrect!!!``````
The above might seem like a good guess, but it’s just not the way the `PVector` object works. If we look at the definition of `add()`...
``````PVector.prototype.add = function(v) {
this.x = this.x + v.x;
this.y = this.y + v.y;
};``````
...we see that this code does not accomplish our goal. First, it does not return a new `PVector` (there is no `return` statement) and second, it changes the value of the `PVector` upon which it is called. In order to add two `PVector` objects together and return the result as a new `PVector`, we must use the "static" `add()` function.
A "static" function is a function that is defined on an object, but it doesn't change properties of the object. So why even define it on the object? Typically, it has something to do with the object, so it is logical to attach it to it. It treats the object more like a namespace. For example, all the static functions on `PVector` perform some sort of manipulation on passed in `PVector` objects and always return back some value. We could define those functions globally as well, but this way, we avoid global functions and have better ways of grouping related functionality.
Let's contrast. Here's how we use the `add()` instance method:
``v.add(u);``
That line of code would modify `v`, so we wouldn't need to save a return value. Conversely, here's how we use the `add()` static function:
``var w = PVector.add(v, u);``
If we didn't save the result of that function into a variable, that line of code would be useless, because the static version doesn't change the objects themselves. `PVector`'s static functions allow us to perform generic mathematical operations on `PVector` objects without having to adjust the value of one of the input `PVectors`.
Here's how we would write the static version of `add()`:
``````PVector.add = function(v1, v2) {
var v3 = new PVector(v1.x + v2.x, v1.y + v2.y);
return v3;
};``````
There are several differences here:
• We define the function directly on the object, not on its prototype
• We never access the `this` keyword inside the function
• We return a value from the function
The PVector object has static versions of `add()`, `sub()`, `mult()`, and `div()`. It also has additional static functions that don't exist as instance methods, like `angleBetween()`, `dot()`, and `cross()`. We'll find ourselves using these functions as we continue making programs with `PVector`.

This "Natural Simulations" course is a derivative of "The Nature of Code" by Daniel Shiffman, used under a Creative Commons Attribution-NonCommercial 3.0 Unported License.

## Want to join the conversation?

• I think I understand, but I'm not completely sure. Please correct my summary if something seems off.

Instance methods alter their object variable while static functions return the value after manipulation without altering. Is that the only major difference between the two types aside from implementation details?

The other question is about calling directly on the object not the prototype. This means I have to use the actual object name, in this case PVector.add(v, u), and not an instance of the object like velocity.add or vector1.add. •  Instance methods don't necessarily alter variables. It's simply a function that gets "shipped out", so to speak, with each instance of that object you create. They are relative to the individual object, hence the keyword `this`.

A static method belongs to the class itself. You can call it without instantiating any members of that class. That's why the method is called from the class name and not from a specific instance.
• var v = new PVector(1,5);
var u = PVector.mult( v , 2);
var w = PVector.sub( v , u);
w.div(3);

why only in first line "new" is used before <i>Pvector</i>?? • The line
``var v = new PVector(1,5);``
is creating a new PVector object

The lines
`` var u = PVector.mult(v, 2);var w = PVector.sub(v, u);``

are using the static functions included with PVector
and then storing the result which is a "new PVector" inside of a variable.

remember that when we use
``v.mult(u)``

we are changing the value of v. There is no value that is actually returned so it can't be stored inside of another variable.

on the other hand
`` var u = PVector.mult(v, 2);``

is a function, that takes two parameters, and returns a PVector object, storing it inside of "u"

The new is used inside of the function itself.
• I can't seem to get the first step in the static challenge. Here's my code. Any help.

var v = new PVector(1,5);
var w = 2;
PVector.mult = function(u,w) {
var u = new PVector(v.x.mult.w,v.y.mult.w);
return u; }; • So, the way I was thinking is that there is no use for a static function that does not return a value. Then I thought about passing things by reference as opposed to by value. Can you do this sort of thing in JavaScript? • I am having some trouble figuring out the Static function challenge.
Could I see someone's spin off, to see what I did wrong?
(1 vote) • I have been having trouble with this one, but it seems to be related to the brower. I've found that reloading the entire page seems to make things I'm fairly certain of actually pass, though I'm stymied at the end and can't tell if it's an error or a code problem.

For the first step, however, I used:

var u = PVector.mult(v, 2);
• In the Mouse stalker challenge, I have worked out I need to use this code:
`var maxDir = new PVector(width - 1, height - 1);//maximum vector to mousevar maxMag = maxDir.mag();`
But why this allowed? maxDir.mag() looks like an instance method to me. Why does it return a value? I was expecting to use something like `var maxMag = PVector.mag(maxDir);` • It's hard for others to explain your misunderstandings...

`mag` is one of many methods associated with PVector instances. It returns the magnitude (or length) of the PVector instance. Lengths are simple numbers, not vectors.

Oddly enough, `PVector` does supply a `mag` static function. It's mechanically generated in the same manner as `get`, `add`, `sub`, etc. It is pretty useless in that it always returns a clone of its first argument, the same as invoking `PVector.get`
• I m trying to understand the implication of" 'We define the function directly on the object, not on its prototype'. Then, a second related question is why instances of PVector don't inherit these static methods too.

There seem to be (at least) 3 ways of adding methods to an object (call it Vector):
1. var Vector = function(x, y) {
this. x = x,
this.y = y,
2. Vector.add = function(a, b) .....
3. Vector.prototype.add = function(a, b) .....

Do they do the same thing? I guess though only the first way (1.) makes the method automatically inheritable in any object based on Vector. For way 3, you have to call Vector's prototype into the new object's prototype. What does 2. do ? I thought it might be equivalent to 1. which would make it automatically inheritable, or not? • Way 1 and way 3 are similar, but way 2 has nothing to do with `this` object. Way 2 is deployed as a way to keep from "polluting the global namespace". One could have written `var VectorAdd(a, b) {...` and been done, but by hanging another property (a static function) onto `Vector` one avoids another `var` and signals that the function is intimately associated with the `Vector` function.

The non-static functions that reference `this` object are "methods". If all their functionality is derived from `this`, then the two ways are functionally equivalent. However, the storage requirements differ. By placing the method in the `prototype` (way 3) one guarantees that a single copy of the method is shared by all constructed objects. If one defines the method within the constructor function (way 1), then each object gets its own unique copy of that function. For example,

``    var a = new Vector(3, 4);    var b = new Vector(12, 5);    prinltn(a.add === b.add);``

prints false with way 1 and true with way 3. So until you know and the uses of "closures", place your methods on the `prototype` property.
• It was not really clear to me what the difference is between defining the function directly on the object or defining it on its its prototype. What happens on the background and why is this difference so important? • Defining another function on the object constructor is no big deal. For example. we could have had something like
``/* * Return a new PVector that is the  * summation of PVectors a and b. */var addPVectors = function(a, b) {    ...};``
But instead we use `PVector.add` is do exactly the same thing. The latter does not pollute the global name space with a bunch more names. It also makes it clear the function is naturally provided by the one, same library that creates & handles PVectors. Furthermore the names of the static functions are easily guessed given the names of the objects' methods.

No big deal. Just good software engineering.
• Why do we have to write `add` instead of `_+=_` or `_ = _+__` on vectors? • The + operator can only add two single numeric values or used to place two strings together. The vector could have one, two or three values. The + sign cannot add several pairs of numbers and special methods are provided that will perform those tasks. The vector object has different methods that are far more complex than just basic addition and subtraction.

Methods
get() Gets the x, y, z component of the vector
set() Sets the x, y, z component of the vector
mag() Calculates the magnitude (length) of the vector and returns the result as a float
sub() Subtracts one vector from another
mult() Multiplies the vector by a scalar
div() Divides the vector by a scalar
dist() Calculate the Euclidean distance between two points
dot() Calculates the dot product
cross() Calculates the cross product
normalize() Normalizes the vector
limit() Limits the magnitude of the vector
angleBetween() Calculates the angle between two vectors
array() Return a representation of the vector as an array

Copied from:
http://processingjs.org/reference/PVector/ 