Above we are extracting the fractional content of a sine wave. The [```sin()```](../glossary/?search=sin) values that fluctuate between ```-1.0``` and ```1.0``` have been chopped behind the floating point, returning all positive values between ```0.0``` and ```1.0```. We can use this effect to get some pseudo-random values by "breaking" this sine wave into smaller pieces. How? By multiplying the resultant of [```sin(x)```](../glossary/?search=sin) by larger numbers. Go ahead and click on the function above and start adding some zeros.
By the time you get to ```100000.0``` ( and the equation looks like this: ```y = fract(sin(x)*100000.0)``` ) you aren't able to distinguish the sine wave any more. The granularity of the fractional part has corrupted the flow of the sine wave into pseudo-random chaos.
Using random can be hard; it is both too chaotic and sometimes not random enough. Take a look at the following graph. To make it, we are using a ```rand()``` function which is implemented exactly like we describe above.
Taking a closer look, you can see the [```sin()```](../glossary/?search=sin) wave crest at ```-1.5707``` and . I bet you now understand why - it's where the maximum and minimum of the sine wave happens.
よく見ると [```sin()```](../glossary/?search=sin) の描く波が ```-1.5707``` と ```-1.5707``` で頂点を迎えています。お分かりですね。これは波が最大値と最小値になる場所です。*talking about "crests" doesn't make much sense here. Better mention the tiny gaps around +PI/2 and -PI/2 and explain why this happens, i.e. these are where the crest of the sine curve are?*
A while ago [Pixelero](https://pixelero.wordpress.com) published an [interesting article about random distribution](https://pixelero.wordpress.com/2008/04/24/various-functions-and-various-distributions-with-mathrandom/). I've added some of the functions he uses in the previous graph for you to play with and see how the distribution can be changed. Uncomment the functions and see what happens.
If you read [Pixelero's article](https://pixelero.wordpress.com/2008/04/24/various-functions-and-various-distributions-with-mathrandom/), it is important to keep in mind that our ```rand()``` function is a deterministic random, also known as pseudo-random. Which means for example ```rand(1.)``` is always going to return the same value. [Pixelero](https://pixelero.wordpress.com/2008/04/24/various-functions-and-various-distributions-with-mathrandom/) makes reference to the ActionScript function ```Math.random()``` which is non-deterministic; every call will return a different value.
Now that we have a better understanding of randomness, it's time to apply it in two dimensions, to both the ```x``` and ```y``` axis. For that we need a way to transform a two dimensional vector into a one dimensional floating point value. There are different ways to do this, but the [```dot()```](../glossary/?search=dot) function is particulary helpful in this case. It returns a single float value between ```0.0``` and ```1.0``` depending on the alignment of two vectors.
Our first step is to apply a grid to it; using the [```floor()```](../glossary/?search=floor) function we will generate an integer table of cells. Take a look at the following code, especially lines 22 and 23.
After scaling the space by 10 (on line 21), we separate the integers of the coordinates from the fractional part. We are familiar with this last operation because we have been using it to subdivide a space into smaller cells that go from ```0.0``` to ```1.0```. By obtaining the integer of the coordinate we isolate a common value for a region of pixels, which will look like a single cell. Then we can use that common integer to obtain a random value for that area. Because our random function is deterministic, the random value returned will be constant for all the pixels in that cell.
Uncomment line 29 to see that we preserve the floating part of the coordinate, so we can still use that as a coordinate system to draw things inside each cell.
Here I'm using the random values of the cells to draw a line in one direction or the other using the ```truchetPattern()``` function from the previous chapter (lines 41 to 47).
[Ryoji Ikeda](http://www.ryojiikeda.com/), Japanese electronic composer and visual artist, has mastered the use of random; it is hard not to be touched and mesmerized by his work. His use of randomness in audio and visual mediums is forged in such a way that it is not annoying chaos but a mirror of the complexity of our technological culture.
* Make rows of moving cells (in opposite directions) with random values. Only display the cells with brighter values. Make the velocity of the rows fluctuate over time.
* Similarly make several rows but each one with a different speed and direction. Hook the position of the mouse to the threshold of which cells to show.
Using random aesthetically can be problematic, especially if you want to make natural-looking simulations. Random is simply too chaotic and very few things look ```random()``` in real life. If you look at a rain pattern or a stock chart, which are both quite random, they are nothing like the random pattern we made at the begining of this chapter. The reason? Well, random values have no correlation between them what so ever, but most natural patterns have some memory of the previous state.
