Aleatory.js

aleatory |ˈeɪlɪət(ə)ri, ˈalɪət(ə)ri|

depending on the throw of a dice or on chance; random.

-- Oxford English Dictionary

Aleatory.js is a Javascript library to work with discrete random variables. The library exposes an expressive and functional API to create, manipulate, compose and query random variables.

Example

Aleatory.js is a library to handle random variables. A random variable is a collection of values, each associated with some probability. The sum of all probabilities is always equal to 1.

As an example, let's consider a dice. A dice can be thought of as a random variable that associates to the numbers 1 to 6 the probability 1/6. This random variable can be created as follows:

var d6 = Aleatory.dice(6);

Now that we have this random variable, we can query it.

// We verify that it is fair...
d6.probabilityAt(1).toFraction();  // 1/6, good.
d6.probabilityAt(6).toFraction();  // Still 1/6. Sounds about right!
d6.probabilityAt(7).toFraction();  // 0, nobody's that lucky!
d6.mean().toFraction();  // 7/2, as expected!

Random variables can also be derived from existing random variables. For instance, let's create a second dice, starting from our fair dice. But this time, we are going to cheat a bit! Whenever the fair dice would give a 1, we will change to a 6.

var cheaterD6 = d6.map(function (n) { 
  if (n === 1) {
    return 6;
  }
  return n;
});

Just as before, let's query this dice!

cheaterD6.probabilityAt(1).toFraction();  // 0, something's fishy!
cheaterD6.probabilityAt(2).toFraction();  // 1/6... sounds okay.
cheaterD6.probabilityAt(6).toFraction();  // Ha! 1/3. That's cheating!

It is worth noting that the original dice d6 remains unchanged and can still be used! All of the methods of Aleatory do not mutate the object.

Let's now define the random variable that corresponds to the sum of those two dice. To create this variable, we make use of the combine method.

var combined = d6.combine(function (n, m) { return n + m; }, cheaterD6);

Now, let's compute to probability to get at least 7 on those two dice.

combined.probability(function (n) { return n >= 7 }).toFraction()  // 13/18, good odds!

Random values can also be sampled from those objects. To do so, one must first create a Generator from the random variable.

var gen = combined.createGenerator();

The generator can then be sampled using its next method. The distribution of values sampled from the generator follows the random variable it was derived from.

var sum = 0;
var n = 1000000;

for (var i = 0; i < n; i++) {
  sum += gen.next();
}

meanNext = sum / n;

meanNext;  // The mean of generated values: 7.8341 (may change, it's random!)

combined.mean().valueOf();  // The mean of the random variable: 7.833333333333333

Aleatory variables have many more methods that can be used to transform them and query them in many interesting ways. Have a look at them in the documentation below!

Install

Using npm

To install this library with npm:

npm install aleatory.js

Using bower

To install this library with bower:

bower install aleatory.js

Use

In Node.js

var Aleatory = require('aleatory.js');

Manually on your website

To use this library on your website, simply download the library and its Fraction.js dependency. You can then include them in your HTML as follows:

<script type="text/javascript" src="fraction.js"></script>
<script type="text/javascript" src="aleatory.js"></script>

Using Require.js

Aleatory.js is compatible with Require.js.

<script type="text/javascript" src="require.js"></script>
<script type="text/javascript">
requirejs(['aleatory.js'], function(Aleatory) {
  // Your code using Aleatory.
});
</script>

Be careful to include Fraction.js before this library.

Documentation

Creating random variables

(static) always(value) → {Aleatory}

Random variable that always returns the same value.

Parameters:

Name	Type	Description
`value`	*	The only value in the random variable.

Returns:

The random variable that assigns to the value the probability 1.

Example:

var forecast = Aleatory.always("sunny");
forecast.probabilityAt("sunny").toFraction();  // 1
forecast.probabilityAt("cloudy").toFraction();  // 0

(static) dice(n) → {Aleatory}

Uniform Aleatory variable over numbers between 1 and n inclusive.

Parameters:

Name	Type	Description
`n`	`number`	The maximum value of the Aleatory variable.

Returns:

The uniform random variable of numbers between 1 and n inclusive.

Example:

var d3 = Aleatory.dice(3);
d3.mean().toFraction();  // 2

(static) uniform(elements) → {Aleatory}

Uniform distribution of the elements. In case of duplicate elements, the probability of each element is proportional to its number of occurences.

Parameters:

Name	Type	Description
`elements`	`Array`	An array of values.

Returns:

The uniform random variable over the elements.

Example:

var colors = Aleatory.uniform(["red", "green", "blue", "yellow"]);
colors.probabilityAt("red").toFraction();  // 1/4

(static) weighted(elements) → {Aleatory}

Weighted distribution of the elements. Each element of elements should be an object with a value field and a weight field. In case of duplicate elements, the probability of each element is proportional to the sum of the weights of its occurences.

Parameters:

Name	Type	Description
`elements`	Array	An array of `{value, weight}` objects.

Returns:

The weighted random variable over the elements.

Example:

var names = Aleatory.weighted([
  { value: "Alice", weight: 6 },
  { value: "Bob", weight: 2 },
  { value: "Charles", weight: 3 },
  { value: "Douglas", weight: 1 }
]);
names.probabilityAt("Bob").toFraction();  // 1/6

Modifying random variables

map(toResult) → {Aleatory}

Applies a function to the values of this random variable. The resulting random variable assigns to each result the probability of this result to be obtained.

Parameters:

Name	Type	Description
`toResult`	`function`	The function to apply on each value.

Returns:

The aleatory variable of the results.

Example:

var result = Aleatory.dice(6).map(function (n) {
  if (n <= 1) {
    return "Critical failure";
  }
  if (n <= 3) {
    return "Failure";
  }
  if (n <= 5) {
    return "Success";
  }
  return "Critical success";
})
result.probabilityAt("Critical failure").toFraction();  // 1/6
result.probabilityAt("Success").toFraction();  // 1/3

flatMap(toAleatory) → {Aleatory}

Applies a function that returns an aleatory variable to each possible value of this aleatory variable. The resulting aleatory variable assigns to each result the probability of this result to be obtained.

Parameters:

Name	Type	Description
`toAleatory`	`function`	The function to apply on each value.

Returns:

The random variable of the results.

Example:

var result = Aleatory.dice(6).flatMap(function (n) {
  if (n <= 3) {
    return Aleatory.always(0);
  }
  return Aleatory.dice(3);
})
result.probabilityAt(0).toFraction();  // 1/2
result.probabilityAt(1).toFraction();  // 1/6

assume(predicate) → {Aleatory}

Returns this random variable conditioned by the predicate. All the values in the resulting random variable satisfy the given predicate. Their probability is proportional to the probability the had in this random variable. As always, the sum of all probabilities in the resulting variable is 1. If none of the values satisfy the predicate, this function returns undefined.

Parameters:

Name	Type	Description
`predicate`	`function`	The predicate that values must hold.

Returns:

The random variable that contains all values which satisfy the predicate.

Example:

var result = Aleatory.dice(6).assume(function (n) {
  return n % 2 == 0;
})
result.probabilityAt(2).toFraction();  // 1/3
result.probabilityAt(5).toFraction();  // 0

combine(combiner, that) → {Aleatory}

Combines two random variables using a combiner function.

Parameters:

Name	Type	Description
`combiner`	`function`	The binary function to combine values of this and that.
`that`	`Aleatory`	The other Aleatory variable.

Returns:

The random variable of the combinations.

Example:

var result = Aleatory.dice(6).combine(function (a, b) { 
  return a - b;
}, Aleatory.dice(3))
result.probabilityAt(2).toFraction();  // 1/6
result.probabilityAt(5).toFraction();  // 1/18

Repeating random experiments

times(n, combiner=addition) → {Aleatory}

Take n independant values from this random variable and combine them using combiner. Returns the random variable of the combinations.

Parameters:

Name	Type	Attributes	Description
`n`	number		Number of times values are taken from `this`.
`combiner`	function	optional	Binary function used to combine values. Should be associative. Defaults to addition.

Returns:

The random variable of combinations.

Example:

var tenD3 = Aleatory.dice(3).times(10);
tenD3.mean().toFraction();  // 20
tenD3.probability(function (n) { return n >= 20; }).toFraction();  // 34001/59049

trials(n) → {Aleatory}

Bernouli trials distribution. Each number i between 0 and n inclusive is associated with the probability of having exactly i successful outcomes in n trials of given Aleatory variable. A value is considered to be a successful outcome if it is "truthy". All "falsy" values (i.e. false, undefined, NaN, null, 0, "") are considered to be failures.

Parameters:

Name	Type	Description
`n`	number	The number of trials.

Returns:

The random variable of the numbers of successful outcomes.

Example:

var trials = Aleatory.dice(6).map(function (n) {
  return n === 1 || n === 6;
}).trials(10);
trials.mean().toFraction();  // 10/3

Measuring random variables

domain() → {Array}

Returns all values with non-zero probability.

Returns:

Values with non-zero probability.

Example:

var twoD6 = Aleatory.dice(6).times(2);
twoD6.domain();  // [ 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12 ]

probabilityAt(value) → {Fraction}

Returns the probability of a certain value.

Parameters:

Name	Type	Description
`value`	*	The value to test.

Returns:

The probability of the value.

Example:

var d6 = Aleatory.dice(6);
d6.probabilityAt(1).toFraction();  // 1/6
d6.probabilityAt(6).toFraction();  // 1/6
d6.probabilityAt(0).toFraction();  // 0
d6.probabilityAt("6").toFraction();  // 0

probability(predicate) → {Fraction}

Returns the probability of a certain predicate being true.

Parameters:

Name	Type	Description
`predicate`	function	The predicate to test.

Returns:

The probability of having a value that satifies the predicate.

Example:

var d6 = Aleatory.dice(6);
d6.probability(function (n) { return n >= 3; }).toFraction();  // 2/3

mean() → {Fraction}

Computes the mean, or expectation of this random variable.

Returns:

The mean of this random variable.

Example:

var d6 = Aleatory.dice(6);
d6.mean().toFraction();  // 7/2

variance() → {Fraction}

Computes the variance of this random variable.

Returns:

The variance of this random variable.

Example:

var d6 = Aleatory.dice(6);
d6.variance().toFraction();  // 35/12

Sampling random values

createGenerator() → {Generator}

Creates a random value generator. The random generator is used to get random values from this Aleatory variable.

Takes time linear in the number of values.

The generator has a single method, called next, which returns a random value from the generator. Each value is returned with probability this.probabilityAt(value). Calling next takes only constant time.

Returns:

A random value generator which follows the distribution of values described by this random variable.

Example:

var gen = Aleatory.dice(6).times(2).createGenerator();
gen.next();  // 7
gen.next();  // 8
gen.next();  // 6
gen.next();  // 10
gen.next();  // 7

JSDoc

The library is fully documented using JSDoc syntax. You can generate the documentation with:

jsdoc aleatory.js

License

This library is released under the Apache-2.0 license. See the LICENSE file for details.

Files

README.md

Latest commit

History

README.md

File metadata and controls

Aleatory.js

Example

Install

Using npm

Using bower

Use

In Node.js

Manually on your website

Using Require.js

Documentation

Creating random variables

(static) always(value) → {Aleatory}

Parameters:

Returns:

Example:

(static) dice(n) → {Aleatory}

Parameters:

Returns:

Example:

(static) uniform(elements) → {Aleatory}

Parameters:

Returns:

Example:

(static) weighted(elements) → {Aleatory}

Parameters:

Returns:

Example:

Modifying random variables

map(toResult) → {Aleatory}

Parameters:

Returns:

Example:

flatMap(toAleatory) → {Aleatory}

Parameters:

Returns:

Example:

assume(predicate) → {Aleatory}

Parameters:

Returns:

Example:

combine(combiner, that) → {Aleatory}

Parameters:

Returns:

Example:

Repeating random experiments

times(n, combiner=addition) → {Aleatory}

Parameters:

Returns:

Example:

trials(n) → {Aleatory}

Parameters:

Returns:

Example:

Measuring random variables

domain() → {Array}

Returns:

Example:

probabilityAt(value) → {Fraction}

Parameters:

Returns:

Example:

probability(predicate) → {Fraction}

Parameters:

Returns:

Example:

mean() → {Fraction}

Returns:

Example:

variance() → {Fraction}

Returns:

Example:

Sampling random values

createGenerator() → {Generator}

Returns: