Weighted Distribution is used to sample from a distribution of objects with different weights. Sampling from this distribution will randomly select an object with a probability proportional to its weight. I followed the specification used by Ryan Lecompte's Weighted Randomizer. I have improved on the runtime efficiency of the sampling algorithm to run in O(log(n)) time rather then O(n). This will be useful for higher performance distribution sampling from distributions with many objects.
Add this line to your application's Gemfile:
gem 'weighted_distribution'
And then execute:
$ bundle
Or install it yourself as:
$ gem install weighted_distribution
To sample from a distribution:
suits = {'heart' => 12, 'spade' => 11, 'diamond' => 10, 'club' => 13}
suit_distribution = WeightedDistribution.new(suits)
# Fetch a single random object.
suit_distribution.sample # => single object
# Fetch the next 10 weighted random objects.
suit_distribution.sample(10) # => array of objects
Licensed with the MIT license (see LICENSE)
The design of the gem and its specification was based largely on Ryan Lecompte's Weighted Randomizer which was based on recipe 5.11 from the Ruby Cookbook