hash_sample

Implements methods for Hash class for getting weighted random samples with and without replacement, as well as regular random samples

gem install hash_sample

    require 'hash_sample'
    loaded_die = {'1' => 0.1, '2' => 0.1, '3' => 0.1, '4' => 0.1, '5' => 0.1, '6' => 0.5}
    # weighted random choice of keys, with replacement (elements can repeat)  
    p loaded_die.weighted_choice      # "6"
    p loaded_die.weighted_choices(1)   # ["6"]
    p loaded_die.wchoices(10)  # ["4", "6", "3", "3", "2", "2", "1", "6", "4", "6"]
    # weighted random choice of keys, without replacement (elements can NOT repeat)
    p loaded_die.weighted_sample      # 6
    p loaded_die.weighted_samples(6)   # ["6", "3", "2", "4", "1", "5"]
    p loaded_die.wsamples(10)  # ["2", "6", "1", "3", "4", "5"]
    # regular random choice of key-value pairs (pairs can NOT repeat)
    p loaded_die.sample       # { '1' => 0.1 }
    p loaded_die.samples(6)    # {'1' => 0.1, '2' => 0.1, '3' => 0.1, '4' => 0.1, '5' => 0.1, '6' => 0.5}

Choose a random key=>value pair or n random pairs from the hash.

The key=>value pairs are chosen by using random and unique indices in order to ensure that each pair doesn't includes more than once

If the hash is empty it returns an empty hash.

If the hash contains less than n unique keys, the copy of whole hash will be returned, none of keys will be lost due to bad luck.

Returns new Hash containing sample key=>value pairs

Weighted random sampling with replacement.

Choose a random key or n random keys from the hash, according to weights defined in hash values.

The samples are drawn by using random and replaced by its copy, so they can be repeated in result.

If the hash is empty the first form returns nil and the second form returns an empty array.

All weights should be Numeric.

Zero or negative weighs will be ignored.

{'_' => 9, 'a' => 1}.wchoices(10)  # ["_", "a", "_", "_", "_", "_", "_", "_", "_", "_"]

Weighted random sampling without replacement.

Choose 1 or n distinct random keys from the hash, according to weights defined in hash values. Drawn items are not put back into the set, so they can not be repeated in result.

If the hash is empty, singular form returns nil and plural returns an empty array.

All weights should be Numeric.

Zero or negative weighs will be ignored.

Hash.new.weighted_sample                   # nil
Hash.new.weighted_samples                  # []
{'_' => 9, 'a' => 1}.weighted_samples(10)  # ["_", "a"]

alias for wchoice

alias for wsample

1. Fork it ( https://github.com/serg123e/hash_sample/fork )
2. Create your feature branch (git checkout -b my-new-feature)
3. Commit your changes (git commit -am 'Add some feature')
4. Push to the branch (git push origin my-new-feature)
5. Create a new Pull Request

1. Efraimidis and Spirakis implementation of random sampling with replacement
2. Weighted Random Sampling (2005; Efraimidis, Spirakis)
3. Abandoned Ruby feature request
4. Inspiring example of using max_by for Enumerables with the same algorithm