Skip to content

Instantly share code, notes, and snippets.

@masonicboom
Created February 19, 2011 05:49
Show Gist options
  • Select an option

  • Save masonicboom/834861 to your computer and use it in GitHub Desktop.

Select an option

Save masonicboom/834861 to your computer and use it in GitHub Desktop.
#!/usr/bin/env ruby
# Count co-occurences of all characters in words given 1-per-line from STDIN.
# count has the form { char_i => { char_j => # of co-occurences }, ... }.
# The block initialization here just sets up a default of 0 co-occurences.
count = Hash.new do
|hash, key| hash[key] = Hash.new { |h, k| h[k] = 0 }
end
STDIN.each do |word|
w = word.chomp
w.chars.to_a.combination(2).each do |char1, char2|
count[char1][char2] += 1
count[char2][char1] += 1
end
end
# Format count into a sparse matrix CSV for consumption by other programs.
index = Hash.new { |hash, key| hash[key] = hash.length+1 }
done = {}
count.each do |char1, other|
other.each do |char2, co_occurences|
ic1 = index[char1]
ic2 = index[char2]
# count includes duplicate data. Here, we strip them.
next if done[ [ic2, ic1] ]
done[ [ic1, ic2] ] = true
#puts "#{ic1},#{ic2},#{co_occurences}\n" # matlab likes this form, with numeric indices rather than unicode characters
puts "#{char1},#{char2},#{co_occurences}\n"
end
end
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment