Skip to content

Commit

Permalink
some air
Browse files Browse the repository at this point in the history
  • Loading branch information
josh committed Jun 10, 2013
1 parent 9822b15 commit 490afdd
Showing 1 changed file with 10 additions and 10 deletions.
20 changes: 10 additions & 10 deletions lib/linguist/classifier.rb
Original file line number Diff line number Diff line change
Expand Up @@ -132,7 +132,7 @@ def language_probability(language)

private
def verbosity
@verbosity ||= (ENV['LINGUIST_DEBUG']||0).to_i
@verbosity ||= (ENV['LINGUIST_DEBUG'] || 0).to_i
end

# Internal: show a table of probabilities for each <token,language> pair.
Expand All @@ -146,19 +146,19 @@ def verbosity
# appear in one language vs. the least-likely language. Dashes
# indicate the least-likely language (and zero points) for each token.
def dump_all_tokens(tokens, languages)
maxlen = tokens.map{|tok| tok.size}.max
maxlen = tokens.map { |tok| tok.size }.max
printf "%#{maxlen}s", ""
puts " #" + languages.map{|lang| sprintf("%10s", lang)}.join
puts " #" + languages.map { |lang| sprintf("%10s", lang) }.join
tokmap = Hash.new(0)
tokens.each{|tok| tokmap[tok] += 1}
tokmap.sort.each{|tok, count|
arr = languages.map{|lang| [lang, token_probability(tok, lang)] }
min = arr.map{|a,b| b}.min
tokens.each { |tok| tokmap[tok] += 1 }
tokmap.sort.each { |tok, count|
arr = languages.map { |lang| [lang, token_probability(tok, lang)] }
min = arr.map { |a,b| b }.min
minlog = Math.log(min)
if !arr.inject(true) {|result, n| result && n[1] == arr[0][1]} # if not all the same
if !arr.inject(true) { |result, n| result && n[1] == arr[0][1] }
printf "%#{maxlen}s%5d", tok, count
puts arr.map{|ent|
ent[1] == min ? " -" : sprintf("%10.3f", count*(Math.log(ent[1])-minlog))
puts arr.map { |ent|
ent[1] == min ? " -" : sprintf("%10.3f", count * (Math.log(ent[1]) - minlog))
}.join
end
}
Expand Down

0 comments on commit 490afdd

Please sign in to comment.