segment_ruby 0.1.1 → 0.1.2
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- checksums.yaml +4 -4
- data/.gitignore +2 -0
- data/data/segment_ruby/test_bigram/2_frequencies.tsv +10 -0
- data/data/segment_ruby/test_bigram/2_total.tsv +1 -0
- data/data/segment_ruby/test_bigram/frequencies.tsv +10 -0
- data/data/segment_ruby/test_bigram/total.tsv +1 -0
- data/data/segment_ruby/test_unigram/frequencies.tsv +10 -0
- data/data/segment_ruby/test_unigram/total.tsv +1 -0
- data/data/segment_ruby/us_names/2_frequencies.tsv.save +0 -0
- data/data/segment_ruby/us_names/2_total.tsv.save +1 -0
- data/data/segment_ruby/us_names/README.md +15 -0
- data/data/segment_ruby/us_names/frequencies.tsv +78637 -0
- data/data/segment_ruby/us_names/total.tsv +1 -0
- data/lib/segment_ruby/version.rb +1 -1
- data/lib/segment_ruby.rb +3 -3
- metadata +13 -2
|
@@ -0,0 +1 @@
|
|
|
1
|
+
175469698
|
data/lib/segment_ruby/version.rb
CHANGED
data/lib/segment_ruby.rb
CHANGED
|
@@ -68,7 +68,7 @@ module SegmentRuby
|
|
|
68
68
|
# bigram log probabilities
|
|
69
69
|
btf = total_file_name('2_')
|
|
70
70
|
bff = freq_file_name('2_')
|
|
71
|
-
@blp = (File.exists?(btf)
|
|
71
|
+
@blp = (File.exists?(btf) && File.exists?(bff) ? ProbabilityDistribution.new(btf, bff) : false)
|
|
72
72
|
end
|
|
73
73
|
|
|
74
74
|
attr_reader :blp, :max_word_length, :model_name, :ulp
|
|
@@ -80,7 +80,7 @@ module SegmentRuby
|
|
|
80
80
|
def log_CPr(w, prev)
|
|
81
81
|
key = [prev, w].join(' ')
|
|
82
82
|
|
|
83
|
-
blp
|
|
83
|
+
blp && blp.has_key?(key) ? blp.log_prob(key) : ulp.log_prob(w)
|
|
84
84
|
end
|
|
85
85
|
|
|
86
86
|
def total_file_name(prefix='')
|
|
@@ -107,7 +107,7 @@ module SegmentRuby
|
|
|
107
107
|
end
|
|
108
108
|
|
|
109
109
|
def segment_r(text, prev, n, memo)
|
|
110
|
-
return [0.0, []] if
|
|
110
|
+
return [0.0, []] if (!text) || (text.size == 0)
|
|
111
111
|
return memo[text] if memo.has_key?(text)
|
|
112
112
|
|
|
113
113
|
log_p_segment = splits(text).map do |first, rem|
|
metadata
CHANGED
|
@@ -1,14 +1,14 @@
|
|
|
1
1
|
--- !ruby/object:Gem::Specification
|
|
2
2
|
name: segment_ruby
|
|
3
3
|
version: !ruby/object:Gem::Version
|
|
4
|
-
version: 0.1.
|
|
4
|
+
version: 0.1.2
|
|
5
5
|
platform: ruby
|
|
6
6
|
authors:
|
|
7
7
|
- Will Fitzgerald
|
|
8
8
|
autorequire:
|
|
9
9
|
bindir: exe
|
|
10
10
|
cert_chain: []
|
|
11
|
-
date: 2016-10-
|
|
11
|
+
date: 2016-10-18 00:00:00.000000000 Z
|
|
12
12
|
dependencies:
|
|
13
13
|
- !ruby/object:Gem::Dependency
|
|
14
14
|
name: bundler
|
|
@@ -100,11 +100,22 @@ files:
|
|
|
100
100
|
- data/segment_ruby/small/README.md
|
|
101
101
|
- data/segment_ruby/small/frequencies.tsv
|
|
102
102
|
- data/segment_ruby/small/total.tsv
|
|
103
|
+
- data/segment_ruby/test_bigram/2_frequencies.tsv
|
|
104
|
+
- data/segment_ruby/test_bigram/2_total.tsv
|
|
105
|
+
- data/segment_ruby/test_bigram/frequencies.tsv
|
|
106
|
+
- data/segment_ruby/test_bigram/total.tsv
|
|
107
|
+
- data/segment_ruby/test_unigram/frequencies.tsv
|
|
108
|
+
- data/segment_ruby/test_unigram/total.tsv
|
|
103
109
|
- data/segment_ruby/twitter/2_frequencies.tsv
|
|
104
110
|
- data/segment_ruby/twitter/2_total.tsv
|
|
105
111
|
- data/segment_ruby/twitter/README.md
|
|
106
112
|
- data/segment_ruby/twitter/frequencies.tsv
|
|
107
113
|
- data/segment_ruby/twitter/total.tsv
|
|
114
|
+
- data/segment_ruby/us_names/2_frequencies.tsv.save
|
|
115
|
+
- data/segment_ruby/us_names/2_total.tsv.save
|
|
116
|
+
- data/segment_ruby/us_names/README.md
|
|
117
|
+
- data/segment_ruby/us_names/frequencies.tsv
|
|
118
|
+
- data/segment_ruby/us_names/total.tsv
|
|
108
119
|
- lib/segment_ruby.rb
|
|
109
120
|
- lib/segment_ruby/version.rb
|
|
110
121
|
- segment_ruby.gemspec
|