te_rex 0.1.0 → 0.1.1

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA1:
-  metadata.gz: 4f48363d758ea116ca8660068ae10d5cee89ddcf
-  data.tar.gz: 186b23231afc32625bde651dcb32509e4125bd3a
+  metadata.gz: bae6e503a7f674b25ede59c7f9ef1c5de54b6a78
+  data.tar.gz: 71e7da095d199a9bc51419a3432912fb461b23c4
 SHA512:
-  metadata.gz: 90a71e60cbb0530737c305025ccedcc26e73ea18c478e8ab4223153743fea14250e3768af67fdf94abf87431ad7657f03b106e821004f871351615170309b7c9
-  data.tar.gz: 98057754a7f75100df6176854a948ff07f6e47712777b1fc35716a8bcab3a93962102c093ddb2efa31db5c1c22839edc20111c5a6be3cda74ae0ead4f700d722
+  metadata.gz: 84220db22732d965161f8e677662b4cb9392f02762d4cf89ef027cdf9622fdc10f8eb26fea72e91301151f6fe24cc71a09029b4d31c77a753b5f8005bfff1868
+  data.tar.gz: 3fb8d3e66ff969b967075d13a898b8a12e81c5456996e890cb77fe6b21ed076be59d4749ca3d3715d429a298cc48969b516ccfc42548fd51f700bd638dc9aada

data/lib/te_rex/bayes_data.rb

@@ -39,8 +39,8 @@ module TeRex
         # Each word in the string is interned and shows count in the document.
         def index_frequency(text)
           cfi = clean_stemmed_filtered_index(text)
-          #cni = clean_filtered_index(text)
-          cfi #.merge(cni)
+          cni = clean_filtered_index(text)
+          cfi.merge(cni)
         end
 
         # Return text with datetime and moneyterms replaced, remove cardinal terms (1st, 23rd, 42nd), remove punctuation.
@@ -88,7 +88,7 @@ module TeRex
           idx = Hash.new(0)
           word_array.each do |word|
             word.downcase!
-            if !TeRex::StopWord::LIST.include?(word) #&& word.length > 2
+            if !TeRex::StopWord::LIST.include?(word) && word.length > 3
               idx[word.intern] += 1
             end
           end

data/lib/te_rex/version.rb

@@ -1,3 +1,3 @@
 module TeRex
-  VERSION = "0.1.0"
+  VERSION = "0.1.1"
 end

data/test/bayes_data_test.rb

@@ -79,6 +79,6 @@ class BayesDataTest < PryTest::Test
 
     assert result[:moneyterm] == 3
     assert result[:sentenc] == 1
-    assert result[:sentence] == 0
+    assert result[:sentence] == 1
   end
 end

data/test/corpus_test.rb

@@ -58,5 +58,4 @@ class CorpusTest < PryTest::Test
     assert @@sent_corpus.training.count == 9
     assert @@sent_corpus.testing.count == 3
   end
-
 end

data/test/trained_bayes_provider_errors_test.rb

@@ -8,7 +8,7 @@ class TrainedBayesProviderErrorsTest < PryTest::Test
   @@credit_data = TeRex::Train::CREDITDATA
   @@credit_decline = TeRex::Train::CREDITDECLINE
   @@credit_service = TeRex::Train::CREDITSERVICE
-  @@unexpected = TeRex::Train::UNEXPECTED
+  #@@unexpected = TeRex::Train::UNEXPECTED
   #@@unk = TeRex::Train::UNKNOWNERROR
 
   @@cls = TeRex::Classifier::Bayes.new(
@@ -19,7 +19,7 @@ class TrainedBayesProviderErrorsTest < PryTest::Test
     {:tag => "CreditDataError",           :msg => "Credit Card data is invalid"},
     {:tag => "CreditDeclineError",        :msg => "Waht? Credit Card declined!"},
     {:tag => "CreditServiceError",        :msg => "External service problem processing"},
-    {:tag => "UnexpectedResponseError",   :msg => "Unexpected response"}
+    #{:tag => "UnexpectedResponseError",   :msg => "Unexpected response"}
     #{:tag => "UnknownError",              :msg => "Unexpected response"},
   )
   @@avail.each {|txt| @@cls.train("AvailabilityError", txt) }
@@ -29,7 +29,7 @@ class TrainedBayesProviderErrorsTest < PryTest::Test
   @@credit_data.each {|txt| @@cls.train("CreditDataError", txt) }
   @@credit_decline.each {|txt| @@cls.train("CreditDeclineError", txt) }
   @@credit_service.each {|txt| @@cls.train("CreditServiceError", txt) }
-  @@unexpected.each {|txt| @@cls.train("UnexpectedResponseError", txt) }
+  #@@unexpected.each {|txt| @@cls.train("UnexpectedResponseError", txt) }
   #@@unk.each {|txt| @@cls.train("UnknownError", txt) }
 
 
@@ -63,6 +63,10 @@ class TrainedBayesProviderErrorsTest < PryTest::Test
     #assert s_unexpected1 == ["UnexpectedResponseError", "Unexpected response"]
   end
 
+  test "total word counts correct" do
+    assert @@cls.total_words == 5696
+  end
+
 
 #  test "Training Data Set Test: cancel-cacancelical examples should classify correctly" do
 #

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: te_rex
 version: !ruby/object:Gem::Version
-  version: 0.1.0
+  version: 0.1.1
 platform: ruby
 authors:
 - Joshua Bowles
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2015-01-14 00:00:00.000000000 Z
+date: 2015-01-20 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: fast-stemmer