frekwenza 0.0.1 → 0.0.2

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA1:
-  metadata.gz: fe93d2259013d43219aac982560c795d7e698a03
-  data.tar.gz: 833373b3981ad562cb81ea3372d6f447c0a9d2bc
+  metadata.gz: 2e95e6c0045abe2259aa61401e009473fcfe39a0
+  data.tar.gz: d040b9ecbc8e842338839a576187a1384c060ef1
 SHA512:
-  metadata.gz: aaef936370715683e87b369fc7f578cc9160be2a694fdcf7cd61b2d8a0f6eed1232adc4b933de5d1c7d68c91674ac792b8a80fde8c45941500041d971e0f969e
-  data.tar.gz: e29aa5916853ac534800acd05e64261155b682f420b20e483e64ddc45bc4e6372571adaa19a8850fd768def67746bf4860f5fca44e038c75f62076ebe23c1c61
+  metadata.gz: 66a7349fa34dd1815397fbbe949fb81d927a88068d1b1e0557ddca728ba65ecc9660d494e4711decace68c827a47af4f728aaffb21dc51ed4234fd1a6e009f82
+  data.tar.gz: 4eb6cb0199039c171863d481f6148bb9cb8b60af39208ceb14a9176c0dc2e8b12997514f4f7a43213cdafa1541f0ac3720f81768cab22fff043a99e1916ba185

data/README.md

@@ -2,6 +2,12 @@
 
 This gem is based on [Mathieu Ripert](https://github.com/mathieuripert)'s work, [ruby-tf-idf](https://github.com/mathieuripert/ruby-tf-idf). Some changes are made in the gem, aside from the project structure.
 
+## Installation
+
+```
+$ gem install frekwenza
+```
+
 ## Usage
 
 Mostly similar with ruby-tf-idf, here's an example right from ruby-tf-idf's readme only with the "RubyTfIdf" part replaced with "Frekwenza".

data/lib/frekwenza/stop_words.rb

@@ -2,9 +2,13 @@ module Frekwenza
   class StopWords
     attr_reader :stop_words
 
-    def initialize(file)
-      string = read(file)
-      build_list(string)
+    def initialize(sw)
+      if sw.kind_of?(String)
+        string = read(sw)
+        build_list(string)
+      else
+        @stop_words = sw
+      end
     end
 
     private

data/lib/frekwenza/tf_idf.rb

@@ -2,14 +2,14 @@ module Frekwenza
   class TfIdf
     attr_reader :tf, :idf, :tf_idf
 
-    def initialize(docs, limit, stop_words_file=nil)
+    def initialize(docs, limit, stop_words=nil)
       @docs = split_docs(docs)
       @tf = []
       @idf = {}
       @tf_idf = []
       @docs_size = @docs.size
       calculate_tf_and_idf
-      calculate_tf_idf(limit, stop_words_file)
+      calculate_tf_idf(limit, stop_words)
     end
 
     private
@@ -38,14 +38,14 @@ module Frekwenza
       end
     end
 
-    def calculate_tf_idf(limit, stop_words_file)
+    def calculate_tf_idf(limit, stop_words)
       @tf.each do |tf_freq|
         tfidf = Hash.new(0)
         tf_freq.each do |k, v|
           tfidf[k] = @idf[k] * v
         end
-        if stop_words_file
-          sw = StopWords.new(stop_words_file)
+        if stop_words
+          sw = StopWords.new(stop_words)
           tfidf.reject!{|k| sw.stop_words.include?(k)}
         end
         tfidf = Hash[tfidf.sort_by{|k, v| -v}[0..limit-1]]

data/lib/frekwenza/version.rb

@@ -1,3 +1,3 @@
 module Frekwenza
-  VERSION = "0.0.1"
+  VERSION = "0.0.2"
 end

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification
 name: frekwenza
 version: !ruby/object:Gem::Version
-  version: 0.0.1
+  version: 0.0.2
 platform: ruby
 authors:
 - Mathieu Ripert
@@ -9,7 +9,7 @@ authors:
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2015-07-25 00:00:00.000000000 Z
+date: 2015-08-18 00:00:00.000000000 Z
 dependencies: []
 description: 'Term Frequency - Inverse Document Frequency '
 email: