From 08bfbc8ee6a14133f202cdb4e4ae837bbe9087ef Mon Sep 17 00:00:00 2001
From: Matteo Wiley <matteowiley@gmail.com>
Date: Wed, 3 Sep 2014 21:08:48 -0500
Subject: [PATCH 1/3] This organizes the data into unique words and author

---
 gutenberg.rb             |  2 +-
 lib/complex_predictor.rb | 28 ++++++++++++++++++++++++++++
 2 files changed, 29 insertions(+), 1 deletion(-)

diff --git a/gutenberg.rb b/gutenberg.rb
index 84d20f6..e955c85 100644
--- a/gutenberg.rb
+++ b/gutenberg.rb
@@ -23,6 +23,6 @@ def run!(predictor_klass, opts={})
   puts "Accuracy: #{accuracy}"
 end
 
-run!(SimplePredictor)
+#run!(SimplePredictor)
 run!(ComplexPredictor, debug: true)
 
diff --git a/lib/complex_predictor.rb b/lib/complex_predictor.rb
index b8921f3..431c63f 100644
--- a/lib/complex_predictor.rb
+++ b/lib/complex_predictor.rb
@@ -1,4 +1,5 @@
 require_relative 'predictor'
+require 'pry-byebug'
 
 class ComplexPredictor < Predictor
   # Public: Trains the predictor on books in our dataset. This method is called
@@ -6,7 +7,34 @@ class ComplexPredictor < Predictor
   #
   # Returns nothing.
   def train!
+    
     @data = {}
+
+    @all_books.each do |category, books|
+      @data[category] = {
+        uniq_words: Hash.new(0),
+        authors: []
+      }
+
+      books.each do |filename, token|
+        token.each do |word|
+          if good_token?(word) && word.length > 5
+            @data[category][:uniq_words][word] += 1
+          end
+        @author_start = token.find_index("by")
+        @data[category][:authors] << token[(@author_start + 1)..(@author_start+2)]
+        end
+      end
+      @data[category][:uniq_words] = @data[category][:uniq_words].sort_by {|key, value| value}.reverse[0..14]
+      binding.pry
+    end
+
+     #@date = {
+     #unq_word: []
+     #book:
+     # 
+     #
+     #}
   end
 
   # Public: Predicts category.

From 0c8225ef0bdf199c241691d6c448ec2c309c20d3 Mon Sep 17 00:00:00 2001
From: Matteo Wiley <matteowiley@gmail.com>
Date: Thu, 4 Sep 2014 09:56:33 -0500
Subject: [PATCH 2/3] This runs with an 96% acc but at 83 seconds

---
 lib/complex_predictor.rb | 23 ++++++++++++++++++-----
 1 file changed, 18 insertions(+), 5 deletions(-)

diff --git a/lib/complex_predictor.rb b/lib/complex_predictor.rb
index 431c63f..00ff986 100644
--- a/lib/complex_predictor.rb
+++ b/lib/complex_predictor.rb
@@ -21,13 +21,16 @@ def train!
           if good_token?(word) && word.length > 5
             @data[category][:uniq_words][word] += 1
           end
+        end
         @author_start = token.find_index("by")
         @data[category][:authors] << token[(@author_start + 1)..(@author_start+2)]
         end
-      end
-      @data[category][:uniq_words] = @data[category][:uniq_words].sort_by {|key, value| value}.reverse[0..14]
-      binding.pry
+        @data[category][:uniq_words] = @data[category][:uniq_words].sort_by {|key, value| value}.reverse[0..14]
+        
+        # binding.pry
+     
     end
+    
 
      #@date = {
      #unq_word: []
@@ -43,8 +46,18 @@ def train!
   #
   # Returns a category.
   def predict(tokens)
-    # Always predict astronomy, for now.
-    :astronomy
+    @match = Hash.new(0)
+    
+    @data.each do |category, uniq|
+      tokens.each do |word|
+        uniq[:uniq_words].each do |uni_w|
+          if uni_w.include?(word)
+            @match[category] += 1
+          end
+        end
+      end
+    end
+    @match.max_by {|category, count| count }.first
   end
 end
 

From 98dfbcc49deee6eca8deb1359a9d3bbe17ebd017 Mon Sep 17 00:00:00 2001
From: Matteo Wiley <matteowiley@gmail.com>
Date: Thu, 4 Sep 2014 10:43:46 -0500
Subject: [PATCH 3/3] This takes it up too 100% and down to 5 secs

---
 lib/complex_predictor.rb | 24 +++++++++++-------------
 1 file changed, 11 insertions(+), 13 deletions(-)

diff --git a/lib/complex_predictor.rb b/lib/complex_predictor.rb
index 00ff986..2b55a08 100644
--- a/lib/complex_predictor.rb
+++ b/lib/complex_predictor.rb
@@ -18,17 +18,21 @@ def train!
 
       books.each do |filename, token|
         token.each do |word|
-          if good_token?(word) && word.length > 5
+          if good_token?(word) && word.length > 7
             @data[category][:uniq_words][word] += 1
           end
         end
         @author_start = token.find_index("by")
         @data[category][:authors] << token[(@author_start + 1)..(@author_start+2)]
         end
-        @data[category][:uniq_words] = @data[category][:uniq_words].sort_by {|key, value| value}.reverse[0..14]
+        @new_hash = Hash.new(0)
+        @most_pop = @data[category][:uniq_words].sort_by {|key, value| value}.reverse[0..100]
+        @most_pop.map! {|word, count| word}
+        # @most_pop.each do |x|
+        #   @new_hash[x.first] = x.last
+        # end
+        @data[category][:uniq_words] = @most_pop
         
-        # binding.pry
-     
     end
     
 
@@ -47,17 +51,11 @@ def train!
   # Returns a category.
   def predict(tokens)
     @match = Hash.new(0)
-    
     @data.each do |category, uniq|
-      tokens.each do |word|
-        uniq[:uniq_words].each do |uni_w|
-          if uni_w.include?(word)
-            @match[category] += 1
-          end
-        end
-      end
+      @match[category] = (@data[category][:uniq_words] & tokens).length
     end
     @match.max_by {|category, count| count }.first
-  end
+   end
+
 end