recommendify 0.1.0 → 0.2.0

data/README.md

@@ -86,7 +86,7 @@ recommender.for("item23")
 
 # remove "item23" from the similarity matrix and the input matrices. you should 
 # do this if your items 'expire', since it will speed up the calculation
-recommender.remove_item!("item23") 
+recommender.delete_item!("item23") 
 ```
 
 ### how it works
@@ -114,7 +114,7 @@ These recommendations were calculated from 2,3mb "profile visit"-data (taken fro
 
 full snippet: http://falbala.23loc.com/~paul/recommendify_out_1.html 
 
-Initially processing the 120.047 `visitor_id->profile_id` pairs currently takes around half an hour on a single core and creates a 126.64mb hashtable in redis. The high memory usage of >100mb for only 5000 items is due to the very long user rows. If you limit the user rows to 100 items (mahout's default) it shrinks to 31mb for the 5k items from example_data.csv. In another real data set with very short user rows (purchase/payment data) it used only 3.4mb for 90k items with very good results. You can try this for yourself; the complete data and code is in `doc/example.rb` and `doc/example_data.csv`. 
+Initially processing the 120.047 `visitor_id->profile_id` pairs currently takes around half an hour with the ruby-only implementation and ~130 seconds with the native/c implementation on a single core. It creates a 24.1mb hashtable in redis (with truncated user_rows a' max 100 items). In another real data set with very short user rows (purchase/payment data) it used only 3.4mb for 90k items with very good results. You can try this for yourself; the complete data and code is in `doc/example.rb` and `doc/example_data.csv`. 
 
 
 

data/doc/example.rb

@@ -11,7 +11,9 @@ class UserRecommender < Recommendify::Base
   #max_neighbors 50
 
   input_matrix :visits, 
-    :similarity_func => :jaccard
+    :similarity_func => :jaccard,
+    :native => true
+
 end
 
 recommender = UserRecommender.new

data/lib/recommendify/base.rb

@@ -77,4 +77,10 @@ class Recommendify::Base
     similarity_matrix.commit_item!(item_id)
   end
 
+  def delete_item!(item_id)
+    input_matrices.map do |k,m|
+      m.delete_item(item_id)
+    end
+  end
+
 end

data/lib/recommendify/cc_matrix.rb

@@ -29,6 +29,11 @@ module Recommendify::CCMatrix
     Recommendify.redis.hkeys(redis_key(:items))
   end
 
+  def delete_item(item_id)    
+    Recommendify.redis.hdel(redis_key(:items), item_id)
+    @ccmatrix.send(:k_delall, item_id)
+  end
+
 private
 
   def all_pairs(keys)

data/lib/recommendify/input_matrix.rb

@@ -44,4 +44,9 @@ class Recommendify::InputMatrix
     raise "implemented in subclass"
   end
 
+  # delete item_id from the matrix
+  def delete_item(item_id)
+    raise "implemented in subclass"
+  end
+
 end

data/lib/recommendify/sparse_matrix.rb

@@ -42,4 +42,12 @@ private
     Recommendify.redis.hincrby(redis_key, key, 1)
   end
 
+  # OPTIMIZE: use scripting/lua in redis 2.6
+  def k_delall(key)
+    Recommendify.redis.hkeys(redis_key).each do |iikey|
+      next unless iikey.split(":").include?(key)
+      Recommendify.redis.hdel(redis_key, iikey)
+    end
+  end
+
 end

data/recommendify.gemspec

@@ -3,7 +3,7 @@ $:.push File.expand_path("../lib", __FILE__)
 
 Gem::Specification.new do |s|
   s.name        = "recommendify"
-  s.version     = "0.1.0"
+  s.version     = "0.2.0"
   s.date        = Date.today.to_s
   s.platform    = Gem::Platform::RUBY
   s.authors     = ["Paul Asmuth"]

data/spec/base_spec.rb

@@ -158,4 +158,21 @@ describe Recommendify::Base do
 
   end
 
+  describe "delete_item!" do
+
+    it "should call delete_item on each input_matrix" do
+      Recommendify::Base.input_matrix(:myfirstinput, :similarity_func => :jaccard)
+      Recommendify::Base.input_matrix(:mysecondinput, :similarity_func => :jaccard)
+      sm = Recommendify::Base.new
+      sm.myfirstinput.should_receive(:delete_item).with("fnorditem")
+      sm.mysecondinput.should_receive(:delete_item).with("fnorditem")
+      sm.delete_item!("fnorditem")
+    end
+
+    it "should delete the item from the similarity matrix"
+
+    it "should delete all occurences of the item in other similarity sets from the similarity matrix"
+
+  end
+
 end

data/spec/cc_matrix_shared.rb

@@ -51,4 +51,25 @@ share_examples_for Recommendify::CCMatrix do
     @matrix.all_items.should include("wurst")
   end
 
+  it "should delete all item<->item pairs on item deletion" do
+    @matrix.ccmatrix["foo", "fnord"] = 2
+    @matrix.add_set("user123", ["foo", "bar", "fnord"])
+    @matrix.add_set("user456", ["fnord", "blubb"])
+    @matrix.ccmatrix["bar", "foo"].should == 1
+    @matrix.ccmatrix["foo", "fnord"].should == 3
+    @matrix.ccmatrix["blubb", "fnord"].should == 1
+    @matrix.delete_item("fnord")
+    @matrix.ccmatrix["bar", "foo"].should == 1
+    @matrix.ccmatrix["foo", "fnord"].should == 0
+    @matrix.ccmatrix["blubb", "fnord"].should == 0  
+  end
+
+  it "should delete the item count on deletion" do    
+    @matrix.add_set("user123", ["foo", "bar", "fnord"])
+    @matrix.add_set("user456", ["fnord", "blubb"])
+    @matrix.send(:item_count, "fnord").should == 2
+    @matrix.delete_item("fnord")
+    @matrix.send(:item_count, "fnord").should == 0
+  end
+
 end

metadata

@@ -2,7 +2,7 @@
 name: recommendify
 version: !ruby/object:Gem::Version 
   prerelease: 
-  version: 0.1.0
+  version: 0.2.0
 platform: ruby
 authors: 
 - Paul Asmuth