lsh 0.1.0 → 0.1.1

data/lib/lsh/math_util_gsl.rb

@@ -31,6 +31,10 @@ module LSH
       @@gsl_random.gaussian(1, dim)
     end
 
+    def self.zeros(dim)
+      GSL::Vector.alloc(dim)
+    end
+
     def self.random_gaussian_matrix(k, l)
       GSL::Matrix.randn(k, l)
     end
@@ -70,6 +74,17 @@ module GSL
       alloc(*o['data'])
     end
 
+    def hash
+      to_a.hash
+    end
+
+    def save(file)
+      fwrite(file)
+    end
+
+    def load(file)
+      fread(file)
+    end
   end
 
 end

data/lib/lsh/math_util_jblas.rb

@@ -32,6 +32,10 @@ module LSH
       JBLAS.randn(k, l)
     end
 
+    def self.zeros(dim)
+      JBLAS::DoubleMatrix.new(1, dim)
+    end
+
     def self.dot(v1, v2)
       (v1 * v2.t)[0,0]
     end

data/lib/lsh/storage/redis_backend.rb

@@ -23,18 +23,25 @@ module LSH
 
     class RedisBackend
 
-      attr_reader :redis
+      attr_reader :redis, :data_dir
 
-      def initialize(params = { :redis => { :host => '127.0.0.1', :port => 6379 } })
+      def initialize(params = { :redis => { :host => '127.0.0.1', :port => 6379 }, :data_dir => 'data' })
         @redis = Redis.new(params[:redis])
+        @data_dir = params[:data_dir]
+        Dir.mkdir(@data_dir) unless File.exists?(@data_dir)
       end
 
       def reset!
         @redis.flushall
+        Dir.foreach(@data_dir) {|f| File.delete(File.join(@data_dir, f)) if f != '.' and f != '..' and f.end_with?('.dat')}
       end
 
       def has_index?
-        projections and parameters and @redis.get("buckets") > 0
+        projections and parameters and number_of_buckets > 0
+      end
+
+      def number_of_buckets
+        @redis.get("buckets") || 0
       end
 
       def projections=(projections)
@@ -68,7 +75,8 @@ module LSH
       end
 
       def add_vector_to_bucket(bucket, hash, vector)
-        @redis.sadd "#{bucket}:#{hash}", vector.to_json
+        vector.save(File.join(@data_dir, vector.hash.to_s+'.dat')) # Writing vector to disk
+        @redis.sadd "#{bucket}:#{hash}", vector.hash.to_s # Only storing vector's hash in Redis
       end
 
       def find_bucket(i)
@@ -76,7 +84,13 @@ module LSH
       end
 
       def query_bucket(bucket, hash)
-        @redis.smembers("#{bucket}:#{hash}").map { |vector_json| JSON.parse(vector_json) }
+        results = []
+        @redis.smembers("#{bucket}:#{hash}").map do |vector_hash|
+          vector = MathUtil.zeros(parameters[:dim])
+          vector.load(File.join(@data_dir, vector_hash+'.dat'))
+          results << vector
+        end
+        results
       end
 
     end

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification
 name: lsh
 version: !ruby/object:Gem::Version
-  version: 0.1.0
+  version: 0.1.1
   prerelease: 
 platform: ruby
 authors:
@@ -9,7 +9,7 @@ authors:
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2012-12-27 00:00:00.000000000 Z
+date: 2012-12-28 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: gsl