From b7cd3ebbb858a8891c35c45896f1bdd525f3534e Mon Sep 17 00:00:00 2001
From: Christoph Helma <helma@in-silico.ch>
Date: Mon, 10 Aug 2015 13:26:06 +0200
Subject: algorithm libraries added, fminer tests pass

---
 lib/similarity.rb | 58 +++++++++++++++++++++++++++++++++++++++++++++++++++++++
 1 file changed, 58 insertions(+)
 create mode 100644 lib/similarity.rb

(limited to 'lib/similarity.rb')

diff --git a/lib/similarity.rb b/lib/similarity.rb
new file mode 100644
index 0000000..91e18db
--- /dev/null
+++ b/lib/similarity.rb
@@ -0,0 +1,58 @@
+=begin
+* Name: similarity.rb
+* Description: Similarity algorithms
+* Author: Andreas Maunz <andreas@maunz.de
+* Date: 10/2012
+=end
+
+module OpenTox
+  module Algorithm
+
+    class Similarity
+
+      #TODO weighted tanimoto
+
+      # Tanimoto similarity
+      # @param [Array] a fingerprints of first compound
+      # @param [Array] b fingerprints of second compound
+      # @return [Float] Tanimoto similarity
+      def self.tanimoto(a,b)
+        bad_request_error "fingerprints #{a} and #{b} don't have equal size" unless a.size == b.size
+        #common = 0.0
+        #a.each_with_index do |n,i|
+          #common += 1 if n == b[i]
+        #end
+        #common/a.size
+        # TODO check if calculation speed can be improved
+        common_p_sum = 0.0
+        all_p_sum = 0.0
+        (0...a.size).each { |idx|
+          common_p_sum += [ a[idx], b[idx] ].min
+          all_p_sum += [ a[idx], b[idx] ].max
+        }
+        common_p_sum/all_p_sum
+      end
+
+
+      # Cosine similarity
+      # @param [Array] a fingerprints of first compound
+      # @param [Array] b fingerprints of second compound
+      # @return [Float] Cosine similarity, the cosine of angle enclosed between vectors a and b
+      def self.cosine(a, b)
+        val = 0.0
+        if a.size>0 and b.size>0
+          if a.size>12 && b.size>12
+            a = a[0..11]
+            b = b[0..11]
+          end
+          a_vec = a.to_gv
+          b_vec = b.to_gv
+          val = a_vec.dot(b_vec) / (a_vec.norm * b_vec.norm)
+        end
+        val
+      end
+
+    end
+
+  end
+end
-- 
cgit v1.2.3