fingerprint csv export

1 files changed, 31 insertions, 2 deletions

diff --git a/lib/dataset.rb b/lib/dataset.rb
index fb1afd2..49208ce 100644
--- a/lib/dataset.rb
+++ b/lib/dataset.rb
@@ -349,6 +349,30 @@ module OpenTox
 
     # Serialisation
     
+    # Convert dataset into csv formatted training data
+    # @return [String]
+    def to_training_csv 
+      
+      export_features = merged_features
+      export_features = transformed_bioactivity_features if export_features.empty? 
+      export_features = bioactivity_features if export_features.empty? 
+      export_feature = export_features.first
+
+      header = ["Canonical SMILES"]
+      header << bioactivity_features.first.name # use original bioactivity name instead of long merged name
+      csv = [header]
+
+      substances.each do |substance|
+        nr_activities = values(substance,bioactivity_features.first).size
+        (0..nr_activities-1).each do |n| # new row for each value
+          row = [substance.smiles]
+          row << values(substance,export_feature)[n] 
+          csv << row
+        end
+      end
+      csv.collect{|r| r.join(",")}.join("\n")
+    end
+    
     # Convert lazar prediction dataset to csv format 
     # @return [String]
     def to_prediction_csv
@@ -376,16 +400,18 @@ module OpenTox
       csv.collect{|r| r.join(",")}.join("\n")
     end
     
-    # Convert dataset into csv formatted training data
+    # Export fingerprints in csv format
     # @return [String]
-    def to_training_csv 
+    def to_fingerprint_csv type=Compound::DEFAULT_FINGERPRINT
       
+      fingerprints = substances.collect{|s| s.fingerprints[type]}.flatten.sort.uniq
       export_features = merged_features
       export_features = transformed_bioactivity_features if export_features.empty? 
       export_features = bioactivity_features if export_features.empty? 
       export_feature = export_features.first
 
       header = ["Canonical SMILES"]
+      header += fingerprints
       header << bioactivity_features.first.name # use original bioactivity name instead of long merged name
       csv = [header]
 
@@ -393,6 +419,9 @@ module OpenTox
         nr_activities = values(substance,bioactivity_features.first).size
         (0..nr_activities-1).each do |n| # new row for each value
           row = [substance.smiles]
+          fingerprints.each do |f|
+            substance.fingerprints[type].include?(f) ? row << 1 : row << 0
+          end
           row << values(substance,export_feature)[n] 
           csv << row
         end