dataset.rb in opentox-ruby-2.0.0

- old
+ new

@@ -72,11 +72,11 @@
 
     # Get all datasets from a service
     # @param [optional,String] uri URI of the dataset service, defaults to service specified in configuration
     # @return [Array] Array of dataset object without data (use one of the load_* methods to pull data from the server)
     def self.all(uri=CONFIG[:services]["opentox-dataset"], subjectid=nil)
-      RestClientWrapper.get(uri,{:accept => "text/uri-list",:subjectid => subjectid}).to_s.each_line.collect{|u| Dataset.new(u, subjectid)}
+      RestClientWrapper.get(uri,{:accept => "text/uri-list",:subjectid => subjectid}).to_s.each_line.collect{|u| Dataset.new(u.chomp, subjectid)}
     end
 
     # Load YAML representation into the dataset
     # @param [String] yaml YAML representation of the dataset
     # @return [OpenTox::Dataset] Dataset object with YAML data
@@ -156,33 +156,46 @@
     end
 
     # Load and return only features from the dataset service
     # @return [Hash]  Features of the dataset
     def load_features(subjectid=nil)
-      parser = Parser::Owl::Dataset.new(@uri, subjectid)
-      @features = parser.load_features(subjectid)
+      if (CONFIG[:yaml_hosts].include?(URI.parse(@uri).host))
+        @features = YAML.load(RestClientWrapper.get(File.join(@uri,"features"), {:accept => "application/x-yaml", :subjectid => subjectid}))
+      else
+        parser = Parser::Owl::Dataset.new(@uri, subjectid)
+        @features = parser.load_features(subjectid)
+      end
       @features
     end
 
+    def feature_classes(feature, subjectid=nil)
+      if Feature.find(feature, subjectid).feature_type == "classification"
+        classes = []
+        @data_entries.each do |c,e|
+          e[feature].each { |v| classes << v.to_s }
+        end
+        classes.uniq.sort
+      else
+        nil
+      end
+    end
+
+=begin
     # Detect feature type(s) in the dataset
     # @return [String] `classification", "regression", "mixed" or unknown`
     def feature_type(subjectid=nil)
       load_features(subjectid)
-      feature_types = @features.collect{|f,metadata| metadata[OT.isA]}.uniq
-      if feature_types.size > 1
-        "mixed"
+      feature_types = @features.collect{|f,metadata| metadata[RDF.type]}.flatten.uniq
+      if feature_types.include?(OT.NominalFeature)
+        "classification"
+      elsif feature_types.include?(OT.NumericFeature)
+        "regression"
       else
-        case feature_types.first
-        when /NominalFeature/
-          "classification"
-        when /NumericFeature/
-          "regression"
-        else
-          "unknown"
-        end
+        "unknown"
       end
     end
+=end
 
     # Get Spreadsheet representation
     # @return [Spreadsheet::Workbook] Workbook which can be written with the spreadsheet gem (data_entries only, metadata will will be discarded))
     def to_spreadsheet
       Serializer::Spreadsheets.new(self).to_spreadsheet
@@ -282,10 +295,10 @@
       if features.size==0
         compounds.each{ |c| dataset.add_compound(c) }
       else
         compounds.each do |c|
           features.each do |f|
-            unless @data_entries[c][f]
+            if @data_entries[c]==nil or @data_entries[c][f]==nil
               dataset.add(c,f,nil)
             else
               @data_entries[c][f].each do |v|
                 dataset.add(c,f,v)
               end