dataset.rb in opentox-ruby-1.0.0

- old
+ new

@@ -12,11 +12,11 @@
     #   dataset = OpenTox::Dataset.new
     # @example Create an empty dataset with URI
     #   dataset = OpenTox::Dataset.new("http:://webservices.in-silico/ch/dataset/1")
     # @param [optional, String] uri Dataset URI
     # @return [OpenTox::Dataset] Dataset object
-    def initialize(uri=nil)
+    def initialize(uri=nil,subjectid=nil)
       super uri
       @features = {}
       @compounds = []
       @data_entries = {}
     end
@@ -24,75 +24,94 @@
     # Create an empty dataset and save it at the dataset service (assigns URI to dataset)
     # @example Create new dataset and save it to obtain a URI 
     #   dataset = OpenTox::Dataset.create
     # @param [optional, String] uri Dataset URI
     # @return [OpenTox::Dataset] Dataset object
-    def self.create(uri=CONFIG[:services]["opentox-dataset"])
-      dataset = Dataset.new
-      dataset.save
+    def self.create(uri=CONFIG[:services]["opentox-dataset"], subjectid=nil)
+      dataset = Dataset.new(nil,subjectid)
+      dataset.save(subjectid)
       dataset
     end
 
     # Create dataset from CSV file (format specification: http://toxcreate.org/help)
     # - loads data_entries, compounds, features
     # - sets metadata (warnings) for parser errors
     # - you will have to set remaining metadata manually
     # @param [String] file CSV file path
     # @return [OpenTox::Dataset] Dataset object with CSV data
-    def self.create_from_csv_file(file) 
-      dataset = Dataset.create
+    def self.create_from_csv_file(file, subjectid=nil) 
+      dataset = Dataset.create(CONFIG[:services]["opentox-dataset"], subjectid)
       parser = Parser::Spreadsheets.new
       parser.dataset = dataset
       parser.load_csv(File.open(file).read)
-      dataset.save
+      dataset.save(subjectid)
       dataset
     end
-
+    
     # Find a dataset and load all data. This can be time consuming, use Dataset.new together with one of the load_* methods for a fine grained control over data loading.
     # @param [String] uri Dataset URI
     # @return [OpenTox::Dataset] Dataset object with all data
-    def self.find(uri)
-      dataset = Dataset.new(uri)
-      dataset.load_all
+    def self.find(uri, subjectid=nil)
+      return nil unless uri
+      dataset = Dataset.new(uri, subjectid)
+      dataset.load_all(subjectid)
       dataset
     end
+    
+    # replaces find as exist check, takes not as long, does NOT raise an un-authorized exception
+    # @param [String] uri Dataset URI
+    # @return [Boolean] true if dataset exists and user has get rights, false else 
+    def self.exist?(uri, subjectid=nil)
+      return false unless uri
+      dataset = Dataset.new(uri, subjectid)
+      begin
+        dataset.load_metadata( subjectid ).size > 0
+      rescue
+        false
+      end
+    end
 
     # 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"])
-      RestClientWrapper.get(uri,:accept => "text/uri-list").to_s.each_line.collect{|u| Dataset.new(u)}
+    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)}
     end
 
     # Load YAML representation into the dataset
     # @param [String] yaml YAML representation of the dataset
     # @return [OpenTox::Dataset] Dataset object with YAML data
     def load_yaml(yaml)
       copy YAML.load(yaml)
     end
 
     def load_rdfxml(rdfxml)
-      load_rdfxml_file Tempfile.open("ot-rdfxml"){|f| f.write(rdfxml)}.path
+      raise "rdfxml data is empty" if rdfxml.to_s.size==0
+      file = Tempfile.new("ot-rdfxml")
+      file.puts rdfxml
+      file.close
+      load_rdfxml_file file
+      file.delete
     end
 
     # Load RDF/XML representation from a file
     # @param [String] file File with RDF/XML representation of the dataset
     # @return [OpenTox::Dataset] Dataset object with RDF/XML data
-    def load_rdfxml_file(file)
-      parser = Parser::Owl::Dataset.new @uri
+    def load_rdfxml_file(file, subjectid=nil)
+      parser = Parser::Owl::Dataset.new @uri, subjectid
       parser.uri = file.path
-      copy parser.load_uri
+      copy parser.load_uri(subjectid)
     end
 
     # Load CSV string (format specification: http://toxcreate.org/help)
     # - loads data_entries, compounds, features
     # - sets metadata (warnings) for parser errors
     # - you will have to set remaining metadata manually
     # @param [String] csv CSV representation of the dataset
     # @return [OpenTox::Dataset] Dataset object with CSV data
-    def load_csv(csv) 
-      save unless @uri # get a uri for creating features
+    def load_csv(csv, subjectid=nil) 
+      save(subjectid) unless @uri # get a uri for creating features
       parser = Parser::Spreadsheets.new
       parser.dataset = self
       parser.load_csv(csv)
     end
 
@@ -100,49 +119,49 @@
     # - loads data_entries, compounds, features
     # - sets metadata (warnings) for parser errors
     # - you will have to set remaining metadata manually
     # @param [Excel] book Excel workbook object (created with roo gem)
     # @return [OpenTox::Dataset] Dataset object with Excel data
-    def load_spreadsheet(book)
-      save unless @uri # get a uri for creating features
+    def load_spreadsheet(book, subjectid=nil)
+      save(subjectid) unless @uri # get a uri for creating features
       parser = Parser::Spreadsheets.new
       parser.dataset = self
       parser.load_spreadsheet(book)
     end
     
     # Load and return only metadata of a Dataset object
     # @return [Hash] Metadata of the dataset
-    def load_metadata
-      add_metadata Parser::Owl::Dataset.new(@uri).load_metadata
+    def load_metadata(subjectid=nil)
+      add_metadata Parser::Owl::Dataset.new(@uri, subjectid).load_metadata(subjectid)
       self.uri = @uri if @uri # keep uri
       @metadata
     end
 
     # Load all data (metadata, data_entries, compounds and features) from URI
-    def load_all
+    def load_all(subjectid=nil)
       if (CONFIG[:yaml_hosts].include?(URI.parse(@uri).host))
-        copy YAML.load(RestClientWrapper.get(@uri, :accept => "application/x-yaml"))
+        copy YAML.load(RestClientWrapper.get(@uri, {:accept => "application/x-yaml", :subjectid => subjectid}))
       else
-        parser = Parser::Owl::Dataset.new(@uri)
-        copy parser.load_uri
+        parser = Parser::Owl::Dataset.new(@uri, subjectid)
+        copy parser.load_uri(subjectid)
       end
     end
 
     # Load and return only compound URIs from the dataset service
     # @return [Array]  Compound URIs in the dataset
-    def load_compounds
-      RestClientWrapper.get(File.join(uri,"compounds"),:accept=> "text/uri-list").to_s.each_line do |compound_uri|
+    def load_compounds(subjectid=nil)
+      RestClientWrapper.get(File.join(uri,"compounds"),{:accept=> "text/uri-list", :subjectid => subjectid}).to_s.each_line do |compound_uri|
         @compounds << compound_uri.chomp
       end
       @compounds.uniq!
     end
 
     # Load and return only features from the dataset service
     # @return [Hash]  Features of the dataset
-    def load_features
-      parser = Parser::Owl::Dataset.new(@uri)
-      @features = parser.load_features
+    def load_features(subjectid=nil)
+      parser = Parser::Owl::Dataset.new(@uri, subjectid)
+      @features = parser.load_features(subjectid)
       @features
     end
 
     # Detect feature type(s) in the dataset
     # @return [String] `classification", "regression", "mixed" or unknown`
@@ -216,11 +235,11 @@
     def add (compound,feature,value)
       @compounds << compound unless @compounds.include? compound
       @features[feature] = {}  unless @features[feature]
       @data_entries[compound] = {} unless @data_entries[compound]
       @data_entries[compound][feature] = [] unless @data_entries[compound][feature]
-      @data_entries[compound][feature] << value
+      @data_entries[compound][feature] << value if value!=nil
     end
 
     # Add/modify metadata, existing entries will be overwritten
     # @example
     #   dataset.add_metadata({DC.title => "any_title", DC.creator => "my_email"})
@@ -240,38 +259,74 @@
     # @param [String] feature Feature URI
     # @param [Hash] metadata Hash with feature metadata
     def add_feature_metadata(feature,metadata)
       metadata.each { |k,v| @features[feature][k] = v }
     end
+    
+    # Add a new compound
+    # @param [String] compound Compound URI
+    def add_compound (compound)
+      @compounds << compound unless @compounds.include? compound
+    end
+    
+    # Creates a new dataset, by splitting the current dataset, i.e. using only a subset of compounds and features
+    # @param [Array] compounds List of compound URIs
+    # @param [Array] features List of feature URIs
+    # @param [Hash] metadata Hash containing the metadata for the new dataset
+    # @param [String] subjectid
+    # @return [OpenTox::Dataset] newly created dataset, already saved
+    def split( compounds, features, metadata, subjectid=nil)
+      LOGGER.debug "split dataset using "+compounds.size.to_s+"/"+@compounds.size.to_s+" compounds"
+      raise "no new compounds selected" unless compounds and compounds.size>0
+      dataset = OpenTox::Dataset.create(CONFIG[:services]["opentox-dataset"],subjectid)
+      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]
+              dataset.add(c,f,nil)
+            else
+              @data_entries[c][f].each do |v|
+                dataset.add(c,f,v)
+              end
+            end
+          end
+        end
+      end
+      dataset.add_metadata(metadata)
+      dataset.save(subjectid)
+      dataset
+    end
 
     # Save dataset at the dataset service 
     # - creates a new dataset if uri is not set
     # - overwrites dataset if uri exists
     # @return [String] Dataset URI
-    def save
+    def save(subjectid=nil)
       # TODO: rewrite feature URI's ??
       @compounds.uniq!
       if @uri
         if (CONFIG[:yaml_hosts].include?(URI.parse(@uri).host))
-          RestClientWrapper.post(@uri,{:content_type =>  "application/x-yaml"},self.to_yaml)
+          RestClientWrapper.post(@uri,self.to_yaml,{:content_type =>  "application/x-yaml", :subjectid => subjectid})
         else
           File.open("ot-post-file.rdf","w+") { |f| f.write(self.to_rdfxml); @path = f.path }
-          task_uri = RestClient.post(@uri, {:file => File.new(@path)},{:accept => "text/uri-list"}).to_s.chomp
+          task_uri = RestClient.post(@uri, {:file => File.new(@path)},{:accept => "text/uri-list" , :subjectid => subjectid}).to_s.chomp
           #task_uri = `curl -X POST -H "Accept:text/uri-list" -F "file=@#{@path};type=application/rdf+xml" http://apps.ideaconsult.net:8080/ambit2/dataset`
           Task.find(task_uri).wait_for_completion
-          self.uri = RestClientWrapper.get(task_uri,:accept => 'text/uri-list')
+          self.uri = RestClientWrapper.get(task_uri,{:accept => 'text/uri-list', :subjectid => subjectid})
         end
       else
         # create dataset if uri is empty
-        self.uri = RestClientWrapper.post(CONFIG[:services]["opentox-dataset"],{}).to_s.chomp
+        self.uri = RestClientWrapper.post(CONFIG[:services]["opentox-dataset"],{:subjectid => subjectid}).to_s.chomp
       end
       @uri
     end
 
     # Delete dataset at the dataset service
-    def delete
-      RestClientWrapper.delete @uri
+    def delete(subjectid=nil)
+      RestClientWrapper.delete(@uri, :subjectid => subjectid)
     end
 
     private
     # Copy a dataset (rewrites URI)
     def copy(dataset)
@@ -291,12 +346,12 @@
   class LazarPrediction < Dataset
 
     # Find a prediction dataset and load all data. 
     # @param [String] uri Prediction dataset URI
     # @return [OpenTox::Dataset] Prediction dataset object with all data
-    def self.find(uri)
-      prediction = LazarPrediction.new(uri)
-      prediction.load_all
+    def self.find(uri, subjectid=nil)
+      prediction = LazarPrediction.new(uri, subjectid)
+      prediction.load_all(subjectid)
       prediction
     end
 
     def value(compound)
       @data_entries[compound.uri].collect{|f,v| v.first if f.match(/prediction/)}.compact.first