bio-band 0.1.3 → 0.1.4

data/lib/bio-band.rb

@@ -7,4 +7,6 @@ require "java"
 require "bio-band/core"
 require "bio-band/weka"
 require "bio-band/apache"
+Random.ancestors[1].instance_eval {remove_const :Random} if defined?(Random)
+java_import 'java.util.Random'
 

data/lib/bio-band/apache/stat/inference.rb

@@ -2,28 +2,28 @@ require 'java'
 
 module Apache
   module Stat
-  	module Inference
+    module Inference
 
-  		java_import 'org.apache.commons.math3.stat.inference.ChiSquareTest'
-  		java_import 'org.apache.commons.math3.stat.inference.MannWhitneyUTest'
-  		java_import 'org.apache.commons.math3.stat.inference.OneWayAnova'
-  		java_import 'org.apache.commons.math3.stat.inference.TTest'
-  		java_import 'org.apache.commons.math3.stat.inference.WilcoxonSignedRankTest'
-  		java_import 'org.apache.commons.math3.stat.StatUtils'
+      java_import 'org.apache.commons.math3.stat.inference.ChiSquareTest'
+      java_import 'org.apache.commons.math3.stat.inference.MannWhitneyUTest'
+      java_import 'org.apache.commons.math3.stat.inference.OneWayAnova'
+      java_import 'org.apache.commons.math3.stat.inference.TTest'
+      java_import 'org.apache.commons.math3.stat.inference.WilcoxonSignedRankTest'
+      java_import 'org.apache.commons.math3.stat.StatUtils'
       java_import 'java.util.ArrayList'
-  		
+      
       # An implementation of the Wilcoxon signed-rank test
       # * *Args*    :
       #   - +Array1+ -> must be a RubyArray.
       #   - +Array2+ -> must be a RubyArray.
-  		def self.wilcoxon_test(array_1,array_2)
-  			obj = WilcoxonSignedRankTest.new
+      def self.wilcoxon_test(array_1,array_2)
+        obj = WilcoxonSignedRankTest.new
         first = Core::Utils::double_to_a(array_1)
         second = Core::Utils::double_to_a(array_2)
-  			val = obj.wilcoxonSignedRank first, second
+        val = obj.wilcoxonSignedRank first, second
         p_val = obj.wilcoxonSignedRankTest first, second, true.to_java(:boolean) 
-  			return val,p_val
-  		end
+        return val,p_val
+      end
 
       # Utility class called by 'chi_square' method in this same package
       class Chi_square
@@ -61,6 +61,7 @@ module Apache
         end
       end    
 
+      # Compare two datasets stored in Ruby Arrays
       def self.chi_square_dataset_compare(observed1,observed2)
         obj = ChiSquareTest.new
         val = obj.chiSquareDataSetsComparison(observed1.to_java(:long),observed2.to_java(:long)) 
@@ -68,12 +69,17 @@ module Apache
         return val,p_value          
       end
 
-      def mann_whitney_u(array1,array2)
+      # An implementation of the Mann-Whitney U test 
+      # (also called Wilcoxon rank-sum test)
+      # * *Args*    :
+      #   - +Array1+ -> must be a RubyArray.
+      #   - +Array2+ -> must be a RubyArray.
+      def self.mann_whitney_u(array1,array2)
         obj = MannWhitneyUTest.new
-        first = array_1.to_java :double
-        second = array_2.to_java :double 
-        value = mannWhitneyU first,second
-        p_value = mannWhitneyUTest first,second
+        first = array1.to_java :double
+        second = array2.to_java :double 
+        value = obj.mannWhitneyU first,second
+        p_value = obj.mannWhitneyUTest first,second
         return value,p_value
       end
 
@@ -140,6 +146,6 @@ module Apache
         p_value = obj.anovaPValue(collection)
         return f_value,p_value
       end
-  	end
+    end
   end
 end

data/lib/bio-band/apache/stat/regression.rb

@@ -2,7 +2,7 @@ require 'java'
 
 module Apache
   module Stat
-  	module Regression
+    module Regression
       java_import "org.apache.commons.math3.stat.regression.SimpleRegression"
 
       # Create a simple regression model on the input data 
@@ -17,6 +17,6 @@ module Apache
       end
 
 
-  	end
+    end
   end
 end

data/lib/bio-band/core/parser/parser.rb

@@ -2,22 +2,22 @@ require 'java'
 
 module Core
   module Parser
-  	# Parse an ARFF file and create an Instances object
-  	def Parser.parse_ARFF(arff_file) 
+    # Parse an ARFF file and create an Instances object
+    def Parser.parse_ARFF(arff_file) 
       java_import 'java.io.FileReader'
       file_in = FileReader.new arff_file
       data_instance = Core::Type::Instances.new file_in
       return data_instance
-  	end
+    end
     # Parse an CSV file and create an Instances object
-  	def Parser.parse_CSV(csv_file) 
-  	  java_import 'weka.core.converters.CSVLoader'
+    def Parser.parse_CSV(csv_file) 
+      java_import 'weka.core.converters.CSVLoader'
       java_import 'java.io.File'
       loader = CSVLoader.new
       file = File.new csv_file 
       loader.setSource(file)
       data_instance = loader.getDataSet
       return data_instance
-  	end
+    end
   end
 end

data/lib/bio-band/core/type/instances.rb

@@ -61,10 +61,18 @@ module Core
         enumerate_instances.each {|inst| yield(inst)}
       end
 
+      def each_row_with_index
+        enumerate_instances.each_with_index {|inst,id| yield(inst,id)}
+      end
+
       def each_column
         enumerate_attributes.each {|attribute| yield(attribute)}
       end
 
+      def each_column_with_index
+        enumerate_attributes.each_with_index {|attribute,id| yield(attribute,id)}
+      end
+
       # Check if this instance's attributes are all Numeric
       def check_numeric_instance
         enumerateAttributes.each do |att|
@@ -226,7 +234,7 @@ module Core
         summary = Ruport::Data::Table::new
         summary.add_column 'Attributes'
         enumerateAttributes.each_with_index do |att,idx| 
-          summary.add_column idx+1
+          summary.add_column idx
         end
  
         att_names = ['Names']
@@ -244,13 +252,15 @@ module Core
         end
         summary << att_types
 
-        puts summary
+        display = []
+        display << summary
 
         unless enumerate_instances.nil?
           count=0
           enumerateInstances.each {|inst| count=count+1}
-          puts "\nNumber of rows: #{count}"
-        end 
+          display << "\nNumber of rows: #{count}"
+        end
+        display 
       end
 
       # Merges two sets of Instances together. The resulting set will have all the
@@ -321,7 +331,7 @@ module Core
       # Return a json String for the current Instances object
       # The output is modeled on the 'datatable' Google charts APIs 
       # More details at: 'https://developers.google.com/chart/interactive/docs/reference#DataTable'
-      def to_json
+      def to_json_format
         dataset_hash = Hash.new
         dataset_hash[:cols] = enumerateAttributes.collect {|attribute| attribute.name}
         dataset_hash[:rows] = enumerateInstances.collect {|instance| instance.toString} 

data/lib/bio-band/weka/attribute_selection/attribute_selection_utils.rb

@@ -6,11 +6,13 @@ module Attribute_selection_Utils
     listOptions.map {|key| "#{key.synopsis} #{key.description}"}.join("\n")
   end 
 
+  #Set options for an evaluator or a search algorithm
   def select_options(options_string)
     options = Utils.splitOptions(options_string)
     set_options(options)
   end
 
+  #Return a short description for the selected evalutator object or search algorithm
   def description
     globalInfo
  end

data/lib/bio-band/weka/attribute_selection/evaluators.rb

@@ -4,12 +4,13 @@ require 'attribute_selection_utils'
 module Weka
   module Attribute_selection
     module Evaluator
+      #This module contains evaluators from the 'weka.attributeSelection' packages
       java_import 'weka.attributeSelection.CfsSubsetEval'
       java_import 'weka.attributeSelection.ChiSquaredAttributeEval'
 
       class CfsSubsetEval
         include Attribute_selection_Utils 
-        java_alias :use_options , :setOptions, [Java::Java.lang.String[]]
+      #  java_alias :use_options , :setOptions, [Java::Java.lang.String[]]
       end
 
       class ChiSquaredAttributeEval

data/lib/bio-band/weka/attribute_selection/search.rb

@@ -3,6 +3,7 @@ require 'attribute_selection_utils'
 
 module Weka
   module Attribute_selection
+    #This module contains search algorithms from the 'weka.attributeSelection' packages
     module Search
       
       java_import 'weka.attributeSelection.GreedyStepwise'

data/lib/bio-band/weka/classifiers/bayes/bayes.rb

@@ -3,6 +3,7 @@ require 'bayes_utils'
 
 module Weka
   module Classifier
+    #This module stores the classifiers from the 'weka.classifiers.bayes' package
     module Bayes
       java_import "weka.classifiers.bayes.NaiveBayes"
       java_import "weka.classifiers.bayes.BayesianLogisticRegression"

data/lib/bio-band/weka/classifiers/bayes/bayes_utils.rb

@@ -16,10 +16,12 @@ module Bayes_utils
     build_classifier(@dataset)
   end
 
+  # set data for instance classifier
   def set_data(data)
     @dataset = data 
   end
 
+  # set class index for the input dataset
   def set_class_index(class_index)
     @class_index = class_index 
   end
@@ -29,6 +31,7 @@ module Bayes_utils
     base.extend(ClassMethods)
   end
 
+  # set classifier options
   def set_options(options)
     options_inst = Utils.splitOptions(options)
     setOptions(options_inst)
@@ -38,14 +41,24 @@ module Bayes_utils
     listOptions.map {|key| "#{key.synopsis} #{key.description}"}.join("\n")
   end
 
+  # return the description reported in the Weka Java doc
   def description
     globalInfo
   end
 
+  # perform crossvalidation on a trained classifier
+  #ARGV:
+  #fold -> 'int' value
   def cross_validate(fold)
-    eval = Weka::Classifier::Evaluation.new self.class.data
-    eval.crossValidateModel(self.class.ancestors[2].new, self.class.data, fold.to_java(:int), Random.new(1))
-    eval.summary
+    if self.class.data
+      eval = Weka::Classifier::Evaluation.new self.class.data
+      eval.crossValidateModel(self.class.ancestors[2].new, self.class.data, fold.to_java(:int), Random.new(1))
+      eval.summary 
+    else
+      eval = Weka::Classifier::Evaluation.new @dataset
+      eval.crossValidateModel(self.class.ancestors[1].new, @dataset, fold.to_java(:int), Random.new(1))
+      eval.summary
+    end
   end
 
   #Class methods module

data/lib/bio-band/weka/classifiers/evaluation.rb

@@ -1,13 +1,13 @@
 module Weka
-	module Classifier
-		java_import 'weka.classifiers.Evaluation'
-    java_import 'java.util.Random'
+  module Classifier
+    java_import 'weka.classifiers.Evaluation'
 
-		class Evaluation
-			def summary
-				toSummaryString
-			end
-		end
+    # Weka Evaluation class to be used with classfiers and clusterers
+    class Evaluation
+      def summary
+        toSummaryString
+      end
+    end
     
-	end
+  end
 end

data/lib/bio-band/weka/classifiers/functions/functions.rb

@@ -3,6 +3,7 @@ require 'functions_utils'
 
 module Weka
 	module Classifier
+    #This module stores the classifiers from the 'weka.classifiers.functions' package
 		module Functions
       java_import 'weka.classifiers.functions.LinearRegression'
       java_import 'weka.classifiers.functions.PLSClassifier'

data/lib/bio-band/weka/classifiers/functions/functions_utils.rb

@@ -20,10 +20,12 @@ module Functions_utils
     build_classifier(@dataset)
   end
 
+  #Set instance data for the instance classifier
   def set_data(data)
     @dataset = data 
   end
 
+  #Set a class index for the input dataset
   def set_class_index(class_index)
     @class_index = class_index 
   end
@@ -33,18 +35,29 @@ module Functions_utils
     setOptions(options_inst)
   end
 
+  #List available options
   def list_options
     listOptions.map {|key| "#{key.synopsis} #{key.description}"}.join("\n")
   end
 
+  #Return a description from the Weka Javadoc for the selected classifier
   def description
     puts globalInfo
   end
 
+  # perform crossvalidation on a trained classifier
+  #ARGV:
+  #fold -> 'int' value
   def cross_validate(fold)
-    eval = Weka::Classifier::Evaluation.new self.class.data
-    eval.crossValidateModel(self.class.ancestors[2].new, self.class.data, fold.to_java(:int), Random.new(1))
-    eval.summary
+    if self.class.data
+      eval = Weka::Classifier::Evaluation.new self.class.data
+      eval.crossValidateModel(self.class.ancestors[2].new, self.class.data, fold.to_java(:int), Random.new(1))
+      eval.summary 
+    else
+      eval = Weka::Classifier::Evaluation.new @dataset
+      eval.crossValidateModel(self.class.ancestors[1].new, @dataset, fold.to_java(:int), Random.new(1))
+      eval.summary
+    end
   end
 
   #Class methods module

data/lib/bio-band/weka/classifiers/lazy/lazy_utils.rb

@@ -20,31 +20,49 @@ module Lazy_utils
     build_classifier(@dataset)
   end
 
+  #Set data for instance classifier
+  #ARGV
+  # data -> an Instances object
   def set_data(data)
     @dataset = data 
   end
 
+  #Set a class index for the input dataset
   def set_class_index(class_index)
     @class_index = class_index 
   end
 
+  #Set options for the selected classfier
+  #ARGS:
+  #options -> a String, i.e. "-K"
   def set_options(options)
     options_inst = Utils.splitOptions(options)
     setOptions(options_inst)
   end
 
+  #List available options
   def list_options
     listOptions.map {|key| "#{key.synopsis} #{key.description}"}.join("\n")
   end
 
+  #Return a description from the Weka JavaDoc for the selected classifier
   def description
     puts globalInfo
   end
 
+  # perform crossvalidation on a trained classifier
+  #ARGV:
+  #fold -> 'int' value
   def cross_validate(fold)
-    eval = Weka::Classifier::Evaluation.new self.class.data
-    eval.crossValidateModel(self.class.ancestors[2].new, self.class.data, fold.to_java(:int), Random.new(1))
-    eval.summary
+    if self.class.data
+      eval = Weka::Classifier::Evaluation.new self.class.data
+      eval.crossValidateModel(self.class.ancestors[2].new, self.class.data, fold.to_java(:int), Random.new(1))
+      eval.summary 
+    else
+      eval = Weka::Classifier::Evaluation.new @dataset
+      eval.crossValidateModel(self.class.ancestors[1].new, @dataset, fold.to_java(:int), Random.new(1))
+      eval.summary
+    end
   end
 
   #Class methods module

data/lib/bio-band/weka/classifiers/mi/mi.rb

@@ -4,6 +4,7 @@ require 'mi_utils'
 module Weka
 	module Classifier
 		module Mi
+      #This module contains classifiers from the 'weka.classifiers.mi' package
       java_import 'weka.classifiers.mi.CitationKNN'
       java_import 'weka.classifiers.mi.MDD'
       java_import 'weka.classifiers.mi.MIBoost'

data/lib/bio-band/weka/classifiers/mi/mi_utils.rb

@@ -20,10 +20,14 @@ module Mi_utils
     build_classifier(@dataset)
   end
 
+  #Set input data for the selected classifier
+  #ARGV:
+  #data -> an Instances class object
   def set_data(data)
     @dataset = data 
   end
 
+  #Set class index for the input dataset
   def set_class_index(class_index)
     @class_index = class_index 
   end
@@ -33,18 +37,29 @@ module Mi_utils
     setOptions(options_inst)
   end
 
+  #List options for the selected classifier
   def list_options
     listOptions.each {|key| puts "#{key.synopsis} #{key.description}"}
   end
 
+  #Return a short description for the current classifier
   def description
     puts globalInfo
   end
 
+  # perform crossvalidation on a trained classifier
+  #ARGV:
+  #fold -> 'int' value
   def cross_validate(fold)
-    eval = Weka::Classifier::Evaluation.new self.class.data
-    eval.crossValidateModel(self.class.ancestors[2].new, self.class.data, fold.to_java(:int), Random.new(1))
-    eval.summary
+    if self.class.data
+      eval = Weka::Classifier::Evaluation.new self.class.data
+      eval.crossValidateModel(self.class.ancestors[2].new, self.class.data, fold.to_java(:int), Random.new(1))
+      eval.summary 
+    else
+      eval = Weka::Classifier::Evaluation.new @dataset
+      eval.crossValidateModel(self.class.ancestors[1].new, @dataset, fold.to_java(:int), Random.new(1))
+      eval.summary
+    end
   end
 
   #Class methods module