bio-band 0.1.0

data/ext/mkrf_conf.rb

@@ -0,0 +1,74 @@
+path = File.expand_path(File.dirname(__FILE__))
+
+def command?(name)
+  `which #{name}`
+  $?.success?
+end
+
+module OS
+  def OS.windows?
+    (/cygwin|mswin|mingw|bccwin|wince|emx/ =~ RbConfig::CONFIG['host_os']) != nil
+  end
+
+  def OS.mac?
+   (/darwin/ =~ RbConfig::CONFIG['host_os']) != nil
+  end
+
+  def OS.unix?
+    !OS.windows?
+  end
+  
+  def OS.linux?
+    OS.unix? and not OS.mac?
+  end
+end
+
+File.open(File.join(path,"Rakefile"),"w") do |rakefile|
+  
+  if OS.windows? == true
+    puts "Sorry, still no support is provided for your OS!" 
+  
+  elsif OS.mac? == true
+    if command?("mvn")==false && command?("brew")==true
+      rakefile.write <<-RAKE
+        task :brew_install do
+          sh "brew install maven"
+        end
+        task :default => [:brew_install]
+      RAKE
+    elsif command?("brew")==false
+      rakefile.write <<-RAKE
+      task :ok_inst do
+        puts "Sorry, Maven could not be installed. Try installing 'brew' first"
+      end
+      task :default => [:ok_inst]
+RAKE
+    else
+      rakefile.write <<-RAKE
+    task :ok_inst do
+      puts "Maven has been detected on your system"
+    end
+    task :default => [:ok_inst]
+RAKE
+    end
+  elsif OS.linux? == true
+    if command?("mvn")==false
+      rakefile.write <<-RAKE
+  	    task :apt_install do
+          sh "sudo apt-get install maven2"
+        end
+        task :default => [:apt_install]
+      RAKE
+    else
+      rakefile.write <<-RAKE
+    task :ok_inst do
+      puts "Maven has been detected on your system"
+    end
+    task :default => [:ok_inst]
+RAKE
+    end
+  end
+end
+
+
+

data/features/create_dataset.feature

@@ -0,0 +1,12 @@
+Feature: Creation of an in-memory dataset
+	In order to perform calculations on a dataset
+	I want to easily build it and store it in memory
+
+	Scenario: creation of a dataset
+		Given a nominal attribute, named "assertion", with values "yes,no" 
+		Given one numeric attribute, named "temperature"
+		Given another numeric attribute, names "days"
+		Given two data rows: "yes,100,30","no,100,0"
+		Then I want to build en empty dataset for my use
+		And I want to populate the dataset by row
+		And I want to print my dataset as a bidimensional Ruby Array

data/features/step_definitions/create_dataset.rb

@@ -0,0 +1,40 @@
+Given(/^a nominal attribute, named "(.*?)", with values "(.*?)"$/) do |arg1, arg2|
+  $first_att = arg1.to_sym
+  $values = arg2.split(',')
+end
+
+Given(/^one numeric attribute, named "(.*?)"$/) do |arg1|
+	$second_att = arg1.to_sym
+end
+
+Given(/^another numeric attribute, names "(.*?)"$/) do |arg1|
+	$third_att = arg1.to_sym
+end
+
+Given(/^two data rows: "(.*?)","(.*?)"$/) do |arg1, arg2|
+	@first_row = arg1.split(',')
+	@second_row = arg2.split(',')
+	@first_row[1] = @first_row[1].to_f
+	@first_row[2] = @first_row[2].to_f
+	@second_row[1] = @second_row[1].to_f
+	@second_row[2] = @second_row[2].to_f
+end
+
+Then(/^I want to build en empty dataset for my use$/) do
+	class Dataset < Core::Type::Instances::Base
+		nominal $first_att, $values
+  	numeric $second_att
+  	string $third_att
+  end
+  @my_instance = Dataset.new
+  @my_instance.summary
+end
+
+Then(/^I want to populate the dataset by row$/) do
+	@my_instance.populate_by_row([@first_row,@second_row])
+	@my_instance.summary
+end
+
+Then(/^I want to print my dataset as a bidimensional Ruby Array$/) do
+	puts @my_instance.to_a2d.inspect
+end

data/features/step_definitions/weka_classifiers.rb

@@ -0,0 +1,42 @@
+Given(/^the Weka "(.*?)" classifier$/) do |arg1|
+  @classifier = Weka::Classifier::Bayes::NaiveBayes.new
+end
+
+Then(/^I want to print a "(.*?)"$/) do |arg1|
+  @classifier.send arg1.to_sym
+end
+
+Then(/^I want to print an options list$/) do
+  @classifier.list_options
+end
+
+Given(/^the unsupervised Weka classifier "(.*?)"$/) do |arg1|
+  class My_classifier < Weka::Classifier::Bayes::NaiveBayes::Base
+  end
+end
+
+Then(/^I want to set option "(.*?)" for it$/) do |arg1|
+  My_classifier.set_options(arg1)
+end
+
+Then(/^I want to set the dataset parsed from "(.*?)"$/) do |arg1|
+  @arff = File.join('resources',arg1)
+  @dataset_ARFF = Core::Parser::parse_ARFF(@arff)
+  My_classifier.set_data(@dataset_ARFF)
+end
+
+Then(/^I want to print a summary for the dataset$/) do
+  @dataset_ARFF.summary
+end
+
+Then(/^I want to set the class index for attribute with index "(.*?)"$/) do |arg1|
+    My_classifier.set_class_index(arg1.to_i)
+end
+
+Then(/^I want to instantiate the classifier for my use$/) do
+  My_classifier.new
+end
+
+
+
+

data/features/step_definitions/weka_clustering.rb

@@ -0,0 +1,30 @@
+Given(/^the ARFF dataset stored in the file "(.*?)"$/) do |arg1|
+  @arff = File.join('resources',arg1)
+end
+
+Given(/^the SimpleKMeans algorithm implementation from Weka$/) do
+  @clusterer = Weka::Classifier::Bayes::NaiveBayes.new
+end
+
+Then(/^I want to parse the data from the file$/) do
+  @dataset_ARFF = Core::Parser::parse_ARFF(@arff)
+end
+
+Then(/^I want to list the options available for SimpleKMeans$/) do
+  puts @clusterer.list_options
+end
+
+Then(/^I want to set K = "(.*?)" as K\-means option$/) do |arg1|
+  class Clustering < Weka::Clusterer::SimpleKMeans::Base
+  end
+  Clustering.set_options "-N #{arg1}"
+end
+
+Then(/^I want to perform clustering on the parsed dataset$/) do
+  Clustering.set_data(@dataset_ARFF)
+  @clustered = Clustering.new
+end
+
+Then(/^I want to report result statistics$/) do
+  puts @clustered
+end

data/features/step_definitions/weka_filters.rb

@@ -0,0 +1,29 @@
+Given(/^the example ARFF file "(.*?)"$/) do |arff_file|
+	@arff = File.join('resources',arff_file)
+end
+
+Given(/^the Weka Attribute Add filter$/) do
+	@filter = Weka::Filter::Unsupervised::Attribute::Add.new
+end
+
+Then(/^I want to parse the file in order to create an Instances class object$/) do
+	@dataset_ARFF = Core::Parser::parse_ARFF(@arff)
+	@dataset_ARFF.summary
+end
+
+Then(/^I want to print the available filter options and usage$/) do
+	@filter.filter_options
+end
+
+Then(/^I want to set the option String "(.*?)"$/) do |arg1|
+	@filter.set_filter_options(arg1)
+	@filter.set_data(@dataset_ARFF)
+end
+
+Then(/^I want to add an attribute \(a column\) to the dataset using the Weka filter Add$/) do
+	@new_inst = @filter.use
+end
+
+Then(/^I want to print a "(.*?)" for the modified dataset$/) do |arg1|
+	@new_inst.send arg1.to_sym
+end

data/features/step_definitions/weka_parsers.rb

@@ -0,0 +1,45 @@
+Given /^the CSV file "(.*?)"$/ do |csv_file|
+	@csv = File.join('resources',csv_file)
+	@dataset_CSV = Core::Parser::parse_CSV(@csv)
+end
+
+Then /^I want to print to stdout the summary for the CSV parsed Instances object$/ do
+	@dataset_CSV.summary
+end
+
+Given /^the ARFF file "(.*?)"$/ do |arff_file|
+	@arff = File.join('resources',arff_file)
+	puts @arff
+	@dataset_ARFF = Core::Parser::parse_ARFF(@arff)
+end
+
+Then /^I want to print to stdout the summary for the ARFF parsed Instances object$/ do
+	@dataset_ARFF.summary
+end
+
+Given(/^the database "(.*?)"$/) do |arg1|
+  @db_connection = arg1
+end
+
+Given(/^a table named "(.*?)"$/) do |arg1|
+	@target_table = arg1
+end
+
+Then(/^I want to extract data from that table$/) do
+  @dataset = Weka::Db.query_mysql(@db_connection,'root','',"select * from #{@target_table}")
+end
+
+Then(/^I want to print to stdout the summary for the parsed Instances object$/) do
+  @dataset.summary
+end
+
+Then(/^I want to convert the data into a bidimensional Ruby Array$/) do 
+	@dataset.to_a2d.should == [["sunny", 85.0, 85.0, "FALSE", "no"], ["sunny", 80.0, 90.0, "TRUE", "no"], ["overcast", 83.0, 86.0, "FALSE", "yes"],
+													 ["rainy", 70.0, 96.0, "FALSE", "yes"], ["rainy", 68.0, 80.0, "FALSE", "yes"], ["rainy", 65.0, 70.0, "TRUE", "no"], 
+													 ["overcast", 64.0, 65.0, "TRUE", "yes"], ["sunny", 72.0, 95.0, "FALSE", "no"], ["sunny", 69.0, 70.0, "FALSE", "yes"], 
+													 ["rainy", 75.0, 80.0, "FALSE", "yes"], ["sunny", 75.0, 70.0, "TRUE", "yes"], ["overcast", 72.0, 90.0, "TRUE", "yes"], 
+													 ["overcast", 81.0, 75.0, "FALSE", "yes"], ["rainy", 71.0, 91.0, "TRUE", "no"]]
+end
+
+
+

data/features/support/env.rb

@@ -0,0 +1,3 @@
+$LOAD_PATH << File.expand_path('../../../lib', __FILE__)
+$LOAD_PATH << File.expand_path('../../../resources/', __FILE__)
+require 'ruby_mining'

data/features/weka_classifiers.feature

@@ -0,0 +1,16 @@
+Feature: Using a Weka classifier
+	In order to classify an instance data
+	I want to use a Weka classifier
+
+	Scenario: Understand options and usage
+		Given the Weka "NaiveBayes" classifier
+		Then I want to print a "description"
+		And I want to print an options list
+
+	Scenario: Use a classifier on a data instance
+		Given the unsupervised Weka classifier "NaiveBayes"
+		Then I want to set option "-K" for it
+		And I want to set the dataset parsed from "weather.numeric.arff"
+		And I want to print a summary for the dataset
+		And I want to set the class index for attribute with index "0"
+		And I want to instantiate the classifier for my use

data/features/weka_clustering.feature

@@ -0,0 +1,14 @@
+Feature: Weka dataset clustering 
+	In order to group similar data vectors in my dataset 
+	I want to use Weka clustering algorithms
+
+	Scenario: use of Kmeans algorithm
+		Given the ARFF dataset stored in the file "weather.numeric.arff"
+    Given the SimpleKMeans algorithm implementation from Weka  
+		Then I want to parse the data from the file
+    And I want to list the options available for SimpleKMeans
+    And I want to set K = "4" as K-means option
+    And I want to perform clustering on the parsed dataset
+    And I want to report result statistics 
+
+

data/features/weka_filters.feature

@@ -0,0 +1,12 @@
+Feature: Filter a dataset (Weka Instances class)
+	In order to manipulate a dataset
+	I want to use the Weka Filter class on it
+
+	Scenario: Use of the Unsupervised Attribute filter 'Add'
+		Given the example ARFF file "weather.numeric.arff"
+		Given the Weka Attribute Add filter
+		Then I want to parse the file in order to create an Instances class object
+		And I want to print the available filter options and usage
+		And I want to set the option String "-T NUM -N dummy"
+		And I want to add an attribute (a column) to the dataset using the Weka filter Add
+		And I want to print a "summary" for the modified dataset

data/features/weka_parsers.feature

@@ -0,0 +1,18 @@
+Feature: Weka basic parsing capabilities
+	In order to perform calculations on a dataset 
+	I want to import data from .ARFF, .CSV files, and external databases
+
+	Scenario: parsing a CSV file
+		Given the CSV file "weather.csv"
+		Then I want to print to stdout the summary for the CSV parsed Instances object 
+
+	Scenario: parsing an ARFF file
+		Given the ARFF file "weather.numeric.arff"
+		Then I want to print to stdout the summary for the ARFF parsed Instances object 
+
+	Scenario: parsing data from a mySQL table
+		Given the database "jdbc:mysql://localhost:3306/Gene_classes"
+		And a table named "test_weka"	
+		Then I want to extract data from that table
+		And I want to print to stdout the summary for the parsed Instances object
+		And I want to convert the data into a bidimensional Ruby Array

data/features/weka_pipeline.feature

@@ -0,0 +1,13 @@
+Feature: Classification pipeline
+	In order to perform text-mining on a dataset
+	I want to parse the data, filter it and then classify it using a Bayesian classifier
+
+	Scenario: Use of Naive Bayes on a filtered data set
+		Given a file containing the training set data "ReutersGrain-train.arff"
+		Given a file containing the test set data "ReutersGrain-test.arff"
+		Then I want to parse them
+		And I want to filter them using the unsupervised filter "StringToWordVector"
+		And I want to build a "NaiveBayes" classifier using training set data
+		And I want to evaluate the performance of the classifier on the test test
+		And I want want to print to stdout a "summary" for the evaluation
+		

data/lib/bio-band.rb

@@ -0,0 +1,10 @@
+start = File.absolute_path '.'
+Dir.chdir(File.join(File.dirname(__FILE__),".."))
+require 'jbundler'
+Dir.chdir start
+require File.join(File.dirname(__FILE__),'..','.jbundler','classpath.rb')
+require "java"
+require "bio-band/core"
+require "bio-band/weka"
+require "bio-band/apache"
+

data/lib/bio-band/apache.rb

@@ -0,0 +1 @@
+require 'bio-band/apache/stat/inference.rb'

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

@@ -0,0 +1,145 @@
+require 'java'
+
+module Apache
+  module Stat
+  	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 '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
+        first = Core::Utils::double_to_a(array_1)
+        second = Core::Utils::double_to_a(array_2)
+  			val = obj.wilcoxonSignedRank first, second
+        p_val = obj.wilcoxonSignedRankTest first, second, true.to_java(:boolean) 
+  			return val,p_val
+  		end
+
+      # Utility class called by 'chi_square' method in this same package
+      class Chi_square
+        def self.chi_square_2d(array_2d)
+          obj = ChiSquareTest.new
+          val = obj.chi_square(array_2d.to_java(Java::long[]))
+          p_value =  obj.chi_square_test(array_2d.to_java(Java::long[]))
+          return val,p_value
+        end
+        
+        def self.chi_square_two_arrays(expected,observed)
+          obj = ChiSquareTest.new
+          val = obj.chi_square(expected.to_java(:double),observed.to_java(:long))
+          p_value = obj.chi_square_test(expected.to_java(:double),observed.to_java(:long))
+          return val,p_value
+        end
+      end
+
+      # 1) Computes the Chi-Square statistic comparing observed and expected frequency counts.
+      # * *Args*    :
+      #   - +Array+ -> must be a bidimensional RubyArray.
+      # 2) Computes the Chi-Square statistic associated with a chi-square test of independence
+      #   based on the input counts array, viewed as a two-way table.
+      # * *Args*    :
+      #   - +Array1+ -> must be a RubyArray.
+      #   - +Array2+ -> must be a RubyArray.
+      def self.chi_square(*args)
+        if args.length == 2
+          Chi_square.chi_square_two_arrays(*args)
+        elsif args.length == 1
+          raise ArgumentError,"RubyArray must be bidimensional" unless args[0].is_2d?
+          Chi_square.chi_square_2d(*args)
+        else
+          raise ArgumentError, 'Function *args should be two RubyArrays or a bidimensional RubyArray'
+        end
+      end    
+
+      def self.chi_square_dataset_compare(observed1,observed2)
+        obj = ChiSquareTest.new
+        val = obj.chiSquareDataSetsComparison(observed1.to_java(:long),observed2.to_java(:long)) 
+        p_value = obj.chiSquareTestDataSetsComparison(observed1.to_java(:long),observed2.to_java(:long))
+        return val,p_value          
+      end
+
+      def 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
+        return value,p_value
+      end
+
+      #Utility class called by 't_test' method in this same package
+      class T_test
+  
+        def self.homoscedastic(array_1,array_2)
+          obj = TTest.new
+          first = array_1.to_java :double
+          second = array_2.to_java :double
+          value = obj.homoscedasticT(first,second)
+          p_value = obj.homoscedasticTTest(first,second)
+          return value, p_value
+        end
+  
+        def self.paired(array_1,array_2)
+          obj = TTest.new
+          first = array_1.to_java :double
+          second =array_2.to_java :double
+          value = obj.pairedT(first,second)
+          p_value = obj.pairedTTest(first,second)
+          return value,p_value
+        end
+
+        def self.t(array_1,array_2)
+          obj = TTest.new
+          first = array_1.to_java :double
+          second =array_2.to_java :double
+          value = obj.t(first,second)
+          p_value =obj.tTest(first,second)
+          return value,p_value
+        end
+      end
+
+      # An implementation for Student's t-tests
+      # * *Args*    :
+      #   - +sample_1+ -> an array of numeric values representing a sample
+      #   - +sample_2+ -> an array of numeric values representing a sample
+      #   - +homoscedastic+ -> set to true for equal variance assumption
+      #   - +paired+ -> set to true if you want to perform a 'paired' t test
+      def self.t_test(sample_1,sample_2,homoscedastic=false,paired=false)
+        if homoscedastic == true
+          T_test.homoscedastic(sample_1,sample_2)
+        elsif paired == true
+          T_test.paired(sample_1,sample_2)
+        else
+          T_test.t(sample_1,sample_2)
+        end 
+      end
+
+      # Implements one-way ANOVA (analysis of variance) statistics.
+      # Tests for differences between two or more categories of univariate data (for example,
+      # the body mass index of accountants, lawyers, doctors and computer programmers). When 
+      # two categories are given, this is equivalent to the TTest.
+      # * *Args*    :
+      #   - +bidimensional_array+ -> a 2d RubyArray
+      def self.one_way_anova(bidimensional_array)
+        collection = ArrayList.new 
+        bidimensional_array.each do |array|
+          collection.add(array.to_java :double)
+        end
+        obj = OneWayAnova.new
+        f_value = obj.anovaFValue(collection)
+        p_value = obj.anovaPValue(collection)
+        return f_value,p_value
+      end
+  	end
+  end
+end