macroape 3.3.2 → 3.3.3

data/lib/macroape/version.rb

@@ -1,3 +1,3 @@
 module Macroape
-  VERSION = "3.3.2"
+  VERSION = "3.3.3"
 end

data/spec/count_distribution_spec.rb

@@ -0,0 +1,52 @@
+require 'spec_helper'
+require 'bioinform'
+require 'macroape/counting'
+
+describe Bioinform::PWM do
+  let :matrix_first do [[1,2,3,4],[10,20,30,40],[100,200,300,400]] end
+  let :matrix_second do [[1,2,3,4],[2,3,4,5]] end
+  let :pwm_first do Bioinform::PWM.new(matrix_first) end
+  let :pwm_second do Bioinform::PWM.new(matrix_second) end
+  
+  context '#count_distribution_after_threshold' do
+    
+    it 'should return hash of score => count for all scores >= threshold' do
+      distribution_first = pwm_first.count_distribution_after_threshold(0)
+      distribution_first.keys.should == Array.product(*matrix_first).map{|score_row| score_row.inject(&:+)}
+      distribution_first.values.uniq.should == [1]
+      
+      distribution_second = pwm_second.count_distribution_after_threshold(0)
+      distribution_second.should == { 3=>1, 4=>2, 5=>3, 6=>4, 7=>3, 8=>2, 9=>1 }
+      
+      distribution_second = pwm_second.count_distribution_after_threshold(5)
+      distribution_second.should == { 5=>3, 6=>4, 7=>3, 8=>2, 9=>1 }
+    end
+    
+    it 'should use existing precalculated hash @count_distribution if it exists' do
+      pwm = pwm_second;
+      pwm.instance_variable_set :@count_distribution, { 3=>10, 4=>20, 5=>30, 6=>40, 7=>30, 8=>20, 9=>10 }
+      
+      distribution_second = pwm.count_distribution_after_threshold(0)
+      distribution_second.should == { 3=>10, 4=>20, 5=>30, 6=>40, 7=>30, 8=>20, 9=>10 }
+      
+      distribution_second = pwm.count_distribution_after_threshold(5)
+      distribution_second.should == { 5=>30, 6=>40, 7=>30, 8=>20, 9=>10 }
+    end
+  end
+  
+  context '#count_distribution' do
+    it 'should return hash of score => count for all available scores' do
+      pwm_second.count_distribution.should == { 3=>1, 4=>2, 5=>3, 6=>4, 7=>3, 8=>2, 9=>1 }
+    end
+    
+    it 'should cache calculation in @count_distribution' do
+      pwm = pwm_second;
+      pwm.instance_variable_set :@count_distribution, { 3=>10, 4=>20, 5=>30, 6=>40, 7=>30, 8=>20, 9=>10 }
+      pwm.count_distribution.should == { 3=>10, 4=>20, 5=>30, 6=>40, 7=>30, 8=>20, 9=>10 }
+      
+      pwm.instance_variable_set :@count_distribution, nil
+      pwm.count_distribution.should == { 3=>1, 4=>2, 5=>3, 6=>4, 7=>3, 8=>2, 9=>1 }
+    end
+  end
+  
+end

data/spec/spec_helper.rb

@@ -0,0 +1,4 @@
+$lib_folder = File.dirname(__FILE__) + '/../lib'
+$LOAD_PATH.unshift $lib_folder
+
+require 'rspec'

data/test/eval_alignment_similarity_test.rb

@@ -1,5 +1,6 @@
 require 'test_helper'
 
+puts "\n\neval_alignment test:"
 class TestEvalAlignmentSimilarity < Test::Unit::TestCase
   def test_process_at_optimal_alignment
     IO.popen(Helpers.exec_cmd('eval_alignment','test/data/KLF4_f2.pat test/data/SP1_f1.pat -1 direct')){|f|

data/test/eval_similarity_test.rb

@@ -1,5 +1,6 @@
 require 'test_helper'
 
+puts "\n\neval_similarity test:"
 class TestEvalSimilarity < Test::Unit::TestCase
   def test_process_pair_of_pwms
     IO.popen(Helpers.exec_cmd('eval_similarity','test/data/KLF4_f2.pat test/data/SP1_f1.pat')){|f|

data/test/find_pvalue_test.rb

@@ -1,5 +1,6 @@
 require 'test_helper'
 
+puts "\n\nfind_pvalue test:"
 class FindPvalueTest < Test::Unit::TestCase
   def test_process_one_threshold
     IO.popen(Helpers.exec_cmd('find_pvalue', 'test/data/KLF4_f2.pat 4.1719')){|f|

data/test/find_threshold_test.rb

@@ -1,5 +1,6 @@
 require 'test_helper'
 
+puts "\n\nfind_threshold test:"
 class FindThresholdTest < Test::Unit::TestCase
   def test_process_several_pvalues
     pvalues = []

data/test/preprocess_collection_test.rb

@@ -2,6 +2,7 @@ require 'test_helper'
 require 'yaml'
 require 'macroape'
 
+puts "\n\npreprocess_collection test:"
 class TestPreprocessCollection < Test::Unit::TestCase
   def test_multipvalue_preproceessing
     system(Helpers.exec_cmd('preprocess_collection','test/data/test_collection -o test/data/test_collection.yaml.tmp -p 0.0005 0.0001 0.00005 --silent'))

data/test/scan_collection_test.rb

@@ -1,5 +1,6 @@
 require 'test_helper'
 
+puts "\n\nscan_collection test:"
 class TestScanCollection < Test::Unit::TestCase
   def test_scan_default_cutoff
     assert_equal File.read('test/data/KLF4_f2_scan_results_default_cutoff.txt').gsub("\r\n", "\n"),

data/test/test_helper.rb

@@ -1,12 +1,12 @@
-$LOAD_PATH.unshift File.dirname(__FILE__) + '/../lib'
-$LOAD_PATH.unshift File.dirname(__FILE__)
+$lib_folder = File.dirname(__FILE__) + '/../lib'
+$LOAD_PATH.unshift $lib_folder
 require 'test/unit'
  
 module Helpers
   def self.obtain_pvalue_by_threshold(args)
-    IO.popen("find_pvalue #{args}",&:read).strip.split.last
+    IO.popen("ruby -I #{$lib_folder} #{$lib_folder}/macroape/exec/find_pvalue.rb #{args}",&:read).strip.split.last
   end
   def self.exec_cmd(executable, param_list)
-    "ruby #{File.dirname(File.absolute_path __FILE__)}/../lib/macroape/exec/#{executable}.rb #{param_list}"
+    "ruby -I #{$lib_folder} #{$lib_folder}/macroape/exec/#{executable}.rb #{param_list}"
   end
 end

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification
 name: macroape
 version: !ruby/object:Gem::Version
-  version: 3.3.2
+  version: 3.3.3
   prerelease: 
 platform: ruby
 authors:
@@ -9,7 +9,7 @@ authors:
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2012-07-07 00:00:00.000000000 Z
+date: 2012-07-19 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: bioinform
@@ -48,7 +48,9 @@ files:
 - Gemfile
 - LICENSE
 - README.md
-- Rakefile
+- Rakefile.rb
+- TODO.txt
+- benchmark/similarity_benchmark.rb
 - bin/eval_alignment
 - bin/eval_similarity
 - bin/find_pvalue
@@ -58,7 +60,7 @@ files:
 - lib/macroape.rb
 - lib/macroape/aligned_pair_intersection.rb
 - lib/macroape/collection.rb
-- lib/macroape/count_by_threshold.rb
+- lib/macroape/counting.rb
 - lib/macroape/exec/eval_alignment.rb
 - lib/macroape/exec/eval_similarity.rb
 - lib/macroape/exec/find_pvalue.rb
@@ -67,9 +69,10 @@ files:
 - lib/macroape/exec/scan_collection.rb
 - lib/macroape/pwm_compare.rb
 - lib/macroape/pwm_compare_aligned.rb
-- lib/macroape/threshold_by_pvalue.rb
 - lib/macroape/version.rb
 - macroape.gemspec
+- spec/count_distribution_spec.rb
+- spec/spec_helper.rb
 - test/data/AHR_si.pat
 - test/data/KLF4_f2.pat
 - test/data/KLF4_f2_scan_results_all.txt
@@ -113,6 +116,8 @@ signing_key:
 specification_version: 3
 summary: PWM comparison tool using MACROAPE approach
 test_files:
+- spec/count_distribution_spec.rb
+- spec/spec_helper.rb
 - test/data/AHR_si.pat
 - test/data/KLF4_f2.pat
 - test/data/KLF4_f2_scan_results_all.txt

data/Rakefile

@@ -1,28 +0,0 @@
-#!/usr/bin/env rake
-require "bundler/gem_tasks"
-
-namespace :spec do
-  task :find_threshold do
-    system("ruby -I ./test test/find_threshold_test.rb")
-  end
-  task :find_pvalue do
-    system("ruby -I ./test test/find_pvalue_test.rb")
-  end
-  task :eval_similarity do
-    system("ruby -I ./test test/eval_similarity_test.rb")
-  end
-  task :eval_alignment_similarity do
-    system("ruby -I ./test test/eval_alignment_similarity_test.rb")
-  end
-  task :preprocess_collection do
-    system("ruby -I ./test test/preprocess_collection_test.rb")
-  end
-  task :scan_collection do
-    system("ruby -I ./test test/scan_collection_test.rb")
-  end
-  task :all => [:find_threshold, :find_pvalue, :eval_similarity,
-                :eval_alignment_similarity, :scan_collection, :preprocess_collection]
-end
-
-desc 'Test all functionality of gem executables'
-task :spec => ['spec:all']

data/lib/macroape/count_by_threshold.rb

@@ -1,16 +0,0 @@
-require 'macroape/threshold_by_pvalue'
-
-module Bioinform
-  class PWM
-    def counts_by_thresholds(*thresholds)
-      scores = count_distribution_after_threshold(thresholds.min)
-      thresholds.map{ |threshold|
-        scores.inject(0.0){|sum,(score,count)|  (score >= threshold) ? sum + count : sum}
-      }
-    end
-
-    def pvalue_by_threshold(threshold)
-      counts_by_thresholds(threshold).first / vocabulary_volume
-    end
-  end
-end