mdarray 0.4.3.pre-java → 0.5.0-java

data/test/{test_boolean.rb → mdarray/test_boolean.rb}

@@ -4,8 +4,6 @@ require 'shoulda'
 
 require 'mdarray'
 
-require_relative 'env'
-
 class MDArrayTest < Test::Unit::TestCase
 
   context "Arithmetic Tests" do

data/test/{test_complete.rb → mdarray/test_complete.rb}

@@ -19,16 +19,8 @@
 # OR MODIFICATIONS.
 ##########################################################################################
 
-require 'simplecov'
-
-# require_relative 'env'
-
-#=begin
-SimpleCov.start do
-  @filters = []
-  add_group "MDArray", "mdarray-0.4.2-java/lib"
-end
-#=end
+# require 'simplecov'
+# require '../env.rb'
 
 # MDArray main object is the homogeneous multidimensional array. It is a table
 # of elements (usually numbers), all of the same type, indexed by a tuple of 
@@ -68,8 +60,5 @@ require_relative 'test_counter'
 require_relative 'test_trigonometry'
 require_relative 'test_views'
 require_relative 'test_printing'
-require_relative 'test_statistics'
+#require_relative 'test_performance'
 #=end
-
-# require_relative 'test_speed'
-

data/test/{test_error.rb → mdarray/test_error.rb}

@@ -40,11 +40,14 @@ class MDArrayTest < Test::Unit::TestCase
 
     should "correct error!" do
 
-      # creates an array from a function (actually a block).  The name fromfunction
-      # is preserved to maintain API compatibility with NumPy (is it necessary?)
-      arr = MDArray.fromfunction("double", [5, 5, 5, 5, 5, 5, 5, 5]) do |x|
-        x.inject(:+)
-      end
+      arr = MDArray.double("VALE3_short.csv", true)
+      arr.print
+
+      arr = MDArray.float("VALE3_short.csv", true)
+      arr.print
+
+      arr = MDArray.int("VALE3_short.csv", true)
+      arr.print
 
     end
     

data/test/{test_operator.rb → mdarray/test_operator.rb}

@@ -76,6 +76,7 @@ class MDArrayTest < Test::Unit::TestCase
       # c = [20 31 42 53]
       # chage the way operators work by changing the binary_operator parameter.
       # @a.binary_operator = BinaryOperator
+
       c = @a + @b
       assert_equal(20, c[0])
 
@@ -122,6 +123,7 @@ class MDArrayTest < Test::Unit::TestCase
 
     end
 
+#=begin
     #-------------------------------------------------------------------------------------
     # Array operations are done elementwise
     #-------------------------------------------------------------------------------------
@@ -261,7 +263,6 @@ class MDArrayTest < Test::Unit::TestCase
     #------------------------------------------------------------------------------------
 
     should "allow users operator's creation" do
-      
 
       # creating a binary_op as it takes two MDArrays as arguments.  The method "total_sum"
       # will make a cumulative sum of all sums of array1[i] + array2[i].  To achieve this
@@ -345,6 +346,9 @@ class MDArrayTest < Test::Unit::TestCase
 
     end
     
+#=end
+    
   end
-  
+
+
 end

data/test/{test_speed.rb → mdarray/test_performance.rb}

@@ -3,7 +3,6 @@ require "test/unit"
 require 'shoulda'
 require 'jruby/profiler'
 
-require 'env'
 require 'benchmark'
 
 require 'mdarray'
@@ -15,52 +14,89 @@ class MDArrayTest < Test::Unit::TestCase
 #=begin
     setup do
       
-      p "starting speed test"
+      p "starting performance testing"
       
     end
 
     #-------------------------------------------------------------------------------------
     #
     #-------------------------------------------------------------------------------------
-
-    should "test addition" do
+=begin
+    should "test basic operations" do
 
       @a = MDArray.typed_arange("double", 10_000_000)
       @b = MDArray.typed_arange("double", 10_000_000)
-
       @c = MDArray.arange(10_000_000)
       @d = MDArray.arange(10_000_000)
 
-      puts Benchmark.measure {
-        @a + @b
-        p @a[10]
-      }
+      binary_operators = [RubyBinaryOperator, FastBinaryOperator]
+      unary_operators = [RubyUnaryOperator, FastUnaryOperator]
 
-      puts Benchmark.measure {
-        @a - @b
-      }
+      puts "FastBinaryOperator and FastUnaryOperator are the default operator for MDArray" 
+      puts "and those are the expected execution times. "
+      puts "FastBinaryOperator sends loops to Java for performance improvement."
+      puts "RubyBinaryOperator and RubyUnaryOperator are left in the code for testing purposes"
+      puts "only and to compare how much improvement can be obtained."
+      puts "In the future, if Ruby becomes more efficient there might be no reason to"
+      puts "implement loops in Java."
+      puts
+      puts "Note that huge improvement (fast_add) can be obtained by implementing operator "
+      puts "methods completely in Java."
 
-      puts Benchmark.measure {
-        @a + @b
-        p @a[10]
-      }
+      binary_operators.each do |operator|
 
-      puts Benchmark.measure {
-        @a.plus(@b)
-        p @a[10]
-      }
+        MDArray.binary_operator = operator
 
-      puts Benchmark.measure {
-        @c.plus(@d)
-      }
+        puts "========================="
+        puts "Benchmarking: #{operator}"
 
-      puts Benchmark.measure {
-        @a.fast_add(@b)
-        p @a[10]
-      }
+        puts "Benchmarking double operations"
+        puts "Benchmarking addition of two large arrays of size 10_000_000"
+        puts Benchmark.measure {
+          c = @a + @b
+        }
+        
+        puts "Benchmarking subtraction of two large arrays of size 10_000_000"
+        puts Benchmark.measure {
+          c = @a - @b
+        }
+        
+        puts "Benchmarking addition of two large arrays of size 10_000_000"
+        puts "Checking if compiler still does improvements after 20_000_000 operations"
+        puts Benchmark.measure {
+          c = @a + @b
+        }
+        
+        puts "========================="
+        puts "Benchmarking fast_add.  fast_add is a Java method to add the two double arrays."
+        puts "Notice that performance can be greatly improved by implementing methods in Java"
+        puts Benchmark.measure {
+          c = @a.fast_add(@b)
+        }
+        puts "========================="
+        
+        puts "Benchmarking int operations"
+        puts "In my machine, there is very little performance difference between double and int"
+        puts "Benchmarking addition of two large arrays of size 10_000_000"
+        puts Benchmark.measure {
+          c = @c + @d
+        }
+        
+        puts "Benchmarking subtraction of two large arrays of size 10_000_000"
+        puts Benchmark.measure {
+          c = @c - @d
+        }
+        
+        puts "Benchmarking addition of two large arrays of size 10_000_000"
+        puts "Checking if compiler still does improvements after 20_000_000 operations"
+        puts Benchmark.measure {
+          c = @c + @d
+        }
 
-    end
+      end
 
+    end
+=end
 
 =begin
     should "test inner product" do
@@ -68,17 +104,16 @@ class MDArrayTest < Test::Unit::TestCase
       @a = MDArray.typed_arange("double", 10_000_000)
       @b = MDArray.typed_arange("double", 10_000_000)
 
-      MDArray.make_binary_op("inner_product", "reduce", 
-                             Proc.new { |sum, val1, val2| sum + (val1 * val2) }, 
-                             nil, 
-                             Proc.new { 0 })
+      # define inner_product for the whole hierarchy as it defines it for double
+      UserFunction.binary_operator("inner_product", "reduce", 
+                                   Proc.new { |sum, val1, val2| sum + (val1 * val2) }, 
+                                   "double", nil, Proc.new { 0 })
 
-      puts "measuring inner_product of two arrays of size: 10.000.000 in seconds"
-      @a.binary_operator = BinaryOperator
+      puts "Measuring inner_product of two arrays of size: 10.000.000 in seconds"
+      puts "Since inner product is more complex than a simple addition or subtraction there"
+      puts "should be a performance impact.  However, this should not be too large"
       puts Benchmark.measure {
-        p @a.inner_product(@b)
-        # p (@a * @b).sum
-        
+        @a.inner_product(@b)
       }
       
     end
@@ -88,7 +123,7 @@ class MDArrayTest < Test::Unit::TestCase
     #
     #-------------------------------------------------------------------------------------
  
-=begin
+#=begin
    should "execute fromfunction" do
 
       # creates an array from a function (actually a block).  The name fromfunction
@@ -107,7 +142,7 @@ class MDArrayTest < Test::Unit::TestCase
       }
 
     end
-=end
+#=end
 
 =begin
       @a = MDArray.double([7, 500, 20, 320])

data/test/{test_printing.rb → mdarray/test_printing.rb}

@@ -2,7 +2,6 @@ require 'rubygems'
 require "test/unit"
 require 'shoulda'
 
-require 'env'
 require 'mdarray'
 
 class MDArrayTest < Test::Unit::TestCase

data/test/mdarray/test_statistics.rb

@@ -0,0 +1,80 @@
+require 'rubygems'
+require "test/unit"
+require 'shoulda'
+
+require 'mdarray'
+
+class MDArrayTest < Test::Unit::TestCase
+
+  context "Statistics Tests" do
+
+    setup do
+
+      # create a byte array filled with 0's
+      @a = MDArray.typed_arange("double", 10_000)
+      @weight = MDArray.arange(10_000)
+
+      # create double array
+      @b = MDArray.double([2, 3, 4])
+
+    end
+
+    #-------------------------------------------------------------------------------------
+    #
+    #-------------------------------------------------------------------------------------
+
+    should "do stats operations" do
+
+      # read file VALE3.  This file has a header that we need to discard.  VALE3 
+      # contains quote values from brazilian company VALE as obtained from Yahoo finance
+      # (quote vale3.SA)
+      vale3 = MDArray.double("VALE3_short.csv", true)
+
+      # in order to use statistics we need to reset_statistics on array vale3.  This 
+      # breaks version 0.4.3 statistics that did not require a call to reset_statistics.
+      vale3.reset_statistics
+
+      # sum all values of vale3.  This does not make sense from a financial point of view.
+      # suming all values including dates... Not doing anything with the value, just
+      # checking that it works!
+      vale3.sum
+
+      # lets get only the open price for the whole period. We slice the vale3 on the
+      # second dimension and get the second column. 
+      open = vale3.slice(1,1)
+      
+      # getting descriptive statistics for the open value.  open is a new MDArray, so
+      # we need to reset_statistics for open as well.
+      open.reset_statistics
+
+      p "Durbin Watson: #{open.durbin_watson}"
+      p "Geometric mean: #{open.geometric_mean}"
+      p "Kurtosis: #{open.kurtosis}"
+      p "Lag1: #{open.lag1}"
+      p "Max: #{open.max}"
+      p "Mean: #{open.mean}"
+      p "Mean deviation: #{open.mean_deviation}"
+      p "Median: #{open.median}"
+      p "Min: #{open.min}"
+      p "Moment3: #{open.moment3}"
+      p "Moment4: #{open.moment4}"
+      p "Product: #{open.product}"
+      p "Skew: #{open.skew}"
+      p "Standard deviation: #{open.standard_deviation}"
+      p "Standard error: #{open.standard_error}"
+      p "Variance: #{open.variance}"
+
+    end
+
+  end
+
+end
+
+=begin
+      assert_equal(49995000, @a.sum)
+      assert_equal(0, @a.min)
+      assert_equal(9999, @a.max)
+      assert_equal(4999.5, @a.mean)
+      assert_equal(6666.333333333333, @a.weighted_mean(@weight))
+=end
+

data/version.rb

@@ -1,2 +1,2 @@
 $gem_name = "mdarray"
-$version="0.4.3.pre"
+$version="0.5.0"

metadata

@@ -1,15 +1,15 @@
 --- !ruby/object:Gem::Specification
 name: mdarray
 version: !ruby/object:Gem::Version
-  prerelease: 6
-  version: 0.4.3.pre
+  prerelease:
+  version: 0.5.0
 platform: java
 authors:
 - Rodrigo Botafogo
 autorequire:
 bindir: bin
 cert_chain: []
-date: 2013-04-19 00:00:00.000000000 Z
+date: 2013-05-16 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: map
@@ -27,6 +27,24 @@ dependencies:
     none: false
   prerelease: false
   type: :runtime
+- !ruby/object:Gem::Dependency
+  name: shoulda
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: !binary |-
+          MA==
+    none: false
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: !binary |-
+          MA==
+    none: false
+  prerelease: false
+  type: :runtime
 - !ruby/object:Gem::Dependency
   name: simplecov
   version_requirements: !ruby/object:Gem::Requirement
@@ -75,7 +93,7 @@ dependencies:
     none: false
   prerelease: false
   type: :development
-description: "\"Multi dimensional array similar to Masahiro Tanaka's narray and numpy.\
+description: "\"Multi dimensional array similar to Masahiro Tanaka's Narray and NumPy.\
   \  \nIt is specifically targeted to JRuby as it uses Java-NetCDF library as base\
   \ Array.\"\n"
 email: rodrigo.a.botafogo@gmail.com
@@ -87,9 +105,16 @@ files:
 - version.rb
 - lib/env.rb
 - lib/mdarray.rb
+- lib/colt/colt.rb
+- lib/colt/colt_mdarray.rb
+- lib/colt/double_descriptive.rb
+- lib/colt/probability.rb
+- lib/colt/stat_list.rb
 - lib/mdarray/access.rb
 - lib/mdarray/counter.rb
 - lib/mdarray/creation.rb
+- lib/mdarray/csv.rb
+- lib/mdarray/fast_operators.rb
 - lib/mdarray/function_creation.rb
 - lib/mdarray/function_map.rb
 - lib/mdarray/hierarchy.rb
@@ -104,24 +129,97 @@ files:
 - lib/mdarray/ruby_operators.rb
 - lib/mdarray/ruby_stats.rb
 - lib/mdarray/views.rb
-- test/arithmetic_casting.rb
+- test/complete.rb
 - test/env.rb
-- test/test_access.rb
-- test/test_boolean.rb
-- test/test_comparison.rb
-- test/test_complete.rb
-- test/test_counter.rb
-- test/test_creation.rb
-- test/test_error.rb
-- test/test_lazy.rb
-- test/test_operator.rb
-- test/test_printing.rb
-- test/test_shape.rb
-- test/test_speed.rb
-- test/test_statistics.rb
-- test/test_trigonometry.rb
-- test/test_views.rb
+- test/colt/test_complete.rb
+- test/colt/test_statistics.rb
+- test/colt/test_stat_list.rb
+- test/mdarray/arithmetic_casting.rb
+- test/mdarray/test_access.rb
+- test/mdarray/test_boolean.rb
+- test/mdarray/test_comparison.rb
+- test/mdarray/test_complete.rb
+- test/mdarray/test_counter.rb
+- test/mdarray/test_creation.rb
+- test/mdarray/test_error.rb
+- test/mdarray/test_operator.rb
+- test/mdarray/test_performance.rb
+- test/mdarray/test_printing.rb
+- test/mdarray/test_shape.rb
+- test/mdarray/test_statistics.rb
+- test/mdarray/test_trigonometry.rb
+- test/mdarray/test_views.rb
+- test/colt/VALE3.csv
+- test/colt/VALE3_short-err.csv
+- test/colt/VALE3_short.csv
+- test/colt/VALE3.xlsx
+- test/colt/VALE3_short.xlsx
+- doc/BinaryOperator.html
+- doc/BitwiseOperators.html
+- doc/BooleanFunctions.html
+- doc/BooleanMDArray.html
+- doc/ByteMDArray.html
+- doc/class_list.html
+- doc/Colt.html
+- doc/ComparisonOperators.html
+- doc/Const.html
+- doc/Csv.html
+- doc/DDescriptive.html
+- doc/DoubleMDArray.html
+- doc/DoubleStatList.html
+- doc/FastBinaryOperator.html
+- doc/FastBoolean.html
+- doc/FastUnaryOperator.html
+- doc/file.README.html
+- doc/file_list.html
+- doc/FloatMDArray.html
+- doc/frames.html
+- doc/FunctionCreation.html
+- doc/FunctionMap.html
+- doc/GenericFunctions.html
+- doc/index.html
+- doc/IntMDArray.html
+- doc/LongMDArray.html
+- doc/MDArray.html
+- doc/MDArrayTest.html
+- doc/method_list.html
+- doc/NonNumericalMDArray.html
+- doc/NumericalMDArray.html
+- doc/NumericFunctions.html
+- doc/Operator.html
+- doc/Proc.html
+- doc/RubyBinaryOperator.html
+- doc/RubyFunctions.html
+- doc/RubyMath.html
+- doc/RubyStats.html
+- doc/RubyUnaryOperator.html
+- doc/ShortMDArray.html
+- doc/StatList.html
+- doc/StringMDArray.html
+- doc/StructureMDArray.html
+- doc/top-level-namespace.html
+- doc/UnaryOperator.html
+- doc/UserFunction.html
+- doc/_index.html
+- doc/css/common.css
+- doc/css/full_list.css
+- doc/css/style.css
+- doc/js/app.js
+- doc/js/full_list.js
+- doc/js/jquery.js
+- doc/MDArray/Counter.html
+- doc/MDArray/IteratorFast.html
+- doc/MDArray/IteratorFastBoolean.html
+- doc/MDArray/IteratorFastByte.html
+- doc/MDArray/IteratorFastChar.html
+- doc/MDArray/IteratorFastDouble.html
+- doc/MDArray/IteratorFastFloat.html
+- doc/MDArray/IteratorFastInt.html
+- doc/MDArray/IteratorFastLong.html
+- doc/MDArray/IteratorFastShort.html
 - vendor/netcdfAll-4.3.16.jar
+- vendor/parallelcolt-0.10.0.jar
+- target/helper.jar
 - README.md
 - LICENSE.txt
 - LICENSE.txt~
@@ -140,10 +238,10 @@ required_ruby_version: !ruby/object:Gem::Requirement
   none: false
 required_rubygems_version: !ruby/object:Gem::Requirement
   requirements:
-  - - !binary |-
-      Pg==
+  - - ">="
     - !ruby/object:Gem::Version
-      version: 1.3.1
+      version: !binary |-
+        MA==
   none: false
 requirements: []
 rubyforge_project:
@@ -152,21 +250,6 @@ signing_key:
 specification_version: 3
 summary: Multi dimensional array similar to narray and numpy.
 test_files:
-- test/arithmetic_casting.rb
+- test/complete.rb
 - test/env.rb
-- test/test_access.rb
-- test/test_boolean.rb
-- test/test_comparison.rb
-- test/test_complete.rb
-- test/test_counter.rb
-- test/test_creation.rb
-- test/test_error.rb
-- test/test_lazy.rb
-- test/test_operator.rb
-- test/test_printing.rb
-- test/test_shape.rb
-- test/test_speed.rb
-- test/test_statistics.rb
-- test/test_trigonometry.rb
-- test/test_views.rb
 has_rdoc: