daru 0.0.1 → 0.0.2

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA1:
-  metadata.gz: 1a2952b88ba0b72a6d7ada74bd03b03b7690f774
-  data.tar.gz: fdace87c3406adda7f141aaab055fb1e40c66e0c
+  metadata.gz: e40b5dd9b236c4f7bbc3e96e0d97f73fe6d940b4
+  data.tar.gz: 377d3ddb6826c72179e84eb2f9cf66cc120b1236
 SHA512:
-  metadata.gz: be50d9aafa90638208605de2bd68c121df1ce72eebe463ed3ec06f548104b5f86f73ea572b7424e03fca518f21f3425dc05608ace18cd3d430111c61b526ce3a
-  data.tar.gz: 10c7243159bad7188f99095308035dc5064bea91d4ba966264f6d84a4b8dfac86c096f126df7c149da331770efdc3c0e8b2a54b48d6bd121ff046d0c9b5d740a
+  metadata.gz: 8e42d5fb359945d2ccab1f19062c1cc2e5123e4a33af54d0aeeda6db966d4d53444f50a40e8af2fbdffdf42e3416dcddf3421815b5e787567e73032365a59f62
+  data.tar.gz: 089564afd6be1eef4f01060b064273ea2ca5efb2dd60d2b63019d2f324bb560e814e1dde855838e54b28cf04b55be8fc02a821bf48adc7e8e65830005618e900

data/Gemfile.lock

@@ -32,4 +32,4 @@ DEPENDENCIES
   bundler
   daru!
   nmatrix (~> 0.1.0.rc5)
-  rspec (~> 3.0)
+  rspec

data/README.md

@@ -2,3 +2,35 @@ daru
 ====
 
 Data Analysis in RUby
+
+## Introduction
+
+daru (Data Analysis in RUby) is a library for storage, analysis and manipulation of data. It aims to be the preferred data analysis library for Ruby. 
+
+Development of daru was started to address the fragmentation of Dataframe-like classes which were created in many ruby gems as per their own needs. 
+
+This creates a hurdle in using these gems together to solve a problem. For example, calculating something in [statsample](https://github.com/clbustos/statsample) and plotting the results in [Nyaplot](https://github.com/domitry/nyaplot).
+
+daru is heavily inspired by `Statsample::Dataset`, `Nyaplot::DataFrame` and the super-awesome pandas, a very mature solution in Python.
+
+## Data Structures
+
+daru employs several data structures for storing and manipulating data:
+* Vector - A basic 1-D vector.
+* DataFrame - A 2-D matrix-like structure which is internally composed of named `Vector` classes.
+
+daru data structures can be constructed by using several Ruby classes. These include `Array`, `Hash`, `Matrix`, [NMatrix](https://github.com/SciRuby/nmatrix) and [MDArray](https://github.com/rbotafogo/mdarray). daru brings a uniform API for handling and manipulating data represented in any of the above Ruby classes.
+
+## Testing
+
+Install jruby using `rvm install jruby`, then run `jruby -S gem install mdarray`, followed by `bundle install`. You will need to install `mdarray` manually because of strange gemspec file behaviour. If anyone can automate this then I'd greatly appreciate it! Then run `rspec` in JRuby to test for MDArray functionality.
+
+Then switch to MRI, do a normal `bundle install` followed by `rspec` for testing everything else with NMatrix functionality.
+
+## Roadmap
+
+* Automate testing for both MRI and JRuby.
+* Enable creation of DataFrame by only specifying an NMatrix/MDArray in initialize. Vector naming happens automatically (alphabetic) or is specified in an Array.
+* Add support for missing values in vectors.
+* Add normal and destructive map iterators.
+* Completely test all functionality for NMatrix and MDArray.

data/daru.gemspec

@@ -5,8 +5,7 @@ require 'version.rb'
 
 DESCRIPTION = <<MSG
 Daru (Data Analysis in RUby) is a library for storage, analysis and manipulation
-of data. It aims to be the preferred data storage library for all Ruby gems, promoting 
-interoperability and easy of use between various ruby packages.
+of data.
 MSG
 
 Gem::Specification.new do |spec|
@@ -25,6 +24,8 @@ Gem::Specification.new do |spec|
   spec.require_paths = ["lib"]
 
   spec.add_development_dependency 'bundler'
-  spec.add_development_dependency 'rspec', '~> 3.0'
-  spec.add_development_dependency 'nmatrix', '~> 0.1.0.rc5'
+  spec.add_development_dependency 'rspec'
+  if RUBY_ENGINE != 'jruby'
+    spec.add_development_dependency 'nmatrix', '~> 0.1.0.rc5'
+  end
 end

data/lib/daru.rb

@@ -1,5 +1,6 @@
 require 'securerandom'
 require 'matrix'
+require 'csv'
 
 require 'daru/vector.rb'
 require 'daru/dataframe.rb'

data/lib/daru/dataframe.rb

@@ -1,7 +1,180 @@
 module Daru
   class DataFrame
-    def initialize source, name=SecureRandom.uuid
-      
+
+    attr_reader :vectors
+
+    attr_reader :fields
+
+    attr_reader :size
+
+    attr_reader :name
+
+    def initialize source, fields=[], name=SecureRandom.uuid
+      if source.empty?
+        @vectors = fields.inject({}){ |a,x| a[x]=Daru::Vector.new; a}
+      else
+        @vectors = source
+      end
+
+      @fields = fields.empty? ? source.keys.sort : fields
+      @name   = name
+ 
+      check_length
+      set_fields_order if @vectors.keys.sort != @fields.sort
+      set_vector_names
+    end
+
+    def self.from_csv file, opts={}
+      opts[:col_sep]           ||= ','
+      opts[:headers]           ||= true
+      opts[:converters]        ||= :numeric
+      opts[:header_converters] ||= :symbol
+
+      csv = CSV.open file, 'r', opts
+
+      yield csv if block_given?
+
+      first = true
+      df    = nil
+
+      csv.each do |row|
+        if first
+          df = Daru::DataFrame.new({}, csv.headers)
+          first = false
+        end
+
+        df.insert_row row
+      end
+
+      df
+    end
+
+    def column name
+      @vectors[name]
+    end
+
+    def delete name
+      @vectors.delete name
+      @fields.delete name
+    end
+
+    def [](name)
+      column name
+    end
+
+    def []=(name, vector)
+      insert_vector name, vector
+    end
+
+    def row index
+      raise Exception, "Expected index to be within bounds" if index > @size
+
+      row = []
+      self.each_column do |column|
+        row << column[index]
+      end
+
+      row
+    end
+
+    def has_vector? vector
+      !!@vectors[vector]
+    end
+
+    def each_row
+      0.upto(@size) do |index|
+        yield row(index)
+      end
+    end
+
+    def each_row_with_index
+      0.upto(@size) do |index|
+        yield row(index), index
+      end
+    end
+
+    def each_column
+      @vectors.values.each do |column|
+        yield column
+      end
+    end
+
+    def insert_vector name, vector
+      raise Exeception, "Expected vector size to be same as DataFrame\ 
+        size." if vector.size != self.size
+
+      @vectors.merge({name => vector})
+      @fields << name
+    end
+
+    def insert_row row
+      raise Exception, "Expected new row to same as the number of rows \ 
+        in the DataFrame" if row.size != @fields.size
+
+      @fields.each_with_index do |field, index|
+        @vectors[field] << row[index]
+      end
+    end
+
+    def to_html(threshold=15)
+      html = '<table>'
+
+      self.each_row_with_index do |row, index|
+        break if index > threshold and index <= @size
+        html += '<tr>'
+        row.each{ |val| html.concat('<td>' + val.to_s + '</td>') }
+        html += '</tr>'
+        if i == threshold
+          html += '<tr>'
+          row.size.times { html.concat('<td>...</td>') }
+          html += '</tr>'
+        end
+      end
+
+      html += '</table>'
+    end
+
+    def to_s
+      to_html
+    end
+    
+    def method_missing(name, *args)
+      if md = name.match(/(.+)\=/)
+        insert_vector name[/(.+)\=/].delete("="), args[0]
+      elsif self.has_vector? name
+        column name
+      else
+        super(name, *args)
+      end
+    end
+
+   private
+    def check_length
+      size = nil
+
+      @vectors.each_value do |vector|
+        if size.nil?
+          size = vector.size
+        elsif size != vector.size
+          raise Exception, "Expected all vectors to be of the same size. Vector \ 
+            #{vector.name} is of size #{vector.size} and another one of size #{size}"    
+        end
+      end
+
+      @size = size
+    end
+
+    def set_fields_order
+      @fields = @vectors.keys & @fields
+      @fields += @vecorts.keys.sort - @fields
+    end
+
+    # Writes names specified in the hash to the actual name of the vector.
+    # Will over-ride any previous name assigned to the vector.
+    def set_vector_names
+      @fields.each do |name|
+        @vectors[name].name = name
+      end
     end
   end
 end

data/lib/daru/vector.rb

@@ -1,3 +1,43 @@
+class Array
+  def daru_vector
+    Daru::Vector.new self
+  end
+
+  alias_method :dv, :daru_vector
+end
+
+class Range
+  def daru_vector
+    Daru::Vector.new self
+  end
+
+  alias_method :dv, :daru_vector
+end
+
+class Hash
+  def daru_vector
+    Daru::Vector.new self.values[0], self.keys[0]
+  end
+
+  alias_method :dv, :daru_vector
+end
+
+class NMatrix
+  def daru_vector
+    Daru::Vector.new self
+  end
+
+  alias_method :dv, :daru_vector
+end
+
+class MDArray
+  def daru_vector
+    Daru::Vector.new self
+  end
+
+  alias_method :dv, :daru_vector
+end
+
 module Daru
   class Vector
     include Enumerable
@@ -6,27 +46,78 @@ module Daru
       @vector.each(&block)
     end
 
-    attr_reader :name
-
-    attr_reader :size
+    attr_accessor :name
 
-    def initialize source, name=SecureRandom.uuid
-      @name = source.is_a?(Hash) ? source.keys[0] : name
+    attr_reader   :size
 
-      @vector = 
-      if source.is_a? Hash
-        source.values[0].is_a?(Range) ? source.values[0].to_a  : source.values[0]
-      elsif source.is_a? Range or source.is_a? Matrix
-        source.to_a.flatten
+    attr_reader   :vector
+ 
+    def initialize source=[], name=nil
+      if source.is_a?(Hash)
+        initialize source.values[0], source.keys[0]
       else
-        source
-      end
+        @name = name || SecureRandom.uuid
+
+        @vector = 
+        case source
+        when Range, Matrix
+          source.to_a.flatten
+        else
+          source
+        end
 
-      @size = @vector.size
+        @size = @vector.size
+      end
     end
 
     def [](index)
       @vector[index]
     end
+
+    def []=(index, value)
+      @vector[index] = value
+    end
+
+    def ==(other)
+      other.vector == @vector and other.name == @name
+    end
+
+    def <<(element)
+      @vector << element
+    end
+
+    def to_json
+      self.to_a.to_json
+    end
+
+    def to_a
+      @vector.to_a
+    end
+
+    def to_nmatrix
+      @vector.to_nm
+    end
+
+    alias_method :to_nm, :to_nmatrix
+
+    def first lim=1
+      lim == 1 ? @vector.first : @vector.first(lim)
+    end
+
+    def to_html threshold=15
+      html = '<table><tr><th>' + @name.to_s + '</th></tr>>'
+
+      @vector.to_a.each_with_index do |el,i|
+        next if threshold < i and i < @arr.length-1
+        content = i == threshold ? '...' : el.to_s
+        html.concat('<tr><td>' + content  + '</td></tr>')
+      end
+
+      html += '</table>'
+    end
+
+    def dup
+      Daru::Vector.new @vector.dup, @name.dup
+    end
   end
 end

data/lib/version.rb

@@ -1,3 +1,3 @@
 module Daru
-  VERSION = "0.0.1"
+  VERSION = "0.0.2"
 end

data/spec/fixtures/matrix_test.csv

@@ -0,0 +1,100 @@
+image_resolution  true_transform  mls
+6.55779 -0.2362347,0.6308649,0.7390552,0,0.6523478,-0.4607318,0.6018043,0,0.7201635,0.6242881,-0.3027024,4262.65,0,0,0,1  0
+2.14746 -0.1539447,-0.2832125,-0.9466212,0,-0.1585857,0.9527035,-0.2592422,0,0.9752699,0.1102116,-0.1915772,1278.86,0,0,0,1 0
+8.31104 0.3832846,0.8818775,-0.2745634,0,0.8566163,-0.4505714,-0.2513841,0,-0.3454004,-0.1388438,-0.9281277,5448.81,0,0,0,1 0
+3.47872 0.3832846,0.8818775,-0.2745634,0,0.8566163,-0.4505714,-0.2513841,0,-0.3454004,-0.1388438,-0.9281277,2179.52,0,0,0,1 0
+4.16725 -0.2362347,0.6308649,0.7390552,0,0.6523478,-0.4607318,0.6018043,0,0.7201635,0.6242881,-0.3027024,2645.34,0,0,0,1  0
+5.79983 -0.2362347,0.6308649,0.7390552,0,0.6523478,-0.4607318,0.6018043,0,0.7201635,0.6242881,-0.3027024,3749.86,0,0,0,1  0
+1.9058  -0.895577,-0.4178617,0.152753,0,-0.1004795,-0.1445008,-0.9843898,0,0.4334117,-0.8969454,0.08742505,1115.36,0,0,0,1  0
+1.9058  -0.2362347,0.6308649,0.7390552,0,0.6523478,-0.4607318,0.6018043,0,0.7201635,0.6242881,-0.3027024,1115.36,0,0,0,1  0
+4.11806 -0.895577,-0.4178617,0.152753,0,-0.1004795,-0.1445008,-0.9843898,0,0.4334117,-0.8969454,0.08742505,2612.06,0,0,0,1  0
+6.26622 -0.2362347,0.6308649,0.7390552,0,0.6523478,-0.4607318,0.6018043,0,0.7201635,0.6242881,-0.3027024,4065.39,0,0,0,1  0
+2.57805 -0.1539447,-0.2832125,-0.9466212,0,-0.1585857,0.9527035,-0.2592422,0,0.9752699,0.1102116,-0.1915772,1570.18,0,0,0,1 0
+4.76151 -0.2362347,0.6308649,0.7390552,0,0.6523478,-0.4607318,0.6018043,0,0.7201635,0.6242881,-0.3027024,3047.39,0,0,0,1  0
+7.11002 -0.895577,-0.4178617,0.152753,0,-0.1004795,-0.1445008,-0.9843898,0,0.4334117,-0.8969454,0.08742505,4636.26,0,0,0,1  0
+5.40811 -0.2362347,0.6308649,0.7390552,0,0.6523478,-0.4607318,0.6018043,0,0.7201635,0.6242881,-0.3027024,3484.84,0,0,0,1  0
+8.19567 -0.1539447,-0.2832125,-0.9466212,0,-0.1585857,0.9527035,-0.2592422,0,0.9752699,0.1102116,-0.1915772,5370.75,0,0,0,1 0
+3.47872 -0.895577,-0.4178617,0.152753,0,-0.1004795,-0.1445008,-0.9843898,0,0.4334117,-0.8969454,0.08742505,2179.52,0,0,0,1  0
+2.91762 -0.2362347,0.6308649,0.7390552,0,0.6523478,-0.4607318,0.6018043,0,0.7201635,0.6242881,-0.3027024,1799.91,0,0,0,1  0
+4.76151 -0.1539447,-0.2832125,-0.9466212,0,-0.1585857,0.9527035,-0.2592422,0,0.9752699,0.1102116,-0.1915772,3047.39,0,0,0,1 0
+8.19567 -0.2362347,0.6308649,0.7390552,0,0.6523478,-0.4607318,0.6018043,0,0.7201635,0.6242881,-0.3027024,5370.75,0,0,0,1  0
+2.91762 -0.1539447,-0.2832125,-0.9466212,0,-0.1585857,0.9527035,-0.2592422,0,0.9752699,0.1102116,-0.1915772,1799.91,0,0,0,1 0
+6.26622 0.3832846,0.8818775,-0.2745634,0,0.8566163,-0.4505714,-0.2513841,0,-0.3454004,-0.1388438,-0.9281277,4065.39,0,0,0,1 0
+4.16725 -0.1539447,-0.2832125,-0.9466212,0,-0.1585857,0.9527035,-0.2592422,0,0.9752699,0.1102116,-0.1915772,2645.34,0,0,0,1 0
+3.78516 -0.895577,-0.4178617,0.152753,0,-0.1004795,-0.1445008,-0.9843898,0,0.4334117,-0.8969454,0.08742505,2386.84,0,0,0,1  0
+3.78516 0.3832846,0.8818775,-0.2745634,0,0.8566163,-0.4505714,-0.2513841,0,-0.3454004,-0.1388438,-0.9281277,2386.84,0,0,0,1 0
+7.47603 -0.2362347,0.6308649,0.7390552,0,0.6523478,-0.4607318,0.6018043,0,0.7201635,0.6242881,-0.3027024,4883.89,0,0,0,1  0
+5.79983 0.3832846,0.8818775,-0.2745634,0,0.8566163,-0.4505714,-0.2513841,0,-0.3454004,-0.1388438,-0.9281277,3749.86,0,0,0,1 0
+2.57805 -0.2362347,0.6308649,0.7390552,0,0.6523478,-0.4607318,0.6018043,0,0.7201635,0.6242881,-0.3027024,1570.18,0,0,0,1  0
+8.31104 -0.1539447,-0.2832125,-0.9466212,0,-0.1585857,0.9527035,-0.2592422,0,0.9752699,0.1102116,-0.1915772,5448.81,0,0,0,1 0
+1.77097 -0.2362347,0.6308649,0.7390552,0,0.6523478,-0.4607318,0.6018043,0,0.7201635,0.6242881,-0.3027024,1024.14,0,0,0,1  0
+5.9596  0.3832846,0.8818775,-0.2745634,0,0.8566163,-0.4505714,-0.2513841,0,-0.3454004,-0.1388438,-0.9281277,3857.95,0,0,0,1 0
+5.40811 0.3832846,0.8818775,-0.2745634,0,0.8566163,-0.4505714,-0.2513841,0,-0.3454004,-0.1388438,-0.9281277,3484.84,0,0,0,1 0
+4.16725 0.3832846,0.8818775,-0.2745634,0,0.8566163,-0.4505714,-0.2513841,0,-0.3454004,-0.1388438,-0.9281277,2645.34,0,0,0,1 0
+7.11002 -0.1539447,-0.2832125,-0.9466212,0,-0.1585857,0.9527035,-0.2592422,0,0.9752699,0.1102116,-0.1915772,4636.26,0,0,0,1 0
+7.47603 -0.895577,-0.4178617,0.152753,0,-0.1004795,-0.1445008,-0.9843898,0,0.4334117,-0.8969454,0.08742505,4883.89,0,0,0,1  0
+13.6304 -0.2362347,0.6308649,0.7390552,0,0.6523478,-0.4607318,0.6018043,0,0.7201635,0.6242881,-0.3027024,9047.63,0,0,0,1  0
+5.9596  -0.895577,-0.4178617,0.152753,0,-0.1004795,-0.1445008,-0.9843898,0,0.4334117,-0.8969454,0.08742505,3857.95,0,0,0,1  0
+4.76151 0.3832846,0.8818775,-0.2745634,0,0.8566163,-0.4505714,-0.2513841,0,-0.3454004,-0.1388438,-0.9281277,3047.39,0,0,0,1 0
+5.40811 -0.1539447,-0.2832125,-0.9466212,0,-0.1585857,0.9527035,-0.2592422,0,0.9752699,0.1102116,-0.1915772,3484.84,0,0,0,1 0
+12.4066 -0.2362347,0.6308649,0.7390552,0,0.6523478,-0.4607318,0.6018043,0,0.7201635,0.6242881,-0.3027024,8219.63,0,0,0,1  0
+5.79983 -0.1539447,-0.2832125,-0.9466212,0,-0.1585857,0.9527035,-0.2592422,0,0.9752699,0.1102116,-0.1915772,3749.86,0,0,0,1 0
+7.11002 -0.2362347,0.6308649,0.7390552,0,0.6523478,-0.4607318,0.6018043,0,0.7201635,0.6242881,-0.3027024,4636.26,0,0,0,1  0
+7.11002 0.3832846,0.8818775,-0.2745634,0,0.8566163,-0.4505714,-0.2513841,0,-0.3454004,-0.1388438,-0.9281277,4636.26,0,0,0,1 0
+2.58728 -0.1539447,-0.2832125,-0.9466212,0,-0.1585857,0.9527035,-0.2592422,0,0.9752699,0.1102116,-0.1915772,1576.42,0,0,0,1 0
+12.4066 -0.895577,-0.4178617,0.152753,0,-0.1004795,-0.1445008,-0.9843898,0,0.4334117,-0.8969454,0.08742505,8219.63,0,0,0,1  0
+3.78516 -0.2362347,0.6308649,0.7390552,0,0.6523478,-0.4607318,0.6018043,0,0.7201635,0.6242881,-0.3027024,2386.84,0,0,0,1  0
+2.58728 0.3832846,0.8818775,-0.2745634,0,0.8566163,-0.4505714,-0.2513841,0,-0.3454004,-0.1388438,-0.9281277,1576.42,0,0,0,1 0
+5.03378 -0.1539447,-0.2832125,-0.9466212,0,-0.1585857,0.9527035,-0.2592422,0,0.9752699,0.1102116,-0.1915772,3231.59,0,0,0,1 0
+5.40811 -0.895577,-0.4178617,0.152753,0,-0.1004795,-0.1445008,-0.9843898,0,0.4334117,-0.8969454,0.08742505,3484.84,0,0,0,1  0
+6.26622 -0.895577,-0.4178617,0.152753,0,-0.1004795,-0.1445008,-0.9843898,0,0.4334117,-0.8969454,0.08742505,4065.39,0,0,0,1  0
+2.58728 -0.2362347,0.6308649,0.7390552,0,0.6523478,-0.4607318,0.6018043,0,0.7201635,0.6242881,-0.3027024,1576.42,0,0,0,1  0
+2.91762 0.3832846,0.8818775,-0.2745634,0,0.8566163,-0.4505714,-0.2513841,0,-0.3454004,-0.1388438,-0.9281277,1799.91,0,0,0,1 0
+5.9596  -0.1539447,-0.2832125,-0.9466212,0,-0.1585857,0.9527035,-0.2592422,0,0.9752699,0.1102116,-0.1915772,3857.95,0,0,0,1 0
+1.77097 0.3832846,0.8818775,-0.2745634,0,0.8566163,-0.4505714,-0.2513841,0,-0.3454004,-0.1388438,-0.9281277,1024.14,0,0,0,1 0
+7.47603 0.3832846,0.8818775,-0.2745634,0,0.8566163,-0.4505714,-0.2513841,0,-0.3454004,-0.1388438,-0.9281277,4883.89,0,0,0,1 0
+5.03378 -0.895577,-0.4178617,0.152753,0,-0.1004795,-0.1445008,-0.9843898,0,0.4334117,-0.8969454,0.08742505,3231.59,0,0,0,1  0
+1.81745 -0.2362347,0.6308649,0.7390552,0,0.6523478,-0.4607318,0.6018043,0,0.7201635,0.6242881,-0.3027024,1055.59,0,0,0,1  0
+1.77097 -0.895577,-0.4178617,0.152753,0,-0.1004795,-0.1445008,-0.9843898,0,0.4334117,-0.8969454,0.08742505,1024.14,0,0,0,1  0
+2.14746 -0.2362347,0.6308649,0.7390552,0,0.6523478,-0.4607318,0.6018043,0,0.7201635,0.6242881,-0.3027024,1278.86,0,0,0,1  0
+8.19567 0.3832846,0.8818775,-0.2745634,0,0.8566163,-0.4505714,-0.2513841,0,-0.3454004,-0.1388438,-0.9281277,5370.75,0,0,0,1 0
+2.91762 -0.895577,-0.4178617,0.152753,0,-0.1004795,-0.1445008,-0.9843898,0,0.4334117,-0.8969454,0.08742505,1799.91,0,0,0,1  0
+1.81745 -0.895577,-0.4178617,0.152753,0,-0.1004795,-0.1445008,-0.9843898,0,0.4334117,-0.8969454,0.08742505,1055.59,0,0,0,1  0
+8.31104 -0.895577,-0.4178617,0.152753,0,-0.1004795,-0.1445008,-0.9843898,0,0.4334117,-0.8969454,0.08742505,5448.81,0,0,0,1  0
+1.81745 0.3832846,0.8818775,-0.2745634,0,0.8566163,-0.4505714,-0.2513841,0,-0.3454004,-0.1388438,-0.9281277,1055.59,0,0,0,1 0
+1.77097 -0.1539447,-0.2832125,-0.9466212,0,-0.1585857,0.9527035,-0.2592422,0,0.9752699,0.1102116,-0.1915772,1024.14,0,0,0,1 0
+2.57805 0.3832846,0.8818775,-0.2745634,0,0.8566163,-0.4505714,-0.2513841,0,-0.3454004,-0.1388438,-0.9281277,1570.18,0,0,0,1 0
+5.03378 -0.2362347,0.6308649,0.7390552,0,0.6523478,-0.4607318,0.6018043,0,0.7201635,0.6242881,-0.3027024,3231.59,0,0,0,1  0
+6.55779 0.3832846,0.8818775,-0.2745634,0,0.8566163,-0.4505714,-0.2513841,0,-0.3454004,-0.1388438,-0.9281277,4262.65,0,0,0,1 0
+6.55779 -0.895577,-0.4178617,0.152753,0,-0.1004795,-0.1445008,-0.9843898,0,0.4334117,-0.8969454,0.08742505,4262.65,0,0,0,1  0
+3.47872 -0.2362347,0.6308649,0.7390552,0,0.6523478,-0.4607318,0.6018043,0,0.7201635,0.6242881,-0.3027024,2179.52,0,0,0,1  0
+4.11806 -0.2362347,0.6308649,0.7390552,0,0.6523478,-0.4607318,0.6018043,0,0.7201635,0.6242881,-0.3027024,2612.06,0,0,0,1  0
+6.55779 -0.1539447,-0.2832125,-0.9466212,0,-0.1585857,0.9527035,-0.2592422,0,0.9752699,0.1102116,-0.1915772,4262.65,0,0,0,1 0
+8.31104 -0.2362347,0.6308649,0.7390552,0,0.6523478,-0.4607318,0.6018043,0,0.7201635,0.6242881,-0.3027024,5448.81,0,0,0,1  0
+2.14746 0.3832846,0.8818775,-0.2745634,0,0.8566163,-0.4505714,-0.2513841,0,-0.3454004,-0.1388438,-0.9281277,1278.86,0,0,0,1 0
+2.32632 0.3832846,0.8818775,-0.2745634,0,0.8566163,-0.4505714,-0.2513841,0,-0.3454004,-0.1388438,-0.9281277,1399.86,0,0,0,1 0
+1.9058  -0.1539447,-0.2832125,-0.9466212,0,-0.1585857,0.9527035,-0.2592422,0,0.9752699,0.1102116,-0.1915772,1115.36,0,0,0,1 0
+6.26622 -0.1539447,-0.2832125,-0.9466212,0,-0.1585857,0.9527035,-0.2592422,0,0.9752699,0.1102116,-0.1915772,4065.39,0,0,0,1 0
+2.32632 -0.1539447,-0.2832125,-0.9466212,0,-0.1585857,0.9527035,-0.2592422,0,0.9752699,0.1102116,-0.1915772,1399.86,0,0,0,1 0
+7.47603 -0.1539447,-0.2832125,-0.9466212,0,-0.1585857,0.9527035,-0.2592422,0,0.9752699,0.1102116,-0.1915772,4883.89,0,0,0,1 0
+4.11806 -0.1539447,-0.2832125,-0.9466212,0,-0.1585857,0.9527035,-0.2592422,0,0.9752699,0.1102116,-0.1915772,2612.06,0,0,0,1 0
+13.6304 -0.895577,-0.4178617,0.152753,0,-0.1004795,-0.1445008,-0.9843898,0,0.4334117,-0.8969454,0.08742505,9047.63,0,0,0,1  0
+2.14746 -0.895577,-0.4178617,0.152753,0,-0.1004795,-0.1445008,-0.9843898,0,0.4334117,-0.8969454,0.08742505,1278.86,0,0,0,1  0
+1.9058  0.3832846,0.8818775,-0.2745634,0,0.8566163,-0.4505714,-0.2513841,0,-0.3454004,-0.1388438,-0.9281277,1115.36,0,0,0,1 0
+5.03378 0.3832846,0.8818775,-0.2745634,0,0.8566163,-0.4505714,-0.2513841,0,-0.3454004,-0.1388438,-0.9281277,3231.59,0,0,0,1 0
+5.9596  -0.2362347,0.6308649,0.7390552,0,0.6523478,-0.4607318,0.6018043,0,0.7201635,0.6242881,-0.3027024,3857.95,0,0,0,1  0
+4.76151 -0.895577,-0.4178617,0.152753,0,-0.1004795,-0.1445008,-0.9843898,0,0.4334117,-0.8969454,0.08742505,3047.39,0,0,0,1  0
+1.81745 -0.1539447,-0.2832125,-0.9466212,0,-0.1585857,0.9527035,-0.2592422,0,0.9752699,0.1102116,-0.1915772,1055.59,0,0,0,1 0
+3.47872 -0.1539447,-0.2832125,-0.9466212,0,-0.1585857,0.9527035,-0.2592422,0,0.9752699,0.1102116,-0.1915772,2179.52,0,0,0,1 0
+2.32632 -0.2362347,0.6308649,0.7390552,0,0.6523478,-0.4607318,0.6018043,0,0.7201635,0.6242881,-0.3027024,1399.86,0,0,0,1  0
+12.4066 0.3832846,0.8818775,-0.2745634,0,0.8566163,-0.4505714,-0.2513841,0,-0.3454004,-0.1388438,-0.9281277,8219.63,0,0,0,1 0
+12.4066 -0.1539447,-0.2832125,-0.9466212,0,-0.1585857,0.9527035,-0.2592422,0,0.9752699,0.1102116,-0.1915772,8219.63,0,0,0,1 0
+2.32632 -0.895577,-0.4178617,0.152753,0,-0.1004795,-0.1445008,-0.9843898,0,0.4334117,-0.8969454,0.08742505,1399.86,0,0,0,1  0
+8.19567 -0.895577,-0.4178617,0.152753,0,-0.1004795,-0.1445008,-0.9843898,0,0.4334117,-0.8969454,0.08742505,5370.75,0,0,0,1  0
+4.11806 0.3832846,0.8818775,-0.2745634,0,0.8566163,-0.4505714,-0.2513841,0,-0.3454004,-0.1388438,-0.9281277,2612.06,0,0,0,1 0
+2.58728 -0.895577,-0.4178617,0.152753,0,-0.1004795,-0.1445008,-0.9843898,0,0.4334117,-0.8969454,0.08742505,1576.42,0,0,0,1  0
+5.79983 -0.895577,-0.4178617,0.152753,0,-0.1004795,-0.1445008,-0.9843898,0,0.4334117,-0.8969454,0.08742505,3749.86,0,0,0,1  0
+2.57805 -0.895577,-0.4178617,0.152753,0,-0.1004795,-0.1445008,-0.9843898,0,0.4334117,-0.8969454,0.08742505,1570.18,0,0,0,1  0
+13.6304 -0.1539447,-0.2832125,-0.9466212,0,-0.1585857,0.9527035,-0.2592422,0,0.9752699,0.1102116,-0.1915772,9047.63,0,0,0,1 0
+3.78516 -0.1539447,-0.2832125,-0.9466212,0,-0.1585857,0.9527035,-0.2592422,0,0.9752699,0.1102116,-0.1915772,2386.84,0,0,0,1 0
+4.16725 -0.895577,-0.4178617,0.152753,0,-0.1004795,-0.1445008,-0.9843898,0,0.4334117,-0.8969454,0.08742505,2645.34,0,0,0,1  0

data/spec/jruby/vector_spec.rb

@@ -1,19 +1,20 @@
+require 'spec_helper.rb'
 # Tests if interpreter is JRuby
 
-# describe Daru::Vector do
-#   context ".initialize" do
-#     it "creates a vector object with an MDArray" do
-#       vector = Daru::Vector.new(MDArray.new([5], [1,2,3,4,5]), :uhura)
+describe Daru::Vector do
+  context ".initialize" do
+    it "creates a vector object with an MDArray" do
+      vector = Daru::Vector.new(MDArray.double([5], [1,2,3,4,5]), :uhura)
 
-#       expect(vector[1])  .to eq(2)
-#       expect(vector.name).to eq(:uhura)
-#     end
+      expect(vector[1])  .to eq(2)
+      expect(vector.name).to eq(:uhura)
+    end
 
-#     it "creates a vector object with a Hash with different values" do
-#       vector = Daru::Vector.new { sulu: MDArray.new([5], [1,2,3,4,5])}
+    it "creates a vector object with a Hash with different values" do
+      vector = Daru::Vector.new({ sulu: MDArray.double([5], [1,2,3,4,5])})
 
-#       expect(vector[1])  .to eq(2)
-#       expect(vector.name).to eq(:sulu) 
-#     end
-#   end
-# end
+      expect(vector[1])  .to eq(2)
+      expect(vector.name).to eq(:sulu) 
+    end
+  end
+end if RUBY_ENGINE == 'jruby'

data/spec/mri/dataframe_spec.rb

@@ -1,5 +1,103 @@
 require 'spec_helper.rb'
 
 describe Daru::DataFrame do
-  
-end
+  context "DataFrame from normal array vectors" do
+    
+    before :each do
+      @df = Daru::DataFrame.new({a: Daru::Vector.new(1..3), 
+        b: (50..52).daru_vector, b_bad: ['Jesse', 'Walter', 'Hank'].daru_vector})
+
+      @vector = Daru::Vector.new 1..3, :a
+    end
+
+    it "checks for the size of DataFrame" do
+      expect(@df.size).to eq(3)
+    end
+
+    it "raises exception for uneven vectors" do
+      expect do
+        df = Daru::DataFrame.new({a: (1..5).dv, b: [1,2,3].daru_vector})
+      end.to raise_error
+    end
+
+    it "returns vector by specifying as method" do
+      expect(@df.a).to eq(@vector)
+    end
+
+    it "returns vector by specifying as index" do
+      expect(@df[:a]).to eq(@vector)
+    end
+
+    it "returns vector by specifying as a column argument" do
+      expect(@df.column(:a)).to eq(@vector)
+    end
+
+    it "returns a row" do
+      r = @df.row 0
+
+      expect(r).to eq([1,50,'Jesse'])
+    end
+
+    it "iterates over columns" do
+      @df.each_column do |col|
+        expect(col.is_a?(Daru::Vector)).to be(true)
+        expect([:a, :b, :b_bad].include? col.name).to be(true)
+      end
+    end
+
+    it "iterates over rows" do
+      @df.each_row do |row|
+        expect(row.size).to be(@df.fields.size)
+      end
+    end
+
+    it "shows column fields" do
+      expect(@df.fields).to eq([:a, :b, :b_bad])
+    end
+
+    it "inserts a new vector" do
+      @df.insert_vector :c, Daru::Vector.new([3,6,9])
+
+      expect(@df.fields.include?(:c)).to be(true)
+    end
+
+    it "inserts a new row" do
+      @df.insert_row [6,6,"Fred"]
+
+      expect(@df.a.vector)    .to eq([1,2,3,6])
+      expect(@df.b.vector)    .to eq([50,51,52,6])
+      expect(@df.b_bad.vector).to eq(['Jesse', 'Walter', 'Hank', 'Fred'])
+    end
+
+    it "raises an error for inappropriate row insertion" do
+      expect { @df.insert_row [1,1] }.to raise_error
+    end
+
+    it "deletes a vector" do
+      @df.delete :a
+
+      expect(@df.fields.include? :a).to be(false)
+    end
+  end
+
+  context "DataFrame loads from files" do
+
+    it "loads a DataFrame from CSV" do
+      df = Daru::DataFrame.from_csv('spec/fixtures/matrix_test.csv', 
+        {col_sep: ' ', headers: true}) do |csv|
+        csv.convert do |field, info|
+          case info[:header]
+          when :true_transform
+            field.split(',').map { |s| s.to_f }
+          else
+            field
+          end
+        end
+      end
+
+      expect(df.fields).to eq([:image_resolution, :true_transform, :mls])
+      expect(df[:image_resolution].first).to eq(6.55779)
+      expect(df.column(:true_transform).first[15]).to eq(1.0)
+    end
+  end
+end if RUBY_ENGINE == 'ruby'

data/spec/mri/vector_spec.rb

@@ -57,23 +57,48 @@ describe Daru::Vector do
     end
   end
 
-  context "#each" do
-    it "executes the given block" do
-      sum = 0
-      vector = Daru::Vector.new 1..5, :mickey
+  context "tests for methods" do # TODO: Name this better
+    before do
+      @anakin = Daru::Vector.new NMatrix.new([5], [1,2,3,4,5]), :anakin
+      @luke   = Daru::Vector.new NMatrix.new([3], [3,4,5,6])  , :luke
+    end
 
-      vector.each{ |e| sum += e}
+    it "checks for an each block" do
+      sum = 0
 
+      @anakin.each{ |e| sum += e}
       expect(sum).to eq(15)
     end
-  end
 
-  context "#==" do
-    it "checks for equality of vectors" do
-      anakin = Daru::Vector.new 1..5, :anakin
-      luke = Daru::Vector.new 3..6, :luke
+    it "checks for inequality of vectors" do
+      expect(@anakin == @luke).to be(false)
+    end
+
+    it "calculates maximum value" do
+      expect(@anakin.max).to eq(5)
+    end
+
+    it "calculates minimmum value" do
+      expect(@anakin.min).to eq(1)
+    end
+
+    it "delegates to the internal array storage" do
+      expect(@anakin.size).to eq(@anakin.to_a.size)
+    end
+
+    it "assigns on index" do
+      @anakin[0] = 666
+      expect(@anakin[0]).to eq(666)
+    end
+
+    it "returns only the vector object" do
+      expect(@anakin.vector == NMatrix.new([5], [1,2,3,4,5])).to be(true)
+    end
+
+    it "tests for equality" do
+      clone = Daru::Vector.new NMatrix.new([5], [1,2,3,4,5]), :clone
 
-      expect(anakin == luke).to be(false)
+      expect(@anakin == clone).to be(false)
     end
   end
-end
+end if RUBY_ENGINE == 'ruby'

data/spec/spec_helper.rb

@@ -1,5 +1,10 @@
 require 'rspec'
-require 'nmatrix'
+
+if RUBY_ENGINE == 'jruby'
+  require 'mdarray'
+else
+  require 'nmatrix'
+end
 
 $LOAD_PATH.unshift(File.join(File.dirname(__FILE__), '..', 'lib'))
 $LOAD_PATH.unshift(File.dirname(__FILE__))

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: daru
 version: !ruby/object:Gem::Version
-  version: 0.0.1
+  version: 0.0.2
 platform: ruby
 authors:
 - Sameer Deshmukh
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2014-10-06 00:00:00.000000000 Z
+date: 2014-10-09 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: bundler
@@ -28,16 +28,16 @@ dependencies:
   name: rspec
   requirement: !ruby/object:Gem::Requirement
     requirements:
-    - - "~>"
+    - - ">="
       - !ruby/object:Gem::Version
-        version: '3.0'
+        version: '0'
   type: :development
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
-    - - "~>"
+    - - ">="
       - !ruby/object:Gem::Version
-        version: '3.0'
+        version: '0'
 - !ruby/object:Gem::Dependency
   name: nmatrix
   requirement: !ruby/object:Gem::Requirement
@@ -52,9 +52,9 @@ dependencies:
     - - "~>"
       - !ruby/object:Gem::Version
         version: 0.1.0.rc5
-description: "Daru (Data Analysis in RUby) is a library for storage, analysis and
-  manipulation\nof data. It aims to be the preferred data storage library for all
-  Ruby gems, promoting \ninteroperability and easy of use between various ruby packages.\n"
+description: |
+  Daru (Data Analysis in RUby) is a library for storage, analysis and manipulation
+  of data.
 email:
 - sameer.deshmukh93@gmail.com
 executables: []
@@ -71,6 +71,7 @@ files:
 - lib/daru/dataframe.rb
 - lib/daru/vector.rb
 - lib/version.rb
+- spec/fixtures/matrix_test.csv
 - spec/jruby/dataframe_spec.rb
 - spec/jruby/vector_spec.rb
 - spec/mri/dataframe_spec.rb
@@ -101,6 +102,7 @@ signing_key:
 specification_version: 4
 summary: Data Analysis in RUby
 test_files:
+- spec/fixtures/matrix_test.csv
 - spec/jruby/dataframe_spec.rb
 - spec/jruby/vector_spec.rb
 - spec/mri/dataframe_spec.rb