daru 0.1.2 → 0.1.3

data/daru.gemspec

@@ -31,7 +31,7 @@ Gem::Specification.new do |spec|
 *************************************************************************
 Thank you for installing daru!
 
-  oOOOOOo 
+  oOOOOOo
  ,|    oO
 //|     |
 \\\\|     |
@@ -39,8 +39,8 @@ Thank you for installing daru!
   `-----`
 
 
-Hope you love daru! For enhanced interactivity and better visualizations, 
-consider using gnuplotrb and nyaplot with iruby. For statistics use the 
+Hope you love daru! For enhanced interactivity and better visualizations,
+consider using gnuplotrb and nyaplot with iruby. For statistics use the
 statsample family.
 
 Read the README for interesting use cases and examples.
@@ -63,9 +63,9 @@ EOF
   spec.add_development_dependency 'nmatrix', '~> 0.2.1'
   spec.add_development_dependency 'distribution', '~> 0.7'
   spec.add_development_dependency 'rb-gsl', '~>1.16'
-  spec.add_development_dependency 'bloomfilter-rb', '~> 2.1'
   spec.add_development_dependency 'dbd-sqlite3'
   spec.add_development_dependency 'dbi'
   spec.add_development_dependency 'activerecord', '~> 4.0'
   spec.add_development_dependency 'sqlite3'
+  spec.add_development_dependency 'rubocop', '>= 0.40.0'
 end

data/lib/daru.rb

@@ -11,7 +11,7 @@ module Daru
     'THU' => 4,
     'FRI' => 5,
     'SAT' => 6
-  }
+  }.freeze
 
   MONTH_DAYS = {
     1 => 31,
@@ -26,19 +26,20 @@ module Daru
     10 => 31,
     11 => 30,
     12 => 31
-  }
-  
-  SPLIT_TOKEN = ','
+  }.freeze
+
+  @lazy_update = false
+
+  SPLIT_TOKEN = ','.freeze
+
   class << self
-    @@lazy_update = false
-    
     # A variable which will set whether Vector metadata is updated immediately or lazily.
     # Call the #update method every time a values are set or removed in order to update
     # metadata like positions of missing values.
     attr_accessor :lazy_update
-    
+
     def create_has_library(library)
-      lib_underscore = library.to_s.gsub(/-/, '_')
+      lib_underscore = library.to_s.tr('-', '_')
       define_singleton_method("has_#{lib_underscore}?") do
         cv = "@@#{lib_underscore}"
         unless class_variable_defined? cv
@@ -58,7 +59,6 @@ module Daru
   create_has_library :gsl
   create_has_library :nmatrix
   create_has_library :nyaplot
-  create_has_library :'bloomfilter-rb'
 end
 
 autoload :Spreadsheet, 'spreadsheet'

data/lib/daru/accessors/array_wrapper.rb

@@ -6,7 +6,7 @@ module Daru
       extend Forwardable
 
       def_delegators :@data, :slice!
-      
+
       def each(&block)
         @data.each(&block)
         self
@@ -66,34 +66,38 @@ module Daru
         ArrayWrapper.new @data.dup, @context
       end
 
+      def compact
+        @data - @context.missing_values
+      end
+
       def mean
-        sum.quo(@size - @context.missing_positions.size).to_f
+        values_to_sum = compact
+        return nil if values_to_sum.empty?
+        sum = values_to_sum.inject :+
+        sum.quo(values_to_sum.size).to_f
       end
 
       def product
-        @data.inject(1) { |m,e| m*e unless e.nil? }
+        compact.inject :*
       end
 
       def max
-        @data.max
+        compact.max
       end
 
       def min
-        @data.min
+        compact.min
       end
 
       def sum
-        @data.inject(0) do |memo ,e|
-          memo += e unless e.nil? #TODO: Remove this conditional somehow!
-          memo
-        end
+        compact.inject :+
       end
 
-     private
+      private
 
       def set_size
         @size = @data.size
       end
     end
   end
-end
+end

data/lib/daru/accessors/dataframe_by_row.rb

@@ -14,4 +14,4 @@ module Daru
       end
     end
   end
-end
+end

data/lib/daru/accessors/gsl_wrapper.rb

@@ -3,33 +3,33 @@ module Daru
     module GSLStatistics
       def vector_standardized_compute(m,sd)
         Daru::Vector.new @data.collect { |x| (x.to_f - m).quo(sd) }, dtype: :gsl,
-          index: @context.index, name: @context.name
+                                                                     index: @context.index, name: @context.name
       end
-      
+
       def vector_centered_compute(m)
-        Daru::Vector.new @data.collect {|x| (x.to_f - m)}, dtype: :gsl,
-          index: @context.index, name: @context.name
+        Daru::Vector.new @data.collect { |x| (x.to_f - m) }, dtype: :gsl,
+                                                             index: @context.index, name: @context.name
       end
 
       def sample_with_replacement(sample=1)
-        r = GSL::Rng.alloc(GSL::Rng::MT19937,rand(10000))
-        Daru::Vector.new(r.sample(@data, sample).to_a, dtype: :gsl, 
-          index: @context.index, name: @context.name)
+        r = GSL::Rng.alloc(GSL::Rng::MT19937,rand(10_000))
+        Daru::Vector.new(r.sample(@data, sample).to_a, dtype: :gsl,
+                                                       index: @context.index, name: @context.name)
       end
-      
+
       def sample_without_replacement(sample=1)
-        r = GSL::Rng.alloc(GSL::Rng::MT19937,rand(10000))
+        r = GSL::Rng.alloc(GSL::Rng::MT19937,rand(10_000))
         r.choose(@data, sample).to_a
       end
 
       def median
-        GSL::Stats::median_from_sorted_data(@data.sort)
+        GSL::Stats.median_from_sorted_data(@data.sort)
       end
-      
+
       def variance_sample(m)
-        @data.variance_m
+        @data.variance(m)
       end
-      
+
       def standard_deviation_sample(m)
         @data.sd(m)
       end
@@ -56,16 +56,27 @@ module Daru
       extend Forwardable
       include Daru::Accessors::GSLStatistics
 
-      def_delegators :@data, :[], :size, :to_a, :each, :mean, 
-        :sum, :prod, :max, :min
-
-      alias :product :prod
+      def_delegators :@data, :[], :size, :to_a, :each
 
       attr_reader :data
 
+      def compact
+        # set missing to [] incase @context is not Daru::Vector
+        missing = @context.missing_values rescue []
+        ::GSL::Vector.alloc(@data.to_a - missing.map(&:to_f))
+      end
+
+      [:mean, :min, :max, :prod, :sum].each do |method|
+        define_method(method) do
+          compact.send(method.to_sym) rescue nil
+        end
+      end
+
+      alias :product :prod
+
       def each(&block)
         @data.each(&block)
-        self  
+        self
       end
 
       def map!(&block)
@@ -110,4 +121,4 @@ module Daru
       end
     end
   end
-end if Daru.has_gsl?
+end if Daru.has_gsl?

data/lib/daru/accessors/mdarray_wrapper.rb

@@ -1,9 +1,7 @@
 module Daru
   module Accessors
-
     # Internal class for wrapping MDArray
     class MDArrayWrapper
-
     end
   end
-end
+end

data/lib/daru/accessors/nmatrix_wrapper.rb

@@ -19,7 +19,7 @@ module Daru
       end
 
       attr_reader :size, :data, :nm_dtype
-      
+
       def initialize vector, context, nm_dtype=:int32
         @size = vector.size
         @data = NMatrix.new [@size*2], vector.to_a, dtype: nm_dtype
@@ -32,47 +32,47 @@ module Daru
         return @data[*index] if index[0] < @size
         nil
       end
- 
+
       def []= index, value
-        raise ArgumentError, "Index #{index} does not exist" if 
-          index > @size and index < @data.size
+        raise ArgumentError, "Index #{index} does not exist" if
+          index > @size && index < @data.size
         resize     if index >= @data.size
         @size += 1 if index == @size
-        
+
         @data = @data.cast(dtype: :object) if value.nil?
         @data[index] = value
-      end 
- 
+      end
+
       def == other
         @data[0...@size] == other[0...@size] and @size == other.size
       end
- 
+
       def delete_at index
         arry = @data.to_a
         arry.delete_at index
         @data = NMatrix.new [(2*@size-1)], arry, dtype: @nm_dtype
         @size -= 1
       end
- 
+
       def index key
         @data.to_a.index key
       end
- 
+
       def << element
         resize if @size >= @data.size
         self[@size] = element
       end
- 
+
       def to_a
         @data[0...@size].to_a
       end
- 
+
       def dup
         NMatrixWrapper.new @data[0...@size].to_a, @context, @nm_dtype
       end
 
-      def resize size = @size*2
-        raise ArgumentError, "Size must be greater than current size" if size < @size
+      def resize size=@size*2
+        raise ArgumentError, 'Size must be greater than current size' if size < @size
 
         @data = NMatrix.new [size], @data.to_a, dtype: @nm_dtype
       end
@@ -98,4 +98,4 @@ module Daru
       end
     end
   end
-end if Daru.has_nmatrix?
+end if Daru.has_nmatrix?

data/lib/daru/core/group_by.rb

@@ -1,12 +1,11 @@
 module Daru
   module Core
     class GroupBy
-
       attr_reader :groups
 
       # Iterate over each group created by group_by. A DataFrame is yielded in
       # block.
-      def each_group &block
+      def each_group
         groups.keys.each do |k|
           yield get_group(k)
         end
@@ -18,7 +17,14 @@ module Daru
         @context = context
         vectors = names.map { |vec| context[vec].to_a }
         tuples  = vectors[0].zip(*vectors[1..-1])
-        keys    = tuples.uniq.sort { |a,b| a && b ? a.compact <=> b.compact : a ? 1 : -1 }
+        keys    =
+          tuples.uniq.sort do |a,b|
+            if a && b
+              a.compact <=> b.compact
+            else
+              a ? 1 : -1
+            end
+          end
 
         keys.each do |key|
           @groups[key] = all_indices_for(tuples, key)
@@ -29,13 +35,13 @@ module Daru
       # Get a Daru::Vector of the size of each group.
       def size
         index =
-        if multi_indexed_grouping?
-          Daru::MultiIndex.from_tuples @groups.keys
-        else
-          Daru::Index.new @groups.keys.flatten
-        end
+          if multi_indexed_grouping?
+            Daru::MultiIndex.from_tuples @groups.keys
+          else
+            Daru::Index.new @groups.keys.flatten
+          end
 
-        values = @groups.values.map { |e| e.size }
+        values = @groups.values.map(&:size)
         Daru::Vector.new(values, index: index, name: :size)
       end
 
@@ -194,11 +200,57 @@ module Daru
         indexes.each do |idx|
           rows << transpose[idx]
         end
+
+        new_index =
+          begin
+            @context.index[indexes]
+          rescue IndexError
+            indexes
+          end
         Daru::DataFrame.rows(
-          rows, index: @context.index[indexes], order: @context.vectors)
+          rows, index: new_index, order: @context.vectors
+        )
+      end
+
+      # Iteratively applies a function to the values in a group and accumulates the result.
+      # @param init (nil) The initial value of the accumulator.
+      # @param block [Proc] A proc or lambda that accepts two arguments.  The first argument
+      #                     is the accumulated result.  The second argument is a DataFrame row.
+      # @example Usage of reduce
+      #   df = Daru::DataFrame.new({
+      #     a: ['a','b'] * 3,
+      #     b: [1,2,3] * 2,
+      #     c: 'A'..'F'
+      #   })
+      #   df.group_by([:a]).reduce('') { |result, row| result += row[:c]; result }
+      #   # =>
+      #   # #<Daru::Vector:70343147159900 @name = nil @metadata = {} @size = 2 >
+      #   #     nil
+      #   #   a ACE
+      #   #   b BDF
+      def reduce(init=nil)
+        result_hash = @groups.each_with_object({}) do |(group, indices), h|
+          group_indices = indices.map { |v| @context.index.to_a[v] }
+
+          grouped_result = init
+          group_indices.each do |idx|
+            grouped_result = yield(grouped_result, @context.row[idx])
+          end
+
+          h[group] = grouped_result
+        end
+
+        index =
+          if multi_indexed_grouping?
+            Daru::MultiIndex.from_tuples result_hash.keys
+          else
+            Daru::Index.new result_hash.keys.flatten
+          end
+
+        Daru::Vector.new(result_hash.values, index: index)
       end
 
-     private
+      private
 
       def select_groups_from method, quantity
         selection     = @context
@@ -219,13 +271,13 @@ module Daru
         multi_index = multi_indexed_grouping?
         rows, order = [], []
 
-        @groups.each do |group, indexes|
+        @groups.each do |_group, indexes|
           single_row = []
           @non_group_vectors.each do |ngvector|
             vec = @context[ngvector]
-            if method_type == :numeric and vec.type == :numeric
+            if method_type == :numeric && vec.type == :numeric
               slice = vec[*indexes]
-              single_row << (slice.is_a?(Numeric) ? slice : slice.send(method))
+              single_row << (slice.is_a?(Daru::Vector) ? slice.send(method) : slice)
             end
           end
 
@@ -234,7 +286,7 @@ module Daru
 
         @non_group_vectors.each do |ngvec|
           order << ngvec if
-            (method_type == :numeric and @context[ngvec].type == :numeric)
+            method_type == :numeric && @context[ngvec].type == :numeric
         end
 
         index = @groups.keys
@@ -256,7 +308,8 @@ module Daru
       end
 
       def multi_indexed_grouping?
-        @groups.keys[0][1] ? true : false
+        return false unless @groups.keys[0]
+        @groups.keys[0].size > 1 ? true : false
       end
     end
   end