red_amber 0.2.1 → 0.2.2

data/lib/red_amber/helper.rb

@@ -17,10 +17,6 @@ module RedAmber
       enum.all?(Integer)
     end
 
-    def sym_or_str?(enum)
-      enum.all? { |e| e.is_a?(Symbol) || e.is_a?(String) }
-    end
-
     def booleans?(enum)
       enum.all? { |e| e.is_a?(TrueClass) || e.is_a?(FalseClass) || e.is_a?(NilClass) }
     end
@@ -29,32 +25,38 @@ module RedAmber
       DataFrame.new(key => vector.data)
     end
 
-    def parse_to_vector(args)
+    def parse_to_vector(args, vsize: size)
       a = args.reduce([]) do |accum, elem|
-        accum.concat(normalize_element(elem))
+        accum.concat(normalize_element(elem, vsize: vsize))
       end
       Vector.new(a)
     end
 
-    def normalize_element(elem)
+    def normalize_element(elem, vsize: size)
       case elem
-      when Numeric, String, Symbol, TrueClass, FalseClass, NilClass
-        [elem]
+      when NilClass
+        [nil]
       when Range
-        both_end = [elem.begin, elem.end]
-        both_end[1] -= 1 if elem.exclude_end? && elem.end.is_a?(Integer)
-
-        if both_end.any?(Integer) || both_end.all?(&:nil?)
-          if both_end.any? { |e| e&.>=(size) || e&.<(-size) }
-            raise DataFrameArgumentError, "Index out of range: #{elem} for 0..#{size - 1}"
+        bg = elem.begin
+        en = elem.end
+        if [bg, en].any?(Integer)
+          bg += vsize if bg&.negative?
+          en += vsize if en&.negative?
+          en -= 1 if en.is_a?(Integer) && elem.exclude_end?
+          if bg&.negative? || (en && en >= vsize)
+            raise DataFrameArgumentError, "Index out of range: #{elem} for 0..#{vsize - 1}"
           end
 
-          (0...size).to_a[elem]
+          Array(0...vsize)[elem]
+        elsif bg.nil? && en.nil?
+          Array(0...vsize)
         else
-          elem.to_a
+          Array[elem]
         end
+      when Enumerator
+        elem.to_a
       else
-        Array(elem)
+        Array[elem]
       end
     end
   end

data/lib/red_amber/vector.rb

@@ -11,27 +11,28 @@ module RedAmber
     include Helper
 
     def initialize(*array)
-      @key = nil # default is 'headless'
-      if array.empty? || array[0].nil?
+      @key = nil # default is 'headless' Vector
+      if array.empty? || array.first.nil?
         Vector.new([])
       else
         array.flatten!
-        case array[0]
-        when Vector
-          @data = array[0].data
-          return
-        when Arrow::Array, Arrow::ChunkedArray
-          @data = array[0]
-          return
-        when Range
-          @data = Arrow::Array.new(Array(array[0]))
-          return
-        end
-        begin
-          @data = Arrow::Array.new(Array(array))
-        rescue Error
-          raise VectorArgumentError, "Invalid argument: #{array}"
-        end
+        @data =
+          case array
+          in [Vector => v]
+            v.data
+          in [Arrow::Array => a]
+            a
+          in [Arrow::ChunkedArray => ca]
+            ca
+          in [Range => r]
+            Arrow::Array.new(Array(r))
+          else
+            begin
+              Arrow::Array.new(Array(array))
+            rescue Error
+              raise VectorArgumentError, "Invalid argument: #{array}"
+            end
+          end
       end
     end
 
@@ -43,19 +44,24 @@ module RedAmber
     end
 
     def inspect(limit: 80)
-      sio = StringIO.new << '['
-      to_a.each_with_object(sio).with_index do |(e, s), i|
-        next_str = "#{s.size > 1 ? ', ' : ''}#{e.inspect}"
-        if (s.size + next_str.size) < limit
-          s << next_str
-        else
-          s << ', ... ' if i < size
-          break
+      if ENV.fetch('RED_AMBER_OUTPUT_MODE', 'Table').casecmp('MINIMUM').zero?
+        # Better performance than `.upcase == 'MINIMUM'``
+        "#{self.class}(:#{type}, size=#{size})"
+      else
+        sio = StringIO.new << '['
+        to_a.each_with_object(sio).with_index do |(e, s), i|
+          next_str = "#{s.size > 1 ? ', ' : ''}#{e.inspect}"
+          if (s.size + next_str.size) < limit
+            s << next_str
+          else
+            s << ', ... ' if i < size
+            break
+          end
         end
-      end
-      sio << ']'
+        sio << ']'
 
-      format "#<#{self.class}(:#{type}, size=#{size}):0x%016x>\n%s\n", object_id, sio.string
+        format "#<#{self.class}(:#{type}, size=#{size}):0x%016x>\n%s\n", object_id, sio.string
+      end
     end
 
     def values
@@ -71,7 +77,7 @@ module RedAmber
     alias_method :indeces, :indices
 
     def to_ary
-      to_a
+      values
     end
 
     def size
@@ -110,6 +116,10 @@ module RedAmber
       type_class == Arrow::StringDataType
     end
 
+    def dictionary?
+      type_class == Arrow::DictionaryDataType
+    end
+
     def temporal?
       type_class < Arrow::TemporalDataType
     end
@@ -126,10 +136,19 @@ module RedAmber
       end
     end
 
+    def map(&block)
+      return enum_for(:map) unless block
+
+      Vector.new(to_a.map(&block))
+    end
+    alias_method :collect, :map
+
+    # undocumented
     def chunked?
       @data.is_a? Arrow::ChunkedArray
     end
 
+    # undocumented
     def n_chunks
       chunked? ? @data.n_chunks : 0
     end

data/lib/red_amber/vector_selectable.rb

@@ -25,7 +25,13 @@ module RedAmber
     end
 
     # TODO: support for option {null_selection_behavior: :drop}
-    def filter(*booleans)
+    def filter(*booleans, &block)
+      if block
+        raise VectorArgumentError, 'Must not specify both arguments and block.' unless booleans.empty?
+
+        booleans = [yield]
+      end
+
       booleans.flatten!
       return Vector.new([]) if booleans.empty?
 
@@ -46,6 +52,8 @@ module RedAmber
 
       filter_by_array(boolean_array) # returns sub Vector
     end
+    alias_method :select, :filter
+    alias_method :find_all, :filter
 
     #   @param indices
     #   @param booleans

data/lib/red_amber/vector_updatable.rb

@@ -8,9 +8,10 @@ module RedAmber
   # Functions to make up some data (especially missing) for new data.
   module VectorUpdatable
     # Replace data
-    # @param arg [Array, Vector, Arrow::Array] index specifier
+    # @param arg [Array, Vector, Arrow::Array] index specifier or boolean
     # @param replacer [Array, Vector, Arrow::Array] new data to replace for.
-    # @return [Vector] Replaced new Vector
+    # @return [Vector] Replaced new Vector.
+    #   If arg has no true, return self.
     def replace(args, replacer)
       args =
         case args
@@ -24,10 +25,12 @@ module RedAmber
       replacer = Array(replacer)
       return self if args.empty? || args[0].nil?
 
-      replacer = nil if replacer.empty?
       vector = parse_to_vector(args)
+      replacer = nil if replacer.empty?
       booleans =
         if vector.boolean?
+          return self unless vector.any
+
           vector
         elsif vector.numeric?
           replacer.sort_by! { |x| args[replacer.index(x)] } if replacer # rubocop:disable Style/SafeNavigation

data/lib/red_amber/version.rb

@@ -1,5 +1,5 @@
 # frozen_string_literal: true
 
 module RedAmber
-  VERSION = '0.2.1'
+  VERSION = '0.2.2'
 end

data/lib/red_amber.rb

@@ -5,6 +5,7 @@ require 'arrow'
 require_relative 'red_amber/helper'
 require_relative 'red_amber/data_frame_displayable'
 require_relative 'red_amber/data_frame_indexable'
+require_relative 'red_amber/data_frame_loadsave'
 require_relative 'red_amber/data_frame_reshaping'
 require_relative 'red_amber/data_frame_selectable'
 require_relative 'red_amber/data_frame_variable_operation'

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: red_amber
 version: !ruby/object:Gem::Version
-  version: 0.2.1
+  version: 0.2.2
 platform: ruby
 authors:
 - Hirokazu SUZUKI (heronshoes)
 autorequire:
 bindir: exe
 cert_chain: []
-date: 2022-09-07 00:00:00.000000000 Z
+date: 2022-10-03 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: red-arrow
@@ -45,16 +45,25 @@ files:
 - doc/CODE_OF_CONDUCT.md
 - doc/DataFrame.md
 - doc/Vector.md
-- doc/examples_of_red_amber.ipynb
 - doc/image/arrow_table_new.png
+- doc/image/basic_verbs.png
 - doc/image/dataframe/assign.png
+- doc/image/dataframe/assign_operation.png
 - doc/image/dataframe/drop.png
 - doc/image/dataframe/pick.png
+- doc/image/dataframe/pick_operation.png
 - doc/image/dataframe/remove.png
 - doc/image/dataframe/rename.png
+- doc/image/dataframe/rename_operation.png
+- doc/image/dataframe/reshaping_DataFrames.png
 - doc/image/dataframe/slice.png
+- doc/image/dataframe/slice_operation.png
 - doc/image/dataframe_model.png
 - doc/image/example_in_red_arrow.png
+- doc/image/group_operation.png
+- doc/image/replace-if_then.png
+- doc/image/reshaping_dataframe.png
+- doc/image/screenshot.png
 - doc/image/tdr.png
 - doc/image/tdr_and_table.png
 - doc/image/tidy_data_in_TDR.png
@@ -69,6 +78,7 @@ files:
 - lib/red_amber/data_frame.rb
 - lib/red_amber/data_frame_displayable.rb
 - lib/red_amber/data_frame_indexable.rb
+- lib/red_amber/data_frame_loadsave.rb
 - lib/red_amber/data_frame_reshaping.rb
 - lib/red_amber/data_frame_selectable.rb
 - lib/red_amber/data_frame_variable_operation.rb