red_amber 0.2.0 → 0.2.1

data/lib/red_amber/data_frame.rb

@@ -159,12 +159,19 @@ module RedAmber
       @vectors || @vectors = init_instance_vars(:vectors)
     end
 
-    # Returns row indices (0...size) in an Array.
+    # Returns row indices (start...(size+start)) in an Array.
     #
+    # @param start [Object]
+    #   Object which have #succ method.
     # @return [Array]
-    #   An Array of all indices of rows.
-    def indices
-      (0...size).to_a
+    #   An Array of indices of the row.
+    # @example
+    #   (when self.size == 5)
+    #   - indices #=> [0, 1, 2, 3, 4]
+    #   - indices(1) #=> [1, 2, 3, 4, 5]
+    #   - indices('a') #=> ['a', 'b', 'c', 'd', 'e']
+    def indices(start = 0)
+      (start..).take(size)
     end
     alias_method :indexes, :indices
 
@@ -225,6 +232,18 @@ module RedAmber
       g
     end
 
+    def method_missing(name, *args, &block)
+      return v(name) if args.empty?
+
+      super
+    end
+
+    def respond_to_missing?(name, include_private)
+      return true if key?(name)
+
+      super
+    end
+
     private
 
     # initialize @variable, @keys, @vectors and return one of them

data/lib/red_amber/data_frame_displayable.rb

@@ -154,7 +154,7 @@ module RedAmber
 
     def format_table(width: 80, head: 5, tail: 3, n_digit: 2)
       original = self
-      indices = size > head + tail ? [*0...head, *(size - tail)...size] : [*0...size]
+      indices = size > head + tail ? [*0..head, *(size - tail)...size] : [*0...size]
       df = slice(indices).assign do
         assigner = { INDEX_KEY => indices.map { |i| (i + 1).to_s } }
         vectors.each_with_object(assigner) do |v, a|
@@ -173,12 +173,12 @@ module RedAmber
       end
 
       df = df.pick { [INDEX_KEY, keys - [INDEX_KEY]] }
-      df = size > head + tail ? df[0, 0, 0...head, 0, -tail..-1] : df[0, 0, 0..-1]
+      df = size > head + tail ? df[0, 0, 0..head, -tail..-1] : df[0, 0, 0..-1]
       df = df.assign do
         vectors.each_with_object({}) do |v, assigner|
           vec = v.replace(0, v.key == INDEX_KEY ? '' : v.key.to_s)
                  .replace(1, v.key == INDEX_KEY ? '' : "<#{original[v.key].type}>")
-          assigner[v.key] = size > head + tail ? vec.replace(head + 2, ':') : vec
+          assigner[v.key] = original.size > head + tail + 1 ? vec.replace(head + 2, ':') : vec
         end
       end
 

data/lib/red_amber/data_frame_reshaping.rb

@@ -5,20 +5,20 @@ module RedAmber
   module DataFrameReshaping
     # Transpose a wide DataFrame.
     #
-    # @param key [Symbol, FalseClass] key of the index column
+    # @param key [Symbol] key of the index column
     #   to transepose into keys.
-    #   If it is false, keys[0] is used.
-    # @param new_key [Symbol, FalseClass] key name of transposed index column.
-    #   If it is false, :name is used. If it already exists, :name1.succ is used.
+    #   If it is not specified, keys[0] is used.
+    # @param new_key [Symbol] key name of transposed index column.
+    #   If it is not specified, :N is used. If it already exists, :N1 or :N1.succ is used.
     # @return [DataFrame] trnsposed DataFrame
-    def transpose(key: keys.first, new_key: :name)
-      raise DataFrameArgumentError, "Not include: #{key}" unless keys.include?(key)
+    def transpose(key: keys.first, name: :N)
+      raise DataFrameArgumentError, "Self does not include: #{key}" unless keys.include?(key)
 
       # Find unused name
       new_keys = self[key].to_a.map { |e| e.to_s.to_sym }
-      new_key = (:name1..).find { |k| !new_keys.include?(k) } if new_keys.include?(new_key)
+      name = (:N1..).find { |k| !new_keys.include?(k) } if new_keys.include?(name)
 
-      hash = { new_key => (keys - [key]) }
+      hash = { name => (keys - [key]) }
       i = keys.index(key)
       each_row do |h|
         k = h.values[i]
@@ -33,7 +33,7 @@ module RedAmber
     # @param name [Symbol, String] key of the column which is come **from values**.
     # @param value [Symbol, String] key of the column which is come **from values**.
     # @return [DataFrame] long DataFrame.
-    def to_long(*keep_keys, name: :name, value: :value)
+    def to_long(*keep_keys, name: :N, value: :V)
       not_included = keep_keys - keys
       raise DataFrameArgumentError, "Not have keys #{not_included}" unless not_included.empty?
 
@@ -63,7 +63,7 @@ module RedAmber
     # @param name [Symbol, String] key of the column which will be expanded **to key names**.
     # @param value [Symbol, String] key of the column which will be expanded **to values**.
     # @return [DataFrame] wide DataFrame.
-    def to_wide(name: :name, value: :value)
+    def to_wide(name: :N, value: :V)
       name = name.to_sym
       raise DataFrameArgumentError, "Invalid key: #{name}" unless keys.include?(name)
 

data/lib/red_amber/data_frame_selectable.rb

@@ -3,8 +3,8 @@
 module RedAmber
   # mix-in for the class DataFrame
   module DataFrameSelectable
-    # select variables: [symbol] or [string]
-    # select observations: [array of index], [range]
+    # select columns: [symbol] or [string]
+    # select rows: [array of index], [range]
     def [](*args)
       args.flatten!
       raise DataFrameArgumentError, 'Empty dataframe' if empty?
@@ -22,17 +22,17 @@ module RedAmber
       raise DataFrameArgumentError, "Invalid argument: #{args}"
     end
 
-    # slice and select some observations to create sub DataFrame
+    # slice and select rows to create sub DataFrame
     def slice(*args, &block)
       slicer = args
       if block
         raise DataFrameArgumentError, 'Must not specify both arguments and block.' unless args.empty?
 
-        slicer = instance_eval(&block)
+        slicer = [instance_eval(&block)]
       end
-      slicer = [slicer].flatten
+      slicer.flatten!
 
-      raise DataFrameArgumentError, 'Empty dataframe' if empty?
+      raise DataFrameArgumentError, 'Self is an empty dataframe' if empty?
       return remove_all_values if slicer.empty? || slicer[0].nil?
 
       vector = parse_to_vector(slicer)
@@ -46,15 +46,59 @@ module RedAmber
       raise DataFrameArgumentError, "Invalid argument #{slicer}"
     end
 
-    # remove selected observations to create sub DataFrame
+    def slice_by(key, keep_key: false, &block)
+      raise DataFrameArgumentError, 'Self is an empty dataframe' if empty?
+      raise DataFrameArgumentError, 'No block given' unless block
+      raise DataFrameArgumentError, "#{key} is no a key of self" unless key?(key)
+      return self if key.nil?
+
+      slicer = instance_eval(&block)
+      return DataFrame.new unless slicer
+
+      if slicer.is_a?(Range)
+        from = slicer.begin
+        from =
+          if from.is_a?(String)
+            self[key].index(from)
+          elsif from.nil?
+            0
+          elsif from < 0
+            size + from
+          else
+            from
+          end
+        to = slicer.end
+        to =
+          if to.is_a?(String)
+            self[key].index(to)
+          elsif to.nil?
+            size - 1
+          elsif to < 0
+            size + to
+          else
+            to
+          end
+        slicer = (from..to).to_a
+      else
+        slicer = slicer.map { |x| x.is_a?(String) ? self[key].index(x) : x }
+      end
+
+      if keep_key
+        take(slicer)
+      else
+        take(slicer).drop(key)
+      end
+    end
+
+    # remove selected rows to create remainer DataFrame
     def remove(*args, &block)
       remover = args
       if block
         raise DataFrameArgumentError, 'Must not specify both arguments and block.' unless args.empty?
 
-        remover = instance_eval(&block)
+        remover = [instance_eval(&block)]
       end
-      remover = [remover].flatten
+      remover.flatten!
 
       raise DataFrameArgumentError, 'Empty dataframe' if empty?
       return self if remover.empty? || remover[0].nil?

data/lib/red_amber/data_frame_variable_operation.rb

@@ -9,12 +9,16 @@ module RedAmber
       if block
         raise DataFrameArgumentError, 'Must not specify both arguments and block.' unless args.empty?
 
-        picker = instance_eval(&block)
+        picker = [instance_eval(&block)]
       end
-      picker = [picker].flatten
+      picker.flatten!
       return DataFrame.new if picker.empty? || picker == [nil]
 
-      picker = keys_by_booleans(picker) if booleans?(picker)
+      key_vector = Vector.new(keys)
+      picker_vector = parse_to_vector(picker)
+
+      picker = key_vector.filter(*picker_vector).to_a if picker_vector.boolean?
+      picker = key_vector.take(*picker_vector).to_a if picker_vector.numeric?
 
       # DataFrame#[] creates a Vector with single key is specified.
       # DataFrame#pick creates a DataFrame with single key.
@@ -29,12 +33,22 @@ module RedAmber
       if block
         raise DataFrameArgumentError, 'Must not specify both arguments and block.' unless args.empty?
 
-        dropper = instance_eval(&block)
+        dropper = [instance_eval(&block)]
       end
-      dropper = [dropper].flatten
-      dropper = keys_by_booleans(dropper) if booleans?(dropper)
+      dropper.flatten!
+
+      key_vector = Vector.new(keys)
+      dropper_vector = parse_to_vector(dropper)
+
+      picker =
+        if dropper_vector.boolean?
+          key_vector.filter(*dropper_vector.primitive_invert).each.map(&:to_sym)
+        elsif dropper_vector.numeric?
+          keys - key_vector.take(*dropper_vector).each.map(&:to_sym)
+        else
+          keys - dropper
+        end
 
-      picker = keys - dropper
       return DataFrame.new if picker.empty?
 
       # DataFrame#[] creates a Vector with single key is specified.
@@ -91,10 +105,20 @@ module RedAmber
 
     def assign_update(*assigner, &block)
       if block
-        raise DataFrameArgumentError, 'Must not specify both arguments and a block' unless assigner.empty?
-
-        assigner = [instance_eval(&block)]
+        assigner_from_block = instance_eval(&block)
+        assigner =
+          if assigner.empty?
+            # block only
+            [assigner_from_block]
+          # If Ruby >= 3.0, one line pattern match can be used
+          # assigner_from_block in [Array, *]
+          elsif multiple_assigner?(assigner_from_block)
+            assigner.zip(assigner_from_block)
+          else
+            assigner.zip([assigner_from_block])
+          end
       end
+
       case assigner
       in [] | [nil] | [{}] | [[]]
         return self
@@ -113,6 +137,8 @@ module RedAmber
       updater = {}
       appender = {}
       key_array_pairs.each do |key, array|
+        raise DataFrameArgumentError, "Empty column data: #{key} => nil" if array.nil?
+
         if keys.include? key
           updater[key] = array
         else
@@ -153,7 +179,7 @@ module RedAmber
         data = updater[key]
         next unless data
 
-        raise DataFrameArgumentError, "Data size mismatch (#{data.size} != #{size})" if data.size != size
+        raise DataFrameArgumentError, "Data size mismatch (#{data.size} != #{size})" if data.nil? || data.size != size
 
         a = Arrow::Array.new(data.is_a?(Vector) ? data.to_a : data)
         fields[i] = Arrow::Field.new(key, a.value_data_type)
@@ -179,8 +205,13 @@ module RedAmber
       end
     end
 
-    def keys_by_booleans(booleans)
-      keys.select.with_index { |_, i| booleans[i] }
+    def multiple_assigner?(assigner)
+      case assigner
+      in [Vector, *] | [Array, *] | [Arrow::Array, *]
+        true
+      else
+        false
+      end
     end
   end
 end

data/lib/red_amber/vector.rb

@@ -122,7 +122,7 @@ module RedAmber
       return enum_for(:each) unless block_given?
 
       size.times do |i|
-        yield self[i]
+        yield data[i]
       end
     end
 

data/lib/red_amber/vector_functions.rb

@@ -34,13 +34,6 @@ module RedAmber
     end
     alias_method :std, :sd
 
-    # option(s) required
-    # - index
-
-    # Returns other than value
-    # - mode
-    # - tdigest
-
     # Return quantile
     #   0.5 quantile (median) is returned by default.
     #   Or return quantile for specified probability (prob).
@@ -88,8 +81,9 @@ module RedAmber
 
     # [Unary element-wise]: vector.func => vector
     unary_element_wise =
-      %i[abs array_sort_indices atan bit_wise_not ceil cos fill_null_backward fill_null_forward floor is_finite
-         is_inf is_nan is_null is_valid round round_to_multiple sign sin tan trunc unique]
+      %i[abs acos asin array_sort_indices atan bit_wise_not ceil cos fill_null_backward \
+         fill_null_forward floor is_finite is_inf is_nan is_null is_valid ln log10 log1p log2 \
+         round round_to_multiple sign sin tan trunc unique]
     unary_element_wise.each do |function|
       define_method(function) do |**options|
         datum = exec_func_unary(function, options)
@@ -129,16 +123,9 @@ module RedAmber
     end
     alias_method :not, :invert
 
-    # NaN support needed
-    # - acos asin ln log10 log1p log2
-
-    # Functions with numerical range check
-    # - abs_checked acos_checked asin_checked cos_checked ln_checked
-    #   log10_checked log1p_checked log2_checked sin_checked tan_checked
-
     # [Binary element-wise]: vector.func(other) => vector
     binary_element_wise =
-      %i[atan2 and_not and_not_kleene bit_wise_and bit_wise_or bit_wise_xor]
+      %i[atan2 and_not and_not_kleene bit_wise_and bit_wise_or bit_wise_xor logb]
     binary_element_wise.each do |function|
       define_method(function) do |other, **options|
         datum = exec_func_binary(function, other, options)
@@ -162,13 +149,6 @@ module RedAmber
       end
     end
 
-    # NaN support needed
-    # - logb
-
-    # Functions with numerical range check
-    # - add_checked divide_checked logb_checked multiply_checked power_checked subtract_checked
-    #   shift_left_checked shift_right_checked
-
     # [Binary element-wise with operator]: vector.func(other) => vector
     binary_element_wise_op = {
       add: '+',
@@ -216,6 +196,23 @@ module RedAmber
       [Vector.new(Array(other) * size), self]
     end
 
+    # < Not implimented yet > ---
+
+    # option(s) required
+    # - index
+
+    # Returns other than value
+    # - mode
+    # - tdigest
+
+    # Functions with numerical range check (unary)
+    # - abs_checked acos_checked asin_checked cos_checked ln_checked
+    #   log10_checked log1p_checked log2_checked sin_checked tan_checked
+
+    # Functions with numerical range check (binary)
+    # - add_checked divide_checked logb_checked multiply_checked power_checked subtract_checked
+    #   shift_left_checked shift_right_checked
+
     # (array functions)
     # dictionary_encode,
     # partition_nth_indices,

data/lib/red_amber/vector_selectable.rb

@@ -82,16 +82,17 @@ module RedAmber
 
     #   @param values [Array, Arrow::Array, Vector]
     def is_in(*values)
-      values.flatten!
+      self_data = chunked? ? data.pack : data
+
       array =
-        case values[0]
-        when Vector
-          values[0].data
-        when Arrow::Array
-          values[0]
+        case values
+        in [Vector] | [Arrow::Array] | [Arrow::ChunkedArray]
+          values[0].to_a
+        else
+          Array(values).flatten
         end
-      array ||= data.class.new(values)
-      Vector.new(data.is_in(array))
+
+      Vector.new(self_data.is_in(array))
     end
 
     # Arrow's support required

data/lib/red_amber/version.rb

@@ -1,5 +1,5 @@
 # frozen_string_literal: true
 
 module RedAmber
-  VERSION = '0.2.0'
+  VERSION = '0.2.1'
 end

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: red_amber
 version: !ruby/object:Gem::Version
-  version: 0.2.0
+  version: 0.2.1
 platform: ruby
 authors:
 - Hirokazu SUZUKI (heronshoes)
 autorequire:
 bindir: exe
 cert_chain: []
-date: 2022-08-15 00:00:00.000000000 Z
+date: 2022-09-07 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: red-arrow