polars-df 0.4.0-x86_64-linux → 0.6.0-x86_64-linux

data/lib/polars/list_expr.rb

@@ -15,7 +15,7 @@ module Polars
     #
     # @example
     #   df = Polars::DataFrame.new({"foo" => [1, 2], "bar" => [["a", "b"], ["c"]]})
-    #   df.select(Polars.col("bar").arr.lengths)
+    #   df.select(Polars.col("bar").list.lengths)
     #   # =>
     #   # shape: (2, 1)
     #   # ┌─────┐
@@ -27,7 +27,7 @@ module Polars
     #   # │ 1   │
     #   # └─────┘
     def lengths
-      Utils.wrap_expr(_rbexpr.arr_lengths)
+      Utils.wrap_expr(_rbexpr.list_lengths)
     end
 
     # Sum all the lists in the array.
@@ -36,7 +36,7 @@ module Polars
     #
     # @example
     #   df = Polars::DataFrame.new({"values" => [[1], [2, 3]]})
-    #   df.select(Polars.col("values").arr.sum)
+    #   df.select(Polars.col("values").list.sum)
     #   # =>
     #   # shape: (2, 1)
     #   # ┌────────┐
@@ -48,7 +48,7 @@ module Polars
     #   # │ 5      │
     #   # └────────┘
     def sum
-      Utils.wrap_expr(_rbexpr.lst_sum)
+      Utils.wrap_expr(_rbexpr.list_sum)
     end
 
     # Compute the max value of the lists in the array.
@@ -57,7 +57,7 @@ module Polars
     #
     # @example
     #   df = Polars::DataFrame.new({"values" => [[1], [2, 3]]})
-    #   df.select(Polars.col("values").arr.max)
+    #   df.select(Polars.col("values").list.max)
     #   # =>
     #   # shape: (2, 1)
     #   # ┌────────┐
@@ -69,7 +69,7 @@ module Polars
     #   # │ 3      │
     #   # └────────┘
     def max
-      Utils.wrap_expr(_rbexpr.lst_max)
+      Utils.wrap_expr(_rbexpr.list_max)
     end
 
     # Compute the min value of the lists in the array.
@@ -78,7 +78,7 @@ module Polars
     #
     # @example
     #   df = Polars::DataFrame.new({"values" => [[1], [2, 3]]})
-    #   df.select(Polars.col("values").arr.min)
+    #   df.select(Polars.col("values").list.min)
     #   # =>
     #   # shape: (2, 1)
     #   # ┌────────┐
@@ -90,7 +90,7 @@ module Polars
     #   # │ 2      │
     #   # └────────┘
     def min
-      Utils.wrap_expr(_rbexpr.lst_min)
+      Utils.wrap_expr(_rbexpr.list_min)
     end
 
     # Compute the mean value of the lists in the array.
@@ -99,7 +99,7 @@ module Polars
     #
     # @example
     #   df = Polars::DataFrame.new({"values" => [[1], [2, 3]]})
-    #   df.select(Polars.col("values").arr.mean)
+    #   df.select(Polars.col("values").list.mean)
     #   # =>
     #   # shape: (2, 1)
     #   # ┌────────┐
@@ -111,7 +111,7 @@ module Polars
     #   # │ 2.5    │
     #   # └────────┘
     def mean
-      Utils.wrap_expr(_rbexpr.lst_mean)
+      Utils.wrap_expr(_rbexpr.list_mean)
     end
 
     # Sort the arrays in the list.
@@ -124,7 +124,7 @@ module Polars
     #       "a" => [[3, 2, 1], [9, 1, 2]]
     #     }
     #   )
-    #   df.select(Polars.col("a").arr.sort)
+    #   df.select(Polars.col("a").list.sort)
     #   # =>
     #   # shape: (2, 1)
     #   # ┌───────────┐
@@ -136,7 +136,7 @@ module Polars
     #   # │ [1, 2, 9] │
     #   # └───────────┘
     def sort(reverse: false)
-      Utils.wrap_expr(_rbexpr.lst_sort(reverse))
+      Utils.wrap_expr(_rbexpr.list_sort(reverse))
     end
 
     # Reverse the arrays in the list.
@@ -149,7 +149,7 @@ module Polars
     #       "a" => [[3, 2, 1], [9, 1, 2]]
     #     }
     #   )
-    #   df.select(Polars.col("a").arr.reverse)
+    #   df.select(Polars.col("a").list.reverse)
     #   # =>
     #   # shape: (2, 1)
     #   # ┌───────────┐
@@ -161,7 +161,7 @@ module Polars
     #   # │ [2, 1, 9] │
     #   # └───────────┘
     def reverse
-      Utils.wrap_expr(_rbexpr.lst_reverse)
+      Utils.wrap_expr(_rbexpr.list_reverse)
     end
 
     # Get the unique/distinct values in the list.
@@ -174,7 +174,7 @@ module Polars
     #       "a" => [[1, 1, 2]]
     #     }
     #   )
-    #   df.select(Polars.col("a").arr.unique)
+    #   df.select(Polars.col("a").list.unique)
     #   # =>
     #   # shape: (1, 1)
     #   # ┌───────────┐
@@ -184,8 +184,8 @@ module Polars
     #   # ╞═══════════╡
     #   # │ [1, 2]    │
     #   # └───────────┘
-    def unique
-      Utils.wrap_expr(_rbexpr.lst_unique)
+    def unique(maintain_order: false)
+      Utils.wrap_expr(_rbexpr.list_unique(maintain_order))
     end
 
     # Concat the arrays in a Series dtype List in linear time.
@@ -202,7 +202,7 @@ module Polars
     #       "b" => [["b", "c"], ["y", "z"]]
     #     }
     #   )
-    #   df.select(Polars.col("a").arr.concat("b"))
+    #   df.select(Polars.col("a").list.concat("b"))
     #   # =>
     #   # shape: (2, 1)
     #   # ┌─────────────────┐
@@ -214,11 +214,11 @@ module Polars
     #   # │ ["x", "y", "z"] │
     #   # └─────────────────┘
     def concat(other)
-      if other.is_a?(Array) && ![Expr, String, Series].any? { |c| other[0].is_a?(c) }
+      if other.is_a?(::Array) && ![Expr, String, Series].any? { |c| other[0].is_a?(c) }
         return concat(Series.new([other]))
       end
 
-      if !other.is_a?(Array)
+      if !other.is_a?(::Array)
         other_list = [other]
       else
         other_list = other.dup
@@ -241,7 +241,7 @@ module Polars
     #
     # @example
     #   df = Polars::DataFrame.new({"foo" => [[3, 2, 1], [], [1, 2]]})
-    #   df.select(Polars.col("foo").arr.get(0))
+    #   df.select(Polars.col("foo").list.get(0))
     #   # =>
     #   # shape: (3, 1)
     #   # ┌──────┐
@@ -254,8 +254,8 @@ module Polars
     #   # │ 1    │
     #   # └──────┘
     def get(index)
-      index = Utils.expr_to_lit_or_expr(index, str_to_lit: false)._rbexpr
-      Utils.wrap_expr(_rbexpr.lst_get(index))
+      index = Utils.parse_as_expression(index)
+      Utils.wrap_expr(_rbexpr.list_get(index))
     end
 
     # Get the value by index in the sublists.
@@ -265,13 +265,35 @@ module Polars
       get(item)
     end
 
+    # Take sublists by multiple indices.
+    #
+    # The indices may be defined in a single column, or by sublists in another
+    # column of dtype `List`.
+    #
+    # @param index [Object]
+    #   Indices to return per sublist
+    # @param null_on_oob [Boolean]
+    #   Behavior if an index is out of bounds:
+    #   True -> set as null
+    #   False -> raise an error
+    #   Note that defaulting to raising an error is much cheaper
+    #
+    # @return [Expr]
+    def take(index, null_on_oob: false)
+      if index.is_a?(::Array)
+        index = Series.new(index)
+      end
+      index = Utils.expr_to_lit_or_expr(index, str_to_lit: false)._rbexpr
+      Utils.wrap_expr(_rbexpr.list_take(index, null_on_oob))
+    end
+
     # Get the first value of the sublists.
     #
     # @return [Expr]
     #
     # @example
     #   df = Polars::DataFrame.new({"foo" => [[3, 2, 1], [], [1, 2]]})
-    #   df.select(Polars.col("foo").arr.first)
+    #   df.select(Polars.col("foo").list.first)
     #   # =>
     #   # shape: (3, 1)
     #   # ┌──────┐
@@ -293,7 +315,7 @@ module Polars
     #
     # @example
     #   df = Polars::DataFrame.new({"foo" => [[3, 2, 1], [], [1, 2]]})
-    #   df.select(Polars.col("foo").arr.last)
+    #   df.select(Polars.col("foo").list.last)
     #   # =>
     #   # shape: (3, 1)
     #   # ┌──────┐
@@ -318,7 +340,7 @@ module Polars
     #
     # @example
     #   df = Polars::DataFrame.new({"foo" => [[3, 2, 1], [], [1, 2]]})
-    #   df.select(Polars.col("foo").arr.contains(1))
+    #   df.select(Polars.col("foo").list.contains(1))
     #   # =>
     #   # shape: (3, 1)
     #   # ┌───────┐
@@ -331,7 +353,7 @@ module Polars
     #   # │ true  │
     #   # └───────┘
     def contains(item)
-      Utils.wrap_expr(_rbexpr.arr_contains(Utils.expr_to_lit_or_expr(item)._rbexpr))
+      Utils.wrap_expr(_rbexpr.list_contains(Utils.expr_to_lit_or_expr(item)._rbexpr))
     end
 
     # Join all string items in a sublist and place a separator between them.
@@ -345,7 +367,7 @@ module Polars
     #
     # @example
     #   df = Polars::DataFrame.new({"s" => [["a", "b", "c"], ["x", "y"]]})
-    #   df.select(Polars.col("s").arr.join(" "))
+    #   df.select(Polars.col("s").list.join(" "))
     #   # =>
     #   # shape: (2, 1)
     #   # ┌───────┐
@@ -357,7 +379,7 @@ module Polars
     #   # │ x y   │
     #   # └───────┘
     def join(separator)
-      Utils.wrap_expr(_rbexpr.lst_join(separator))
+      Utils.wrap_expr(_rbexpr.list_join(separator))
     end
 
     # Retrieve the index of the minimal value in every sublist.
@@ -370,7 +392,7 @@ module Polars
     #       "a" => [[1, 2], [2, 1]]
     #     }
     #   )
-    #   df.select(Polars.col("a").arr.arg_min)
+    #   df.select(Polars.col("a").list.arg_min)
     #   # =>
     #   # shape: (2, 1)
     #   # ┌─────┐
@@ -382,7 +404,7 @@ module Polars
     #   # │ 1   │
     #   # └─────┘
     def arg_min
-      Utils.wrap_expr(_rbexpr.lst_arg_min)
+      Utils.wrap_expr(_rbexpr.list_arg_min)
     end
 
     # Retrieve the index of the maximum value in every sublist.
@@ -395,7 +417,7 @@ module Polars
     #       "a" => [[1, 2], [2, 1]]
     #     }
     #   )
-    #   df.select(Polars.col("a").arr.arg_max)
+    #   df.select(Polars.col("a").list.arg_max)
     #   # =>
     #   # shape: (2, 1)
     #   # ┌─────┐
@@ -407,7 +429,7 @@ module Polars
     #   # │ 0   │
     #   # └─────┘
     def arg_max
-      Utils.wrap_expr(_rbexpr.lst_arg_max)
+      Utils.wrap_expr(_rbexpr.list_arg_max)
     end
 
     # Calculate the n-th discrete difference of every sublist.
@@ -421,7 +443,7 @@ module Polars
     #
     # @example
     #   s = Polars::Series.new("a", [[1, 2, 3, 4], [10, 2, 1]])
-    #   s.arr.diff
+    #   s.list.diff
     #   # =>
     #   # shape: (2,)
     #   # Series: 'a' [list[i64]]
@@ -430,7 +452,7 @@ module Polars
     #   #         [null, -8, -1]
     #   # ]
     def diff(n: 1, null_behavior: "ignore")
-      Utils.wrap_expr(_rbexpr.lst_diff(n, null_behavior))
+      Utils.wrap_expr(_rbexpr.list_diff(n, null_behavior))
     end
 
     # Shift values by the given period.
@@ -442,7 +464,7 @@ module Polars
     #
     # @example
     #   s = Polars::Series.new("a", [[1, 2, 3, 4], [10, 2, 1]])
-    #   s.arr.shift
+    #   s.list.shift
     #   # =>
     #   # shape: (2,)
     #   # Series: 'a' [list[i64]]
@@ -451,7 +473,7 @@ module Polars
     #   #         [null, 10, 2]
     #   # ]
     def shift(periods = 1)
-      Utils.wrap_expr(_rbexpr.lst_shift(periods))
+      Utils.wrap_expr(_rbexpr.list_shift(periods))
     end
 
     # Slice every sublist.
@@ -466,7 +488,7 @@ module Polars
     #
     # @example
     #   s = Polars::Series.new("a", [[1, 2, 3, 4], [10, 2, 1]])
-    #   s.arr.slice(1, 2)
+    #   s.list.slice(1, 2)
     #   # =>
     #   # shape: (2,)
     #   # Series: 'a' [list[i64]]
@@ -477,7 +499,7 @@ module Polars
     def slice(offset, length = nil)
       offset = Utils.expr_to_lit_or_expr(offset, str_to_lit: false)._rbexpr
       length = Utils.expr_to_lit_or_expr(length, str_to_lit: false)._rbexpr
-      Utils.wrap_expr(_rbexpr.lst_slice(offset, length))
+      Utils.wrap_expr(_rbexpr.list_slice(offset, length))
     end
 
     # Slice the first `n` values of every sublist.
@@ -489,7 +511,7 @@ module Polars
     #
     # @example
     #   s = Polars::Series.new("a", [[1, 2, 3, 4], [10, 2, 1]])
-    #   s.arr.head(2)
+    #   s.list.head(2)
     #   # =>
     #   # shape: (2,)
     #   # Series: 'a' [list[i64]]
@@ -510,7 +532,7 @@ module Polars
     #
     # @example
     #   s = Polars::Series.new("a", [[1, 2, 3, 4], [10, 2, 1]])
-    #   s.arr.tail(2)
+    #   s.list.tail(2)
     #   # =>
     #   # shape: (2,)
     #   # Series: 'a' [list[i64]]
@@ -523,6 +545,33 @@ module Polars
       slice(offset, n)
     end
 
+    # Count how often the value produced by ``element`` occurs.
+    #
+    # @param element [Expr]
+    #   An expression that produces a single value
+    #
+    # @return [Expr]
+    #
+    # @example
+    #   df = Polars::DataFrame.new({"listcol" => [[0], [1], [1, 2, 3, 2], [1, 2, 1], [4, 4]]})
+    #   df.select(Polars.col("listcol").list.count_match(2).alias("number_of_twos"))
+    #   # =>
+    #   # shape: (5, 1)
+    #   # ┌────────────────┐
+    #   # │ number_of_twos │
+    #   # │ ---            │
+    #   # │ u32            │
+    #   # ╞════════════════╡
+    #   # │ 0              │
+    #   # │ 0              │
+    #   # │ 2              │
+    #   # │ 1              │
+    #   # │ 0              │
+    #   # └────────────────┘
+    def count_match(element)
+      Utils.wrap_expr(_rbexpr.list_count_match(Utils.expr_to_lit_or_expr(element)._rbexpr))
+    end
+
     # Convert the series of type `List` to a series of type `Struct`.
     #
     # @param n_field_strategy ["first_non_null", "max_width"]
@@ -535,7 +584,7 @@ module Polars
     #
     # @example
     #   df = Polars::DataFrame.new({"a" => [[1, 2, 3], [1, 2]]})
-    #   df.select([Polars.col("a").arr.to_struct])
+    #   df.select([Polars.col("a").list.to_struct])
     #   # =>
     #   # shape: (2, 1)
     #   # ┌────────────┐
@@ -548,7 +597,7 @@ module Polars
     #   # └────────────┘
     def to_struct(n_field_strategy: "first_non_null", name_generator: nil)
       raise Todo if name_generator
-      Utils.wrap_expr(_rbexpr.lst_to_struct(n_field_strategy, name_generator, 0))
+      Utils.wrap_expr(_rbexpr.list_to_struct(n_field_strategy, name_generator, 0))
     end
 
     # Run any polars expression against the lists' elements.
@@ -568,7 +617,7 @@ module Polars
     # @example
     #   df = Polars::DataFrame.new({"a" => [1, 8, 3], "b" => [4, 5, 2]})
     #   df.with_column(
-    #     Polars.concat_list(["a", "b"]).arr.eval(Polars.element.rank).alias("rank")
+    #     Polars.concat_list(["a", "b"]).list.eval(Polars.element.rank).alias("rank")
     #   )
     #   # =>
     #   # shape: (3, 3)
@@ -582,7 +631,7 @@ module Polars
     #   # │ 3   ┆ 2   ┆ [2.0, 1.0] │
     #   # └─────┴─────┴────────────┘
     def eval(expr, parallel: false)
-       Utils.wrap_expr(_rbexpr.lst_eval(expr._rbexpr, parallel))
+      Utils.wrap_expr(_rbexpr.list_eval(expr._rbexpr, parallel))
     end
   end
 end

data/lib/polars/list_name_space.rb

@@ -1,9 +1,9 @@
 module Polars
-  # Series.arr namespace.
+  # Series.list namespace.
   class ListNameSpace
     include ExprDispatch
 
-    self._accessor = "arr"
+    self._accessor = "list"
 
     # @private
     def initialize(series)
@@ -16,7 +16,7 @@ module Polars
     #
     # @example
     #   s = Polars::Series.new([[1, 2, 3], [5]])
-    #   s.arr.lengths
+    #   s.list.lengths
     #   # =>
     #   # shape: (2,)
     #   # Series: '' [u32]
@@ -119,13 +119,13 @@ module Polars
     #
     # @example
     #   s = Polars::Series.new([["foo", "bar"], ["hello", "world"]])
-    #   s.arr.join("-")
+    #   s.list.join("-")
     #   # =>
     #   # shape: (2,)
     #   # Series: '' [str]
     #   # [
-    #   #     "foo-bar"
-    #   #     "hello-world"
+    #   #         "foo-bar"
+    #   #         "hello-world"
     #   # ]
     def join(separator)
       super
@@ -180,7 +180,7 @@ module Polars
     #
     # @example
     #   s = Polars::Series.new("a", [[1, 2, 3, 4], [10, 2, 1]])
-    #   s.arr.diff
+    #   s.list.diff
     #   # =>
     #   # shape: (2,)
     #   # Series: 'a' [list[i64]]
@@ -201,7 +201,7 @@ module Polars
     #
     # @example
     #   s = Polars::Series.new("a", [[1, 2, 3, 4], [10, 2, 1]])
-    #   s.arr.shift
+    #   s.list.shift
     #   # =>
     #   # shape: (2,)
     #   # Series: 'a' [list[i64]]
@@ -225,7 +225,7 @@ module Polars
     #
     # @example
     #   s = Polars::Series.new("a", [[1, 2, 3, 4], [10, 2, 1]])
-    #   s.arr.slice(1, 2)
+    #   s.list.slice(1, 2)
     #   # =>
     #   # shape: (2,)
     #   # Series: 'a' [list[i64]]
@@ -246,7 +246,7 @@ module Polars
     #
     # @example
     #   s = Polars::Series.new("a", [[1, 2, 3, 4], [10, 2, 1]])
-    #   s.arr.head(2)
+    #   s.list.head(2)
     #   # =>
     #   # shape: (2,)
     #   # Series: 'a' [list[i64]]
@@ -267,7 +267,7 @@ module Polars
     #
     # @example
     #   s = Polars::Series.new("a", [[1, 2, 3, 4], [10, 2, 1]])
-    #   s.arr.tail(2)
+    #   s.list.tail(2)
     #   # =>
     #   # shape: (2,)
     #   # Series: 'a' [list[i64]]
@@ -291,7 +291,7 @@ module Polars
     #
     # @example
     #   df = Polars::DataFrame.new({"a" => [[1, 2, 3], [1, 2]]})
-    #   df.select([Polars.col("a").arr.to_struct])
+    #   df.select([Polars.col("a").list.to_struct])
     #   # =>
     #   # shape: (2, 1)
     #   # ┌────────────┐
@@ -323,7 +323,7 @@ module Polars
     # @example
     #   df = Polars::DataFrame.new({"a" => [1, 8, 3], "b" => [4, 5, 2]})
     #   df.with_column(
-    #     Polars.concat_list(["a", "b"]).arr.eval(Polars.element.rank).alias("rank")
+    #     Polars.concat_list(["a", "b"]).list.eval(Polars.element.rank).alias("rank")
     #   )
     #   # =>
     #   # shape: (3, 3)

data/lib/polars/rolling_group_by.rb

@@ -10,7 +10,8 @@ module Polars
       period,
       offset,
       closed,
-      by
+      by,
+      check_sorted
     )
       period = Utils._timedelta_to_pl_duration(period)
       offset = Utils._timedelta_to_pl_duration(offset)
@@ -21,12 +22,13 @@ module Polars
       @offset = offset
       @closed = closed
       @by = by
+      @check_sorted = check_sorted
     end
 
     def agg(aggs)
       @df.lazy
         .groupby_rolling(
-          index_column: @time_column, period: @period, offset: @offset, closed: @closed, by: @by
+          index_column: @time_column, period: @period, offset: @offset, closed: @closed, by: @by, check_sorted: @check_sorted
         )
         .agg(aggs)
         .collect(no_optimization: true, string_cache: false)