polars-df 0.5.0 → 0.6.0

data/lib/polars/lazy_functions.rb

@@ -14,7 +14,7 @@ module Polars
 
       if name.is_a?(DataType)
         Utils.wrap_expr(_dtype_cols([name]))
-      elsif name.is_a?(Array)
+      elsif name.is_a?(::Array)
         if name.length == 0 || Utils.strlike?(name[0])
           name = name.map { |v| v.is_a?(Symbol) ? v.to_s : v }
           Utils.wrap_expr(RbExpr.cols(name))
@@ -36,7 +36,7 @@ module Polars
     # @example A horizontal rank computation by taking the elements of a list
     #   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)
@@ -156,9 +156,8 @@ module Polars
         column.sum
       elsif Utils.strlike?(column)
         col(column.to_s).sum
-      elsif column.is_a?(Array)
+      elsif column.is_a?(::Array)
         exprs = Utils.selection_to_rbexpr_list(column)
-        # TODO
         Utils.wrap_expr(_sum_exprs(exprs))
       else
         fold(lit(0).cast(:u32), ->(a, b) { a + b }, column).alias("sum")
@@ -283,18 +282,33 @@ module Polars
     # Return an expression representing a literal value.
     #
     # @return [Expr]
-    def lit(value)
-      if value.is_a?(Polars::Series)
+    def lit(value, dtype: nil, allow_object: nil)
+      if value.is_a?(::Time) || value.is_a?(::DateTime)
+        time_unit = dtype&.time_unit || "ns"
+        time_zone = dtype.&time_zone
+        e = lit(Utils._datetime_to_pl_timestamp(value, time_unit)).cast(Datetime.new(time_unit))
+        if time_zone
+          return e.dt.replace_time_zone(time_zone.to_s)
+        else
+          return e
+        end
+      elsif value.is_a?(::Date)
+        return lit(::Time.utc(value.year, value.month, value.day)).cast(Date)
+      elsif value.is_a?(Polars::Series)
         name = value.name
         value = value._s
-        e = Utils.wrap_expr(RbExpr.lit(value))
+        e = Utils.wrap_expr(RbExpr.lit(value, allow_object))
         if name == ""
           return e
         end
         return e.alias(name)
+      elsif (defined?(Numo::NArray) && value.is_a?(Numo::NArray)) || value.is_a?(::Array)
+        return lit(Series.new("", value))
+      elsif dtype
+        return Utils.wrap_expr(RbExpr.lit(value, allow_object)).cast(dtype)
       end
 
-      Utils.wrap_expr(RbExpr.lit(value))
+      Utils.wrap_expr(RbExpr.lit(value, allow_object))
     end
 
     # Cumulatively sum values in a column/Series, or horizontally across list of columns/expressions.
@@ -625,23 +639,42 @@ module Polars
     # @return [Expr, Series]
     #
     # @example
-    #   df.lazy.filter(Polars.col("foo") < Polars.arange(0, 100)).collect
+    #   Polars.arange(0, 3, eager: true)
+    #   # =>
+    #   # shape: (3,)
+    #   # Series: 'arange' [i64]
+    #   # [
+    #   #         0
+    #   #         1
+    #   #         2
+    #   # ]
+    #
+    # @example
+    #   df = Polars::DataFrame.new({"a" => [1, 2], "b" => [3, 4]})
+    #   df.select(Polars.arange(Polars.col("a"), Polars.col("b")))
+    #   # =>
+    #   # shape: (2, 1)
+    #   # ┌───────────┐
+    #   # │ arange    │
+    #   # │ ---       │
+    #   # │ list[i64] │
+    #   # ╞═══════════╡
+    #   # │ [1, 2]    │
+    #   # │ [2, 3]    │
+    #   # └───────────┘
     def arange(low, high, step: 1, eager: false, dtype: nil)
       low = Utils.expr_to_lit_or_expr(low, str_to_lit: false)
       high = Utils.expr_to_lit_or_expr(high, str_to_lit: false)
       range_expr = Utils.wrap_expr(RbExpr.arange(low._rbexpr, high._rbexpr, step))
 
-      if !dtype.nil? && dtype != "i64"
+      if !dtype.nil? && !["i64", Int64].include?(dtype)
         range_expr = range_expr.cast(dtype)
       end
 
       if !eager
         range_expr
       else
-        DataFrame.new
-          .select(range_expr)
-          .to_series
-          .rename("arange", in_place: true)
+        DataFrame.new.select(range_expr.alias("arange")).to_series
       end
     end
 
@@ -658,7 +691,7 @@ module Polars
     #
     # @return [Expr]
     def arg_sort_by(exprs, reverse: false)
-      if !exprs.is_a?(Array)
+      if !exprs.is_a?(::Array)
         exprs = [exprs]
       end
       if reverse == true || reverse == false
@@ -997,19 +1030,24 @@ module Polars
     #   Only used in `eager` mode. As expression, use `alias`.
     #
     # @return [Expr]
-    def repeat(value, n, eager: false, name: nil)
+    def repeat(value, n, dtype: nil, eager: false, name: nil)
+      if !name.nil?
+        warn "the `name` argument is deprecated. Use the `alias` method instead."
+      end
+
+      if n.is_a?(Integer)
+        n = lit(n)
+      end
+
+      value = Utils.parse_as_expression(value, str_as_lit: true)
+      expr = Utils.wrap_expr(RbExpr.repeat(value, n._rbexpr, dtype))
+      if !name.nil?
+        expr = expr.alias(name)
+      end
       if eager
-        if name.nil?
-          name = ""
-        end
-        dtype = py_type_to_dtype(type(value))
-        Series._repeat(name, value, n, dtype)
-      else
-        if n.is_a?(Integer)
-          n = lit(n)
-        end
-        Utils.wrap_expr(RbExpr.repeat(value, n._rbexpr))
+        return select(expr).to_series
       end
+      expr
     end
 
     # Return indices where `condition` evaluates `true`.
@@ -1124,13 +1162,11 @@ module Polars
       end
 
       if unit == "d"
-        expr = column.cast(:date)
+        expr = column.cast(Date)
       elsif unit == "s"
-        raise Todo
-        # expr = (column.cast(:i64) * 1_000_000).cast(Datetime("us"))
+        expr = (column.cast(Int64) * 1_000_000).cast(Datetime.new("us"))
       elsif Utils::DTYPE_TEMPORAL_UNITS.include?(unit)
-        raise Todo
-        # expr = column.cast(Datetime(unit))
+        expr = column.cast(Datetime.new(unit))
       else
         raise ArgumentError, "'unit' must be one of {{'ns', 'us', 'ms', 's', 'd'}}, got '#{unit}'."
       end

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)
     #   # ┌─────┐
@@ -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)
     #   # ┌────────┐
@@ -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)
     #   # ┌────────┐
@@ -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)
     #   # ┌────────┐
@@ -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)
     #   # ┌────────┐
@@ -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)
     #   # ┌───────────┐
@@ -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)
     #   # ┌───────────┐
@@ -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)
     #   # ┌───────────┐
@@ -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)
     #   # ┌──────┐
@@ -254,7 +254,7 @@ module Polars
     #   # │ 1    │
     #   # └──────┘
     def get(index)
-      index = Utils.expr_to_lit_or_expr(index, str_to_lit: false)._rbexpr
+      index = Utils.parse_as_expression(index)
       Utils.wrap_expr(_rbexpr.list_get(index))
     end
 
@@ -280,7 +280,7 @@ module Polars
     #
     # @return [Expr]
     def take(index, null_on_oob: false)
-      if index.is_a?(Array)
+      if index.is_a?(::Array)
         index = Series.new(index)
       end
       index = Utils.expr_to_lit_or_expr(index, str_to_lit: false)._rbexpr
@@ -293,7 +293,7 @@ module Polars
     #
     # @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)
     #   # ┌──────┐
@@ -315,7 +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)
     #   # ┌──────┐
@@ -340,7 +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)
     #   # ┌───────┐
@@ -367,7 +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)
     #   # ┌───────┐
@@ -392,7 +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)
     #   # ┌─────┐
@@ -417,7 +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)
     #   # ┌─────┐
@@ -443,7 +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]]
@@ -464,7 +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]]
@@ -488,7 +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]]
@@ -511,7 +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]]
@@ -532,7 +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]]
@@ -554,7 +554,7 @@ module Polars
     #
     # @example
     #   df = Polars::DataFrame.new({"listcol" => [[0], [1], [1, 2, 3, 2], [1, 2, 1], [4, 4]]})
-    #   df.select(Polars.col("listcol").arr.count_match(2).alias("number_of_twos"))
+    #   df.select(Polars.col("listcol").list.count_match(2).alias("number_of_twos"))
     #   # =>
     #   # shape: (5, 1)
     #   # ┌────────────────┐
@@ -584,7 +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)
     #   # ┌────────────┐
@@ -617,7 +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)

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)