Bump Polars 0.37 (#861)

datasets/iris.csv

@@ -149,4 +149,3 @@ sepal_length,sepal_width,petal_length,petal_width,species
 6.5,3.0,5.2,2.0,Iris-virginica
 6.2,3.4,5.4,2.3,Iris-virginica
 5.9,3.0,5.1,1.8,Iris-virginica
-

lib/explorer/backend/lazy_series.ex

@@ -606,7 +606,12 @@ defmodule Explorer.Backend.LazySeries do
 
   @impl true
   def format(list) do
-    series_list = Enum.map(list, &series_or_lazy_series!/1)
+    series_list =
+      Enum.map(list, fn
+        s when is_binary(s) -> s
+        s -> series_or_lazy_series!(s)
+      end)
+
     data = new(:format, [series_list], :string, aggregations?(series_list))
 
     Backend.Series.new(data, :string)

lib/explorer/polars_backend/native.ex

@@ -314,7 +314,6 @@ defmodule Explorer.PolarsBackend.Native do
   def s_fill_missing_with_atom(_s, _value), do: err()
   def s_fill_missing_with_date(_s, _value), do: err()
   def s_fill_missing_with_datetime(_s, _value), do: err()
-  def s_format(_series_list), do: err()
   def s_greater(_s, _rhs), do: err()
   def s_greater_equal(_s, _rhs), do: err()
   def s_head(_s, _length), do: err()

lib/explorer/polars_backend/series.ex

@@ -132,10 +132,26 @@ defmodule Explorer.PolarsBackend.Series do
 
   @impl true
   def format(list) do
-    polars_series = for s <- list, do: s.data
+    {_, df_args, params} =
+      Enum.reduce(list, {0, [], []}, fn s, {counter, df_args, params} ->
+        if is_binary(s) or Kernel.is_nil(s) do
+          {counter, df_args, [s | params]}
+        else
+          counter = counter + 1
+          name = "#{counter}"
+          column = Explorer.Backend.LazySeries.new(:column, [name], :string)
+          {counter, [{name, s} | df_args], [column | params]}
+        end
+      end)
 
-    Shared.apply(:s_format, [polars_series])
-    |> Shared.create_series()
+    df = Explorer.PolarsBackend.DataFrame.from_series(df_args)
+    format_expr = Explorer.Backend.LazySeries.new(:format, [Enum.reverse(params)], :string)
+    out_dtypes = Map.put(df.dtypes, "result", :string)
+    out_names = ["result" | df.names]
+    out_df = %{df | dtypes: out_dtypes, names: out_names}
+
+    Explorer.PolarsBackend.DataFrame.mutate_with(df, out_df, [{"result", format_expr}])
+    |> Explorer.PolarsBackend.DataFrame.pull("result")
   end
 
   @impl true

lib/explorer/series.ex

@@ -1274,13 +1274,27 @@ defmodule Explorer.Series do
 
   """
   @doc type: :element_wise
-  def categorise(%Series{dtype: l_dtype} = series, %Series{dtype: dtype} = categories)
-      when K.and(K.in(l_dtype, [:string | @integer_types]), K.in(dtype, [:string, :category])),
+  def categorise(%Series{dtype: l_dtype} = series, %Series{dtype: :category} = categories)
+      when K.in(l_dtype, [:string | @integer_types]),
       do: apply_series(series, :categorise, [categories])
 
+  def categorise(%Series{dtype: l_dtype} = series, %Series{dtype: :string} = categories)
+      when K.in(l_dtype, [:string | @integer_types]) do
+    if nil_count(categories) != 0 do
+      raise(ArgumentError, "categories as strings cannot have nil values")
+    end
+
+    if count(categories) != n_distinct(categories) do
+      raise(ArgumentError, "categories as strings cannot have duplicated values")
+    end
+
+    categories = cast(categories, :category)
+    apply_series(series, :categorise, [categories])
+  end
+
   def categorise(%Series{dtype: l_dtype} = series, [head | _] = categories)
       when K.and(K.in(l_dtype, [:string | @integer_types]), is_binary(head)),
-      do: apply_series(series, :categorise, [from_list(categories, dtype: :string)])
+      do: categorise(series, from_list(categories, dtype: :string))
 
   # Slice and dice
 
@@ -2086,13 +2100,20 @@ defmodule Explorer.Series do
       iex> s1 = Explorer.Series.from_list([<<1>>, <<239, 191, 19>>], dtype: :binary)
       iex> s2 = Explorer.Series.from_list([<<3>>, <<4>>], dtype: :binary)
       iex> Explorer.Series.format([s1, s2])
-      ** (RuntimeError) Polars Error: invalid utf-8 sequence
+      ** (RuntimeError) Polars Error: invalid utf8
   """
   @doc type: :shape
   @spec format([Series.t() | String.t()]) :: Series.t()
   def format([_ | _] = list) do
     list = cast_to_string(list)
-    impl!(list).format(list)
+
+    if impl = impl!(list) do
+      impl.format(list)
+    else
+      [hd | rest] = list
+      s = Series.from_list([hd], dtype: :string)
+      impl!([s]).format([s | rest])
+    end
   end
 
   defp cast_to_string(list) do
@@ -2103,8 +2124,8 @@ defmodule Explorer.Series do
       %Series{} = s ->
         cast(s, :string)
 
-      value when is_binary(value) ->
-        from_list([value], dtype: :string)
+      value when K.or(is_binary(value), K.is_nil(value)) ->
+        value
 
       other ->
         raise ArgumentError,