Add drop_first parameter to Series.to_dummies

py-polars/polars/series/series.py

@@ -2109,14 +2109,18 @@ def quantile(
         """
         return self._s.quantile(quantile, interpolation)
 
-    def to_dummies(self, separator: str = "_") -> DataFrame:
+    def to_dummies(
+        self, *, separator: str = "_", drop_first: bool = False
+    ) -> DataFrame:
         """
         Get dummy/indicator variables.
 
         Parameters
         ----------
         separator
             Separator/delimiter used when generating column names.
+        drop_first
+            Remove the first category from the variable being encoded.
 
         Examples
         --------
@@ -2132,8 +2136,20 @@ def to_dummies(self, separator: str = "_") -> DataFrame:
         │ 0   ┆ 1   ┆ 0   │
         │ 0   ┆ 0   ┆ 1   │
         └─────┴─────┴─────┘
-        """
-        return wrap_df(self._s.to_dummies(separator))
+
+        >>> s.to_dummies(drop_first=True)
+        shape: (3, 2)
+        ┌─────┬─────┐
+        │ a_2 ┆ a_3 │
+        │ --- ┆ --- │
+        │ u8  ┆ u8  │
+        ╞═════╪═════╡
+        │ 0   ┆ 0   │
+        │ 1   ┆ 0   │
+        │ 0   ┆ 1   │
+        └─────┴─────┘
+        """
+        return wrap_df(self._s.to_dummies(separator, drop_first))
 
     @overload
     def cut(

py-polars/tests/unit/series/test_series.py

@@ -1542,6 +1542,16 @@ def test_to_dummies() -> None:
     assert_frame_equal(result, expected)
 
 
+def test_to_dummies_drop_first() -> None:
+    s = pl.Series("a", [1, 2, 3])
+    result = s.to_dummies(drop_first=True)
+    expected = pl.DataFrame(
+        {"a_2": [0, 1, 0], "a_3": [0, 0, 1]},
+        schema={"a_2": pl.UInt8, "a_3": pl.UInt8},
+    )
+    assert_frame_equal(result, expected)
+
+
 def test_chunk_lengths() -> None:
     s = pl.Series("a", [1, 2, 2, 3])
     # this is a Series with one chunk, of length 4