Add dseries.struct.explode

python/dask_cudf/dask_cudf/accessors.py

@@ -37,6 +37,32 @@ def field(self, key):
             meta=self.d_series._meta._constructor([], dtype=typ),
         )
 
+    def explode(self):
+        """
+        Creates a dataframe view of the struct column, one column per field.
+
+        Returns
+        -------
+        DataFrame
+
+        Examples
+        --------
+        >>> import cudf, dask_cudf as dgd
+        >>> ds = dgd.from_cudf(cudf.Series(
+        ...     [{'a': 42, 'b': 'str1', 'c': [-1]},
+        ...      {'a': 0,  'b': 'str2', 'c': [400, 500]},
+        ...      {'a': 7,  'b': '',     'c': []}]), npartitions=2)
+        >>> ds.struct.explode().compute()
+            a     b           c
+        0  42  str1        [-1]
+        1   0  str2  [400, 500]
+        2   7                []
+        """
+        return self.d_series.map_partitions(
+            lambda s: s.struct.explode(),
+            meta=self.d_series._meta.struct.explode(),
+        )
+
 
 class ListMethods:
     def __init__(self, d_series):

python/dask_cudf/dask_cudf/tests/test_accessor.py

@@ -500,3 +500,18 @@ def test_dask_struct_field_Int_Error(data):
 
     with pytest.raises(IndexError):
         got.struct.field(1000).compute()
+
+
+@pytest.mark.parametrize(
+    "data",
+    [
+        [{}, {}, {}],
+        [{"a": 100, "b": "abc"}, {"a": 42, "b": "def"}, {"a": -87, "b": ""}],
+        [{"a": [1, 2, 3], "b": {"c": 101}}, {"a": [4, 5], "b": {"c": 102}}],
+    ],
+)
+def test_struct_explode(data):
+    expect = pd.DataFrame(data)
+    got = dgd.from_cudf(Series(data), 2).struct.explode().compute()
+
+    assert_eq(expect, got)