cimport is_list_like

Author: jreback

doc/source/user_guide/reshaping.rst

@@ -828,7 +828,7 @@ You can join this with the original to get an expanded ``DataFrame``.
 
 .. ipython:: python
 
-   df[['keys']]join(df['values'].explode())
+   df[['keys']].join(df['values'].explode())
 
 
 This routine will replace empty lists with ``np.nan`` and preserve scalar entries. The dtype of the resulting ``Series`` is always ``object``.

pandas/_libs/lib.pxd

@@ -0,0 +1 @@
+cdef bint c_is_list_like(object, bint)

pandas/_libs/lib.pyx

@@ -887,7 +887,7 @@ def is_period(val: object) -> bool:
     return util.is_period_object(val)
 
 
-cpdef is_list_like(obj: object, allow_sets: bool = True):
+def is_list_like(obj: object, allow_sets: bool = True):
     """
     Check if the object is list-like.
 
@@ -926,15 +926,19 @@ cpdef is_list_like(obj: object, allow_sets: bool = True):
     >>> is_list_like(np.array(2)))
     False
     """
+    return c_is_list_like(obj, allow_sets)
 
+
+
+cdef inline bint c_is_list_like(object obj, bint allow_sets):
     return (
         isinstance(obj, abc.Iterable)
         and
         # we do not count strings/unicode/bytes as list-like
         not isinstance(obj, (str, bytes))
         and
         # exclude zero-dimensional numpy arrays, effectively scalars
-        not (isinstance(obj, np.ndarray) and obj.ndim == 0)
+        not (util.is_array(obj) and obj.ndim == 0)
         and
         # exclude sets if allow_sets is False
         not (allow_sets is False and isinstance(obj, abc.Set))

pandas/_libs/reshape.pyx

@@ -4,7 +4,7 @@ from cython import Py_ssize_t
 from numpy cimport (int8_t, int16_t, int32_t, int64_t, uint8_t, uint16_t,
                     uint32_t, uint64_t, float32_t, float64_t, ndarray)
 import numpy
-from pandas._libs.lib import is_list_like
+from pandas._libs.lib cimport c_is_list_like
 
 
 ctypedef fused reshape_t:
@@ -121,7 +121,7 @@ def explode(ndarray[object] values):
     counts = numpy.zeros(n, dtype='int64')
     for i in range(n):
         v = values[i]
-        if is_list_like(v):
+        if c_is_list_like(v, False):
             if len(v):
                 counts[i] += len(v)
             else:
@@ -135,7 +135,7 @@ def explode(ndarray[object] values):
     for i in range(n):
         v = values[i]
 
-        if is_list_like(v):
+        if c_is_list_like(v, False):
             if len(v):
                 for j in range(len(v)):
                     result[count] = v[j]

pandas/core/series.py

@@ -27,6 +27,7 @@
     is_datetime64_dtype,
     is_datetimelike,
     is_dict_like,
+    is_object_dtype,
     is_extension_array_dtype,
     is_extension_type,
     is_hashable,
@@ -3635,7 +3636,7 @@ def reorder_levels(self, order):
         result.index = result.index.reorder_levels(order)
         return result
 
-    def explode(self) -> 'Series':
+    def explode(self) -> "Series":
         """
         Create new Series expanding a list-like column.
 
@@ -3648,7 +3649,7 @@ def explode(self) -> 'Series':
         See Also
         --------
         Series.str.split: Split string values on specified separator.
-        Series.unstakc: Unstack, a.k.a. pivot, Series with MultiIndex to produce DataFrame.
+        Series.unstack: Unstack, a.k.a. pivot, Series with MultiIndex to produce DataFrame.
 
         Examples
         --------
@@ -3680,7 +3681,7 @@ def explode(self) -> 'Series':
         Scalars will be returned unchanged.
         Empty list-likes will result in a np.nan for that row.
         """
-        if not len(self):
+        if not len(self) or not is_object_dtype(self):
             return self.copy()
 
         values, counts = reshape.explode(np.asarray(self.array))

pandas/tests/series/test_explode.py

@@ -1,3 +1,4 @@
+import pytest
 import numpy as np
 
 import pandas as pd
@@ -68,6 +69,14 @@ def test_invert_array():
     tm.assert_series_equal(result, df["a"].rename())
 
 
+@pytest.mark.parametrize(
+    "s", [pd.Series([1, 2, 3]), pd.Series(pd.date_range("2019", periods=3, tz="UTC"))]
+)
+def non_object_dtype(s):
+    result = s.explode()
+    tm.assert_series_equal(result, s)
+
+
 def test_typical_usecase():
 
     df = pd.DataFrame([{"var1": "a,b,c", "var2": 1}, {"var1": "d,e,f", "var2": 2}])