Fix Series construction with dtype=str

doc/source/whatsnew/v0.23.0.txt

@@ -1035,6 +1035,7 @@ Reshaping
 - Bug in :class:`Series` constructor with ``Categorical`` where a ```ValueError`` is not raised when an index of different length is given (:issue:`19342`)
 - Bug in :meth:`DataFrame.astype` where column metadata is lost when converting to categorical or a dictionary of dtypes (:issue:`19920`)
 - Bug in :func:`cut` and :func:`qcut` where timezone information was dropped (:issue:`19872`)
+- Bug in :class:`Series` constructor with a ``dtype=str``, previously raised in some cases (:issue:`19853`)
 
 Other
 ^^^^^

pandas/core/series.py

@@ -4059,9 +4059,10 @@ def _try_cast(arr, take_fast_path):
     if issubclass(subarr.dtype.type, compat.string_types):
         # GH 16605
         # If not empty convert the data to dtype
-        if not isna(data).all():
-            data = np.array(data, dtype=dtype, copy=False)
-
-        subarr = np.array(data, dtype=object, copy=copy)
+        # GH 19853: If data is a scalar, subarr has already the result
+        if not np.isscalar(data):
+            if not np.all(isna(data)):
+                data = np.array(data, dtype=dtype, copy=False)
+            subarr = np.array(data, dtype=object, copy=copy)
 
     return subarr

pandas/tests/series/test_constructors.py

@@ -110,6 +110,11 @@ def test_constructor_empty(self, input_class):
             empty2 = Series(input_class(), index=lrange(10), dtype='float64')
             assert_series_equal(empty, empty2)
 
+            # GH 19853 : with empty string, index and dtype str
+            empty = Series('', dtype=str, index=range(3))
+            empty2 = Series('', index=range(3))
+            assert_series_equal(empty, empty2)
+
     @pytest.mark.parametrize('input_arg', [np.nan, float('nan')])
     def test_constructor_nan(self, input_arg):
         empty = Series(dtype='float64', index=lrange(10))