handle all-na Series/Index

Author: h-vetinari

pandas/core/strings.py

@@ -2268,7 +2268,8 @@ def cat(self, others=None, sep=None, na_rep=None, join=None):
             result = cat_core(all_cols, sep)
 
         if isinstance(self._orig, Index):
-            result = Index(result, name=self._orig.name)
+            # add dtype for case that result is all-NA
+            result = Index(result, dtype='object', name=self._orig.name)
         else:  # Series
             result = Series(result, index=data.index, name=self._orig.name)
         return result

pandas/tests/test_strings.py

@@ -367,13 +367,37 @@ def test_str_cat_align_mixed_inputs(self, join):
         with tm.assert_raises_regex(ValueError, rgx):
             s.str.cat([t, z], join=join)
 
-    def test_str_cat_raises(self):
+    @pytest.mark.parametrize('box', [Series, Index])
+    def test_str_cat_raises(self, box):
         # non-strings hiding behind object dtype
-        s = Series([1, 2, 3, 4], dtype='object')
+        s = box([1, 2, 3, 4], dtype='object')
         message = 'Can only use .str accessor with string values'
         with tm.assert_raises_regex(AttributeError, message):
             s.str
 
+    @pytest.mark.parametrize('box', [Series, Index])
+    @pytest.mark.parametrize('other', [Series, Index])
+    def test_str_cat_all_na(self, box, other):
+        # check that all NaNs in caller / target work
+        s = Index(['a', 'b', 'c', 'd'])
+        s = s if box == Index else Series(s, index=s)
+        t = other([np.nan] * 4, dtype='object')
+        # add index of s for alignment
+        t = t if other == Index else Series(t, index=s)
+
+        # all-NA target
+        expected = Index([np.nan] * 4, dtype='object')
+        expected = expected if box == Index else Series(expected,
+                                                        index=s.index)
+        result = s.str.cat(t, join='left')
+        assert_series_or_index_equal(result, expected)
+
+        # all-NA caller (only for Series)
+        if other == Series:
+            expected = Series([np.nan] * 4, dtype='object', index=t.index)
+            result = t.str.cat(s, join='left')
+            tm.assert_series_equal(result, expected)
+
     def test_str_cat_special_cases(self):
         s = Series(['a', 'b', 'c', 'd'])
         t = Series(['d', 'a', 'e', 'b'], index=[3, 0, 4, 1])