don't raise exceptions splitting a blank string

Author: Montana Low

pandas/core/strings.py

@@ -1633,7 +1633,8 @@ def cons_row(x):
             if result:
                 # propagate nan values to match longest sequence (GH 18450)
                 max_len = max(len(x) for x in result)
-                result = [x * max_len if x[0] is np.nan else x for x in result]
+                result = [x * max_len if len(x) == 0 or x[0] is np.nan
+                          else x for x in result]
 
         if not isinstance(expand, bool):
             raise ValueError("expand must be True or False")

pandas/tests/test_strings.py

@@ -1992,6 +1992,19 @@ def test_rsplit(self):
         exp = Series([['a_b', 'c'], ['c_d', 'e'], NA, ['f_g', 'h']])
         tm.assert_series_equal(result, exp)
 
+    def test_split_blank_string(self):
+        # expand blank split GH 20067
+        values = Series([''], name='test')
+        result = values.str.split(expand=True)
+        exp = DataFrame([[]])
+        tm.assert_frame_equal(result, exp)
+
+        values = Series(['a b c', 'a b', '', ' '], name='test')
+        result = values.str.split(expand=True)
+        exp = DataFrame([['a', 'b', 'c'], ['a', 'b', np.nan],
+                         [np.nan, np.nan, np.nan], [np.nan, np.nan, np.nan]])
+        tm.assert_frame_equal(result, exp)
+
     def test_split_noargs(self):
         # #1859
         s = Series(['Wes McKinney', 'Travis  Oliphant'])