BUG: to_records() fails for MultiIndex DF (pandas-dev#21064)

fersarr · May 16, 2018 · 81d32b8 · 81d32b8
1 parent a327920
commit 81d32b8
Show file tree

Hide file tree

Showing 3 changed files with 31 additions and 2 deletions.
diff --git a/doc/source/whatsnew/v0.24.0.txt b/doc/source/whatsnew/v0.24.0.txt
@@ -131,7 +131,7 @@ Indexing
 MultiIndex
 ^^^^^^^^^^
 
--
+- Bug :func:`to_records` fails for empty MultiIndex DF (:issue:`21064`)
 -
 -
 

diff --git a/pandas/core/frame.py b/pandas/core/frame.py
@@ -1392,7 +1392,13 @@ def to_records(self, index=True, convert_datetime64=None):
             else:
                 if isinstance(self.index, MultiIndex):
                     # array of tuples to numpy cols. copy copy copy
-                    ix_vals = lmap(np.array, zip(*self.index.values))
+                    tuples = self.index.values
+                    if len(tuples):
+                        ix_vals = lmap(np.array, zip(*tuples))
+                    else:
+                        # empty MultiIndex DF
+                        ix_vals = [np.array([], dtype=lev.dtype)
+                                   for lev in self.index.levels]
                 else:
                     ix_vals = [self.index.values]
 

diff --git a/pandas/tests/frame/test_convert_to.py b/pandas/tests/frame/test_convert_to.py
@@ -328,3 +328,26 @@ def test_to_dict_index_dtypes(self, into, expected):
         result = DataFrame.from_dict(result, orient='index')[cols]
         expected = DataFrame.from_dict(expected, orient='index')[cols]
         tm.assert_frame_equal(result, expected)
+
+    def test_to_records_with_multiindex(self):
+        size = 4
+        tup = zip(*[['a', 'a', 'b', 'b'], ['x', 'y', 'x', 'y']])
+        index = MultiIndex.from_tuples(tup)
+        df = DataFrame(np.random.randn(size, size), index=index)
+
+        records = df.to_records(index=True)
+        assert len(records) == size
+        # first record includes index
+        assert records[0][0] == 'a'
+        assert records[0][1] == 'x'
+        # last record includes index
+        assert records[size - 1][0] == 'b'
+        assert records[size - 1][1] == 'y'
+
+    def test_to_records_with_empty_multiindex(self):
+        # GH 21064
+        multi = MultiIndex([['a'], ['b']], labels=[[], []])
+        df = DataFrame(columns=['A'], index=multi)
+
+        records = df.to_records(index=True)
+        assert len(records) == 0