BUG: Fix keyerror bug when indexing multiindex columns with NaT values (

#60463) * BUG: Fix keyerror bug when indexing multiindex columns with NaT values * BUG: Update whatsnew/v3.0.0.rst * BUG: Move new test to test_multilevel.py
pandas-dev · Dec 3, 2024 · a6f721e · a6f721e
1 parent 40131a6
commit a6f721e
Show file tree

Hide file tree

Showing 3 changed files with 27 additions and 4 deletions.
diff --git a/doc/source/whatsnew/v3.0.0.rst b/doc/source/whatsnew/v3.0.0.rst
@@ -667,6 +667,7 @@ Indexing
 ^^^^^^^^
 - Bug in :meth:`DataFrame.__getitem__` returning modified columns when called with ``slice`` in Python 3.12 (:issue:`57500`)
 - Bug in :meth:`DataFrame.from_records` throwing a ``ValueError`` when passed an empty list in ``index`` (:issue:`58594`)
+- Bug in :meth:`MultiIndex.insert` when a new value inserted to a datetime-like level gets cast to ``NaT`` and fails indexing (:issue:`60388`)
 - Bug in printing :attr:`Index.names` and :attr:`MultiIndex.levels` would not escape single quotes (:issue:`60190`)
 
 Missing

diff --git a/pandas/core/indexes/multi.py b/pandas/core/indexes/multi.py
@@ -4084,11 +4084,10 @@ def insert(self, loc: int, item) -> MultiIndex:
                 # have to insert into level
                 # must insert at end otherwise you have to recompute all the
                 # other codes
-                if isna(k):  # GH 59003
+                lev_loc = len(level)
+                level = level.insert(lev_loc, k)
+                if isna(level[lev_loc]):  # GH 59003, 60388
                     lev_loc = -1
-                else:
-                    lev_loc = len(level)
-                    level = level.insert(lev_loc, k)
             else:
                 lev_loc = level.get_loc(k)
 

diff --git a/pandas/tests/test_multilevel.py b/pandas/tests/test_multilevel.py
@@ -295,6 +295,29 @@ def test_multiindex_insert_level_with_na(self, na):
         df[na, "B"] = 1
         tm.assert_frame_equal(df[na], DataFrame([1], columns=["B"]))
 
+    def test_multiindex_dt_with_nan(self):
+        # GH#60388
+        df = DataFrame(
+            [
+                [1, np.nan, 5, np.nan],
+                [2, np.nan, 6, np.nan],
+                [np.nan, 3, np.nan, 7],
+                [np.nan, 4, np.nan, 8],
+            ],
+            index=Series(["a", "b", "c", "d"], dtype=object, name="sub"),
+            columns=MultiIndex.from_product(
+                [
+                    ["value1", "value2"],
+                    [datetime.datetime(2024, 11, 1), datetime.datetime(2024, 11, 2)],
+                ],
+                names=[None, "Date"],
+            ),
+        )
+        df = df.reset_index()
+        result = df[df.columns[0]]
+        expected = Series(["a", "b", "c", "d"], name=("sub", np.nan))
+        tm.assert_series_equal(result, expected)
+
 
 class TestSorted:
     """everything you wanted to test about sorting"""