Fix multiindex loc nan

doc/source/whatsnew/v1.4.0.rst

@@ -595,7 +595,7 @@ Groupby/resample/rolling
 - Fixed bug in :meth:`Series.rolling` and :meth:`DataFrame.rolling` for centered datetimelike windows with uneven nanosecond (:issue:`43997`)
 - Bug in :meth:`GroupBy.nth` failing on ``axis=1`` (:issue:`43926`)
 - Fixed bug in :meth:`Series.rolling` and :meth:`DataFrame.rolling` not respecting right bound on centered datetime-like windows, if the index contain duplicates (:issue:`#3944`)
-
+- Bug in :meth:`DataFrame.groupby` when grouping on multiple columns where at least one includes ``np.nan`` which resulted in a ``KeyError`` when the ``np.nan`` containing index was selected with :meth:`Series.loc` (:issue:`43814`)
 
 Reshaping
 ^^^^^^^^^

pandas/core/groupby/grouper.py

@@ -14,6 +14,7 @@
 
 import numpy as np
 
+from pandas._libs.tslibs.nattype import NaTType
 from pandas._typing import (
     ArrayLike,
     NDFrameT,
@@ -34,6 +35,7 @@
     Categorical,
     ExtensionArray,
 )
+from pandas.core.base import isna
 import pandas.core.common as com
 from pandas.core.frame import DataFrame
 from pandas.core.groupby import ops
@@ -486,6 +488,7 @@ def __init__(
         self._observed = observed
         self.in_axis = in_axis
         self._dropna = dropna
+        self._na_placeholder = None
 
         self._passed_categorical = False
 
@@ -690,6 +693,13 @@ def _codes_and_uniques(self) -> tuple[np.ndarray, ArrayLike]:
             codes, uniques = algorithms.factorize(
                 self.grouping_vector, sort=self._sort, na_sentinel=na_sentinel
             )
+            # GH43943, store placeholder for (np.nan, pd.NaT, np.datetime64("NaT", "ns"),
+            # np.timedelta64("NaT", "ns")), will later be replaced by -1 in
+            # pandas/core/groupby:reconstructed_codes
+            if not self._dropna:
+                if isna(self.grouping_vector).any():
+                    self._na_placeholder = max(codes)
+
         return codes, uniques
 
     @cache_readonly

pandas/core/groupby/ops.py

@@ -859,7 +859,21 @@ def ngroups(self) -> int:
     def reconstructed_codes(self) -> list[np.ndarray]:
         codes = self.codes
         ids, obs_ids, _ = self.group_info
-        return decons_obs_group_ids(ids, obs_ids, self.shape, codes, xnull=True)
+        reconstructed_codes = decons_obs_group_ids(
+            ids, obs_ids, self.shape, codes, xnull=True
+        )
+
+        def transform_codes(code_level, grouping):
+            if grouping._na_placeholder is not None:
+                return np.where(code_level == grouping._na_placeholder, -1, code_level)
+            else:
+                return code_level
+
+        transformed_rec_codes = [
+            transform_codes(code_level, grouping)
+            for code_level, grouping in zip(reconstructed_codes, self._groupings)
+        ]
+        return transformed_rec_codes
 
     @final
     @cache_readonly

pandas/tests/groupby/test_groupby_dropna.py

@@ -370,3 +370,51 @@ def test_groupby_nan_included():
         tm.assert_numpy_array_equal(result_values, expected_values)
     assert np.isnan(list(result.keys())[2])
     assert list(result.keys())[0:2] == ["g1", "g2"]
+
+
+@pytest.mark.parametrize(
+    "na",
+    [np.nan, pd.NaT, pd.NA, np.datetime64("NaT", "ns"), np.timedelta64("NaT", "ns")],
+)
+def test_groupby_codes_with_nan_in_multiindex(na):
+    # GH 43814
+    df = pd.DataFrame(
+        {
+            "temp_playlist": [0, 0, 0, 0],
+            "objId": ["o1", na, "o1", na],
+            "x": [1, 2, 3, 4],
+        }
+    )
+
+    grouped_df = df.groupby(by=["temp_playlist", "objId"], dropna=False)["x"].sum()
+    expected = pd.MultiIndex.from_arrays(
+        [[0, 0], ["o1", na]], names=["temp_playlist", "objId"]
+    )
+    result = grouped_df.index
+    assert all((res == ex).all() for res, ex in zip(result.codes, expected.codes))
+
+
+def test_groupby_multiindex_multiply_with_series():
+    # GH#36060
+    df = pd.DataFrame(
+        {
+            "animal": ["Falcon", "Falcon", "Parrot", "Parrot"],
+            "type": [np.nan, np.nan, np.nan, np.nan],
+            "speed": [380.0, 370.0, 24.0, 26.0],
+        }
+    )
+    speed = df.groupby(["animal", "type"], dropna=False)["speed"].first()
+    # Reconstruct same index to allow for multiplication.
+    ix_wing = pd.MultiIndex.from_tuples(
+        [("Falcon", np.nan), ("Parrot", np.nan)], names=["animal", "type"]
+    )
+    wing = pd.Series([42, 44], index=ix_wing)
+
+    result = wing * speed
+    expected = pd.Series(
+        [15960.0, 1056.0],
+        index=pd.MultiIndex.from_tuples(
+            [("Falcon", np.nan), ("Parrot", np.nan)], names=["animal", "type"]
+        ),
+    )
+    tm.assert_series_equal(result, expected)