BUG: Index.drop raising Error when Index has duplicates

doc/source/whatsnew/v1.2.0.rst

@@ -636,6 +636,7 @@ MultiIndex
 - Bug in :meth:`DataFrame.reset_index` with ``NaT`` values in index raises ``ValueError`` with message ``"cannot convert float NaN to integer"`` (:issue:`36541`)
 - Bug in :meth:`DataFrame.combine_first` when used with :class:`MultiIndex` containing string and ``NaN`` values raises ``TypeError`` (:issue:`36562`)
 - Bug in :meth:`MultiIndex.drop` dropped ``NaN`` values when non existing key was given as input (:issue:`18853`)
+- Bug in :meth:`MultiIndex.drop` dropping more values than expected when index has duplicates and is not sorted (:issue:`33494`)
 
 I/O
 ^^^
@@ -758,6 +759,7 @@ Other
 - Passing an array with 2 or more dimensions to the :class:`Series` constructor now raises the more specific ``ValueError``, from a bare ``Exception`` previously (:issue:`35744`)
 - Bug in ``accessor.DirNamesMixin``, where ``dir(obj)`` wouldn't show attributes defined on the instance (:issue:`37173`).
 - Bug in :meth:`Series.nunique` with ``dropna=True`` was returning incorrect results when both ``NA`` and ``None`` missing values were present (:issue:`37566`)
+- Bug in :meth:`Index.drop` raising ``InvalidIndexError`` when index has duplicates (:issue:`38051`)
 
 .. ---------------------------------------------------------------------------
 

pandas/core/indexes/base.py

@@ -5508,7 +5508,7 @@ def drop(self, labels, errors: str_t = "raise"):
         """
         arr_dtype = "object" if self.dtype == "object" else None
         labels = com.index_labels_to_array(labels, dtype=arr_dtype)
-        indexer = self.get_indexer(labels)
+        indexer = self.get_indexer_for(labels)
         mask = indexer == -1
         if mask.any():
             if errors != "ignore":

pandas/core/indexes/multi.py

@@ -2169,7 +2169,8 @@ def drop(self, codes, level=None, errors="raise"):
                 if isinstance(loc, int):
                     inds.append(loc)
                 elif isinstance(loc, slice):
-                    inds.extend(range(loc.start, loc.stop))
+                    step = loc.step if loc.step is not None else 1
+                    inds.extend(range(loc.start, loc.stop, step))
                 elif com.is_bool_indexer(loc):
                     if self.lexsort_depth == 0:
                         warnings.warn(

pandas/tests/indexes/multi/test_drop.py

@@ -1,3 +1,5 @@
+import warnings
+
 import numpy as np
 import pytest
 
@@ -147,3 +149,13 @@ def test_drop_with_nan_in_index(nulls_fixture):
     msg = r"labels \[Timestamp\('2001-01-01 00:00:00'\)\] not found in level"
     with pytest.raises(KeyError, match=msg):
         mi.drop(pd.Timestamp("2001"), level="date")
+
+
+def test_drop_with_non_monotonic_duplicates():
+    # GH#33494
+    mi = MultiIndex.from_tuples([(1, 2), (2, 3), (1, 2)])
+    with warnings.catch_warnings():
+        warnings.simplefilter("ignore", PerformanceWarning)
+        result = mi.drop((1, 2))
+    expected = MultiIndex.from_tuples([(2, 3)])
+    tm.assert_index_equal(result, expected)

pandas/tests/indexes/test_base.py

@@ -1496,6 +1496,13 @@ def test_drop_tuple(self, values, to_drop):
             with pytest.raises(KeyError, match=msg):
                 removed.drop(drop_me)
 
+    def test_drop_with_duplicates_in_index(self):
+        # GH38051
+        index = Index([0, 1, 0, 1])
+        result = index.drop(0)
+        expected = Index([1, 1])
+        tm.assert_index_equal(result, expected)
+
     @pytest.mark.parametrize(
         "attr",
         [