BUG: #10645 in using MultiIndex.__contains__

doc/source/whatsnew/v0.17.0.txt

@@ -404,4 +404,5 @@ Bug Fixes
 - Bug in ``io.common.get_filepath_or_buffer`` which caused reading of valid S3 files to fail if the bucket also contained keys for which the user does not have read permission (:issue:`10604`)
 - Bug in vectorised setting of timestamp columns with python ``datetime.date`` and numpy ``datetime64`` (:issue:`10408`, :issue:`10412`)
 
+- Bug in ``MultiIndex.__contains__`` throws an ``IndexError`` for large multiindices (:issue:`10645`)
 - Bug in ``pd.DataFrame`` when constructing an empty DataFrame with a string dtype (:issue:`9428`)

pandas/index.pyx

@@ -143,6 +143,11 @@ cdef class IndexEngine:
                 return self._get_loc_duplicates(val)
             values = self._get_index_values()
             loc = _bin_search(values, val) # .searchsorted(val, side='left')
+
+            # GH10675
+            if len(values) <= loc or 0 > loc:
+                raise KeyError(val)
+
             if util.get_value_at(values, loc) != val:
                 raise KeyError(val)
             return loc

pandas/tests/test_index.py

@@ -1253,6 +1253,14 @@ def test_get_loc(self):
         with tm.assertRaises(TypeError):
             idx.get_loc('a', method='nearest')
 
+    def test_get_loc_keyerror(self):
+        # GH10645
+        mi = pd.MultiIndex.from_arrays([range(100), range(100)])
+        self.assertRaises(KeyError, lambda: mi.get_loc((1000001, 0)))
+
+        mi = pd.MultiIndex.from_arrays([range(1000000), range(1000000)])
+        self.assertRaises(KeyError, lambda: mi.get_loc((1000001, 0)))
+
     def test_slice_locs(self):
         for dtype in [int, float]:
             idx = Index(np.array([0, 1, 2, 5, 6, 7, 9, 10], dtype=dtype))