Fix duplicate bin detection for large cut inputs

python/cudf/cudf/core/cut.py

@@ -147,7 +147,7 @@ def cut(
 
     # bins can either be an int, sequence of scalars or an intervalIndex
     if isinstance(bins, Sequence):
-        if len(set(bins)) is not len(bins):
+        if len(set(bins)) != len(bins):
             if duplicates == "raise":
                 raise ValueError(
                     f"Bin edges must be unique: {bins!r}.\n"

python/cudf/cudf/tests/general_functions/test_cut.py

@@ -230,6 +230,29 @@ def test_cut_drop_duplicates_raises(
         )
 
 
+def test_cut_unique_bins_larger_than_integer_cache():
+    x = [1, 2]
+    bins = list(range(257))
+
+    expected = pd.cut(x=x, bins=bins, labels=False)
+    actual = cut(x=x, bins=bins, labels=False)
+
+    assert_eq(expected, actual)
+
+
+def test_cut_duplicate_bins_with_distinct_large_integers():
+    x = [1, 2]
+    # Construct equal bin edges independently so they are distinct objects.
+    duplicate_edge = int("257")
+    bins = [0, duplicate_edge, int("257"), 258]
+
+    assert bins[1] == bins[2]
+    assert bins[1] is not bins[2]
+
+    with pytest.raises(ValueError, match="Bin edges must be unique"):
+        cut(x=x, bins=bins, labels=False)
+
+
 @pytest.mark.parametrize(
     "x",
     [