BUG: Fix qcut for nullable integers

doc/source/whatsnew/v1.0.1.rst

@@ -131,6 +131,7 @@ ExtensionArray
 ^^^^^^^^^^^^^^
 
 - Bug in dtype being lost in ``__invert__``  (``~`` operator) for extension-array backed ``Series`` and ``DataFrame`` (:issue:`23087`)
+- Bug where :meth:`qcut` would raise when passed a nullable integer. (:issue:`31389`)
 -
 
 

pandas/core/reshape/tile.py

@@ -202,17 +202,10 @@ def cut(
     """
     # NOTE: this binning code is changed a bit from histogram for var(x) == 0
 
-    # for handling the cut for datetime and timedelta objects
     original = x
     x = _preprocess_for_cut(x)
     x, dtype = _coerce_to_type(x)
 
-    # To support cut(IntegerArray), we convert to object dtype with NaN
-    # Will properly support in the future.
-    # https://github.com/pandas-dev/pandas/pull/31290
-    if is_extension_array_dtype(x.dtype) and is_integer_dtype(x.dtype):
-        x = x.to_numpy(dtype=object, na_value=np.nan)
-
     if not np.iterable(bins):
         if is_scalar(bins) and bins < 1:
             raise ValueError("`bins` should be a positive integer.")
@@ -434,7 +427,7 @@ def _bins_to_cuts(
 
 def _coerce_to_type(x):
     """
-    if the passed data is of datetime/timedelta or bool type,
+    if the passed data is of datetime/timedelta, bool or nullable int type,
     this method converts it to numeric so that cut or qcut method can
     handle it
     """
@@ -451,6 +444,12 @@ def _coerce_to_type(x):
     elif is_bool_dtype(x):
         # GH 20303
         x = x.astype(np.int64)
+    # To support cut and qcut for IntegerArray we convert to float dtype.
+    # Will properly support in the future.
+    # https://github.com/pandas-dev/pandas/pull/31290
+    # https://github.com/pandas-dev/pandas/issues/31389
+    elif is_extension_array_dtype(x) and is_integer_dtype(x):
+        x = x.to_numpy(dtype=np.float64, na_value=np.nan)
 
     if dtype is not None:
         # GH 19768: force NaT to NaN during integer conversion

pandas/tests/arrays/test_integer.py

@@ -1061,19 +1061,6 @@ def test_value_counts_na():
     tm.assert_series_equal(result, expected)
 
 
-@pytest.mark.parametrize("bins", [3, [0, 5, 15]])
-@pytest.mark.parametrize("right", [True, False])
-@pytest.mark.parametrize("include_lowest", [True, False])
-def test_cut(bins, right, include_lowest):
-    a = np.random.randint(0, 10, size=50).astype(object)
-    a[::2] = np.nan
-    result = pd.cut(
-        pd.array(a, dtype="Int64"), bins, right=right, include_lowest=include_lowest
-    )
-    expected = pd.cut(a, bins, right=right, include_lowest=include_lowest)
-    tm.assert_categorical_equal(result, expected)
-
-
 def test_array_setitem_nullable_boolean_mask():
     # GH 31446
     ser = pd.Series([1, 2], dtype="Int64")

pandas/tests/reshape/test_cut.py

@@ -612,3 +612,16 @@ def test_cut_incorrect_labels(labels):
     msg = "Bin labels must either be False, None or passed in as a list-like argument"
     with pytest.raises(ValueError, match=msg):
         cut(values, 4, labels=labels)
+
+
+@pytest.mark.parametrize("bins", [3, [0, 5, 15]])
+@pytest.mark.parametrize("right", [True, False])
+@pytest.mark.parametrize("include_lowest", [True, False])
+def test_cut_nullable_integer(bins, right, include_lowest):
+    a = np.random.randint(0, 10, size=50).astype(float)
+    a[::2] = np.nan
+    result = cut(
+        pd.array(a, dtype="Int64"), bins, right=right, include_lowest=include_lowest
+    )
+    expected = cut(a, bins, right=right, include_lowest=include_lowest)
+    tm.assert_categorical_equal(result, expected)

pandas/tests/reshape/test_qcut.py

@@ -3,6 +3,7 @@
 import numpy as np
 import pytest
 
+import pandas as pd
 from pandas import (
     Categorical,
     DatetimeIndex,
@@ -286,3 +287,14 @@ def test_qcut_bool_coercion_to_int(bins, box, compare):
     expected = qcut(data_expected, bins, duplicates="drop")
     result = qcut(data_result, bins, duplicates="drop")
     compare(result, expected)
+
+
+@pytest.mark.parametrize("q", [2, 5, 10])
+def test_qcut_nullable_integer(q, any_nullable_int_dtype):
+    arr = pd.array(np.arange(100), dtype=any_nullable_int_dtype)
+    arr[::2] = pd.NA
+
+    result = qcut(arr, q)
+    expected = qcut(arr.astype(float), q)
+
+    tm.assert_categorical_equal(result, expected)