Backport PR #55369 on branch 2.1.x (MAINT: Remove np.int_ and np.uint) (#55387)

Backport PR #55369: MAINT: Remove `np.int_` and `np.uint`

Co-authored-by: Mateusz Sokół <[email protected]>

Author: meeseeksmachine

doc/source/user_guide/enhancingperf.rst

@@ -184,7 +184,7 @@ can be improved by passing an ``np.ndarray``.
       ...: cpdef np.ndarray[double] apply_integrate_f(np.ndarray col_a, np.ndarray col_b,
       ...:                                            np.ndarray col_N):
       ...:     assert (col_a.dtype == np.float64
-      ...:             and col_b.dtype == np.float64 and col_N.dtype == np.int_)
+      ...:             and col_b.dtype == np.float64 and col_N.dtype == np.dtype(int))
       ...:     cdef Py_ssize_t i, n = len(col_N)
       ...:     assert (len(col_a) == len(col_b) == n)
       ...:     cdef np.ndarray[double] res = np.empty(n)

pandas/compat/numpy/__init__.py

@@ -21,6 +21,21 @@
     )
 
 
+np_long: type
+np_ulong: type
+
+if _nlv >= Version("2.0.0.dev0"):
+    try:
+        np_long = np.long  # type: ignore[attr-defined]
+        np_ulong = np.ulong  # type: ignore[attr-defined]
+    except AttributeError:
+        np_long = np.int_
+        np_ulong = np.uint
+else:
+    np_long = np.int_
+    np_ulong = np.uint
+
+
 __all__ = [
     "np",
     "_np_version",

pandas/core/algorithms.py

@@ -1641,7 +1641,7 @@ def safe_sort(
         else:
             mask = None
     else:
-        reverse_indexer = np.empty(len(sorter), dtype=np.int_)
+        reverse_indexer = np.empty(len(sorter), dtype=int)
         reverse_indexer.put(sorter, np.arange(len(sorter)))
         # Out of bound indices will be masked with `-1` next, so we
         # may deal with them here without performance loss using `mode='wrap'`

pandas/tests/arrays/boolean/test_reduction.py

@@ -1,6 +1,8 @@
 import numpy as np
 import pytest
 
+from pandas.compat.numpy import np_long
+
 import pandas as pd
 
 
@@ -51,7 +53,7 @@ def test_reductions_return_types(dropna, data, all_numeric_reductions):
         s = s.dropna()
 
     if op in ("sum", "prod"):
-        assert isinstance(getattr(s, op)(), np.int_)
+        assert isinstance(getattr(s, op)(), np_long)
     elif op == "count":
         # Oddly on the 32 bit build (but not Windows), this is intc (!= intp)
         assert isinstance(getattr(s, op)(), np.integer)

pandas/tests/dtypes/cast/test_infer_dtype.py

@@ -170,7 +170,7 @@ def test_infer_dtype_from_scalar(value, expected):
 @pytest.mark.parametrize(
     "arr, expected",
     [
-        ([1], np.int_),
+        ([1], np.dtype(int)),
         (np.array([1], dtype=np.int64), np.int64),
         ([np.nan, 1, ""], np.object_),
         (np.array([[1.0, 2.0]]), np.float64),

pandas/tests/extension/base/dim2.py

@@ -237,7 +237,7 @@ def get_reduction_result_dtype(dtype):
                 return NUMPY_INT_TO_DTYPE[np.dtype(int)]
             else:
                 # i.e. dtype.kind == "u"
-                return NUMPY_INT_TO_DTYPE[np.dtype(np.uint)]
+                return NUMPY_INT_TO_DTYPE[np.dtype("uint")]
 
         if method in ["sum", "prod"]:
             # std and var are not dtype-preserving

pandas/tests/frame/indexing/test_indexing.py

@@ -108,11 +108,11 @@ def test_setitem_list(self, float_frame):
             data["A"] = newcolumndata
 
     def test_setitem_list2(self):
-        df = DataFrame(0, index=range(3), columns=["tt1", "tt2"], dtype=np.int_)
+        df = DataFrame(0, index=range(3), columns=["tt1", "tt2"], dtype=int)
         df.loc[1, ["tt1", "tt2"]] = [1, 2]
 
         result = df.loc[df.index[1], ["tt1", "tt2"]]
-        expected = Series([1, 2], df.columns, dtype=np.int_, name=1)
+        expected = Series([1, 2], df.columns, dtype=int, name=1)
         tm.assert_series_equal(result, expected)
 
         df["tt1"] = df["tt2"] = "0"

pandas/tests/frame/methods/test_shift.py

@@ -1,6 +1,7 @@
 import numpy as np
 import pytest
 
+from pandas.compat.numpy import np_long
 import pandas.util._test_decorators as td
 
 import pandas as pd
@@ -471,22 +472,22 @@ def test_shift_axis1_multiple_blocks_with_int_fill(self):
         df1 = DataFrame(rng.integers(1000, size=(5, 3), dtype=int))
         df2 = DataFrame(rng.integers(1000, size=(5, 2), dtype=int))
         df3 = pd.concat([df1.iloc[:4, 1:3], df2.iloc[:4, :]], axis=1)
-        result = df3.shift(2, axis=1, fill_value=np.int_(0))
+        result = df3.shift(2, axis=1, fill_value=np_long(0))
         assert len(df3._mgr.blocks) == 2
 
         expected = df3.take([-1, -1, 0, 1], axis=1)
-        expected.iloc[:, :2] = np.int_(0)
+        expected.iloc[:, :2] = np_long(0)
         expected.columns = df3.columns
 
         tm.assert_frame_equal(result, expected)
 
         # Case with periods < 0
         df3 = pd.concat([df1.iloc[:4, 1:3], df2.iloc[:4, :]], axis=1)
-        result = df3.shift(-2, axis=1, fill_value=np.int_(0))
+        result = df3.shift(-2, axis=1, fill_value=np_long(0))
         assert len(df3._mgr.blocks) == 2
 
         expected = df3.take([2, 3, -1, -1], axis=1)
-        expected.iloc[:, -2:] = np.int_(0)
+        expected.iloc[:, -2:] = np_long(0)
         expected.columns = df3.columns
 
         tm.assert_frame_equal(result, expected)

pandas/tests/frame/test_constructors.py

@@ -1823,7 +1823,7 @@ def test_constructor_single_value(self):
             DataFrame("a", [1, 2], ["a", "c"], float)
 
     def test_constructor_with_datetimes(self):
-        intname = np.dtype(np.int_).name
+        intname = np.dtype(int).name
         floatname = np.dtype(np.float64).name
         objectname = np.dtype(np.object_).name
 

pandas/tests/frame/test_reductions.py

@@ -10,6 +10,10 @@
     IS64,
     is_platform_windows,
 )
+from pandas.compat.numpy import (
+    np_long,
+    np_ulong,
+)
 import pandas.util._test_decorators as td
 
 import pandas as pd
@@ -1712,11 +1716,11 @@ class TestEmptyDataFrameReductions:
         "opname, dtype, exp_value, exp_dtype",
         [
             ("sum", np.int8, 0, np.int64),
-            ("prod", np.int8, 1, np.int_),
+            ("prod", np.int8, 1, np_long),
             ("sum", np.int64, 0, np.int64),
             ("prod", np.int64, 1, np.int64),
             ("sum", np.uint8, 0, np.uint64),
-            ("prod", np.uint8, 1, np.uint),
+            ("prod", np.uint8, 1, np_ulong),
             ("sum", np.uint64, 0, np.uint64),
             ("prod", np.uint64, 1, np.uint64),
             ("sum", np.float32, 0, np.float32),