From 11dfd16e988044df9f58e1408e8e81c4064defe8 Mon Sep 17 00:00:00 2001
From: Matthew Roeschke <10647082+mroeschke@users.noreply.github.com>
Date: Thu, 5 Sep 2024 17:37:12 -0700
Subject: [PATCH 1/4] Add labeling APIs to pylibcudf

---
 python/cudf/cudf/_lib/labeling.pyx            | 40 ++++---------
 python/pylibcudf/pylibcudf/CMakeLists.txt     |  1 +
 python/pylibcudf/pylibcudf/__init__.pxd       |  1 +
 python/pylibcudf/pylibcudf/__init__.py        |  3 +
 python/pylibcudf/pylibcudf/labeling.pxd       | 13 ++++
 python/pylibcudf/pylibcudf/labeling.pyx       | 59 +++++++++++++++++++
 .../pylibcudf/pylibcudf/libcudf/labeling.pxd  |  8 +--
 .../pylibcudf/tests/test_labeling.py          | 20 +++++++
 8 files changed, 111 insertions(+), 34 deletions(-)
 create mode 100644 python/pylibcudf/pylibcudf/labeling.pxd
 create mode 100644 python/pylibcudf/pylibcudf/labeling.pyx
 create mode 100644 python/pylibcudf/pylibcudf/tests/test_labeling.py

diff --git a/python/cudf/cudf/_lib/labeling.pyx b/python/cudf/cudf/_lib/labeling.pyx
index 2e1959a348d..3966cce8981 100644
--- a/python/cudf/cudf/_lib/labeling.pyx
+++ b/python/cudf/cudf/_lib/labeling.pyx
@@ -1,16 +1,11 @@
 # Copyright (c) 2021-2024, NVIDIA CORPORATION.
 
-from cudf.core.buffer import acquire_spill_lock
-
 from libcpp cimport bool as cbool
-from libcpp.memory cimport unique_ptr
-from libcpp.utility cimport move
 
-from pylibcudf.libcudf.column.column cimport column
-from pylibcudf.libcudf.column.column_view cimport column_view
-from pylibcudf.libcudf.labeling cimport inclusive, label_bins as cpp_label_bins
+import pylibcudf as plc
 
 from cudf._lib.column cimport Column
+from cudf.core.buffer import acquire_spill_lock
 
 
 # Note that the parameter input shadows a Python built-in in the local scope,
@@ -19,26 +14,11 @@ from cudf._lib.column cimport Column
 @acquire_spill_lock()
 def label_bins(Column input, Column left_edges, cbool left_inclusive,
                Column right_edges, cbool right_inclusive):
-    cdef inclusive c_left_inclusive = \
-        inclusive.YES if left_inclusive else inclusive.NO
-    cdef inclusive c_right_inclusive = \
-        inclusive.YES if right_inclusive else inclusive.NO
-
-    cdef column_view input_view = input.view()
-    cdef column_view left_edges_view = left_edges.view()
-    cdef column_view right_edges_view = right_edges.view()
-
-    cdef unique_ptr[column] c_result
-
-    with nogil:
-        c_result = move(
-            cpp_label_bins(
-                input_view,
-                left_edges_view,
-                c_left_inclusive,
-                right_edges_view,
-                c_right_inclusive,
-            )
-        )
-
-    return Column.from_unique_ptr(move(c_result))
+    plc_column = plc.labeling.label_bins(
+        input.to_pylibcudf(mode="read"),
+        left_edges.to_pylibcudf(mode="read"),
+        left_inclusive,
+        right_edges.to_pylibcudf(mode="read"),
+        right_inclusive
+    )
+    return Column.from_pylibcudf(plc_column)
diff --git a/python/pylibcudf/pylibcudf/CMakeLists.txt b/python/pylibcudf/pylibcudf/CMakeLists.txt
index a4f17344cb0..f07c8897e34 100644
--- a/python/pylibcudf/pylibcudf/CMakeLists.txt
+++ b/python/pylibcudf/pylibcudf/CMakeLists.txt
@@ -27,6 +27,7 @@ set(cython_sources
     groupby.pyx
     interop.pyx
     join.pyx
+    labeling.pyx
     lists.pyx
     merge.pyx
     null_mask.pyx
diff --git a/python/pylibcudf/pylibcudf/__init__.pxd b/python/pylibcudf/pylibcudf/__init__.pxd
index 841efa59bda..b7cf6413c05 100644
--- a/python/pylibcudf/pylibcudf/__init__.pxd
+++ b/python/pylibcudf/pylibcudf/__init__.pxd
@@ -13,6 +13,7 @@ from . cimport (
     filling,
     groupby,
     join,
+    labeling,
     lists,
     merge,
     null_mask,
diff --git a/python/pylibcudf/pylibcudf/__init__.py b/python/pylibcudf/pylibcudf/__init__.py
index d3878a89a6a..84b1c29f791 100644
--- a/python/pylibcudf/pylibcudf/__init__.py
+++ b/python/pylibcudf/pylibcudf/__init__.py
@@ -24,6 +24,7 @@
     interop,
     io,
     join,
+    labeling,
     lists,
     merge,
     null_mask,
@@ -67,7 +68,9 @@
     "gpumemoryview",
     "groupby",
     "interop",
+    "io",
     "join",
+    "labeling",
     "lists",
     "merge",
     "null_mask",
diff --git a/python/pylibcudf/pylibcudf/labeling.pxd b/python/pylibcudf/pylibcudf/labeling.pxd
new file mode 100644
index 00000000000..0415178b3a9
--- /dev/null
+++ b/python/pylibcudf/pylibcudf/labeling.pxd
@@ -0,0 +1,13 @@
+# Copyright (c) 2024, NVIDIA CORPORATION.
+from libcpp cimport bool
+
+from .column cimport Column
+
+
+cpdef Column label_bins(
+    Column input,
+    Column left_edges,
+    bool left_inclusive,
+    Column right_edges,
+    bool right_inclusive
+)
diff --git a/python/pylibcudf/pylibcudf/labeling.pyx b/python/pylibcudf/pylibcudf/labeling.pyx
new file mode 100644
index 00000000000..8c269b51ca4
--- /dev/null
+++ b/python/pylibcudf/pylibcudf/labeling.pyx
@@ -0,0 +1,59 @@
+# Copyright (c) 2024, NVIDIA CORPORATION.
+
+from libcpp.memory cimport unique_ptr
+from libcpp.utility cimport move
+from pylibcudf.libcudf cimport labeling as cpp_labeling
+from pylibcudf.libcudf.column.column cimport column
+from pylibcudf.libcudf.labeling cimport inclusive
+
+from .column cimport Column
+
+
+cpdef Column label_bins(
+    Column input,
+    Column left_edges,
+    bool left_inclusive,
+    Column right_edges,
+    bool right_inclusive
+):
+    """Labels elements based on membership in the specified bins.
+
+    Parameters
+    ----------
+    input : Column
+        Column of input elements to label according to the specified bins.
+    left_edges : Column
+        Column of the left edge of each bin.
+    left_inclusive : bool
+        Whether or not the left edge is inclusive.
+    right_edges : Column
+        Column of the right edge of each bin.
+    right_inclusive : bool
+        Whether or not the right edge is inclusive.
+
+    Returns
+    -------
+    Column
+        Column of integer labels of the elements in `input`
+        according to the specified bins.
+    """
+    cdef unique_ptr[column] c_result
+    cdef inclusive c_left_inclusive = (
+        inclusive.YES if left_inclusive else inclusive.NO
+    )
+    cdef inclusive c_right_inclusive = (
+        inclusive.YES if right_inclusive else inclusive.NO
+    )
+
+    with nogil:
+        c_result = move(
+            cpp_labeling.label_bins(
+                input.view(),
+                left_edges.view(),
+                c_left_inclusive,
+                right_edges.view(),
+                c_right_inclusive,
+            )
+        )
+
+    return Column.from_libcudf(move(c_result))
diff --git a/python/pylibcudf/pylibcudf/libcudf/labeling.pxd b/python/pylibcudf/pylibcudf/libcudf/labeling.pxd
index ec6ef6b2a41..026ea0df9ef 100644
--- a/python/pylibcudf/pylibcudf/libcudf/labeling.pxd
+++ b/python/pylibcudf/pylibcudf/libcudf/labeling.pxd
@@ -1,14 +1,14 @@
 # Copyright (c) 2021-2024, NVIDIA CORPORATION.
-
+from libcpp cimport bool
 from libcpp.memory cimport unique_ptr
 from pylibcudf.libcudf.column.column cimport column
 from pylibcudf.libcudf.column.column_view cimport column_view
 
 
 cdef extern from "cudf/labeling/label_bins.hpp" namespace "cudf" nogil:
-    ctypedef enum inclusive:
-        YES "cudf::inclusive::YES"
-        NO "cudf::inclusive::NO"
+    cpdef enum class inclusive(bool):
+        YES
+        NO
 
     cdef unique_ptr[column] label_bins (
         const column_view &input,
diff --git a/python/pylibcudf/pylibcudf/tests/test_labeling.py b/python/pylibcudf/pylibcudf/tests/test_labeling.py
new file mode 100644
index 00000000000..0a956eab70b
--- /dev/null
+++ b/python/pylibcudf/pylibcudf/tests/test_labeling.py
@@ -0,0 +1,20 @@
+# Copyright (c) 2024, NVIDIA CORPORATION.
+
+import pyarrow as pa
+import pylibcudf as plc
+import pytest
+
+
+@pytest.mark.parametrize("left_inclusive", [True, False])
+@pytest.mark.parametrize("right_inclusive", [True, False])
+def test_label_bins(left_inclusive, right_inclusive):
+    in_col = plc.interop.from_arrow(pa.array([1, 2, 3]))
+    left_edges = plc.interop.from_arrow(pa.array([0, 5]))
+    right_edges = plc.interop.from_arrow(pa.array([4, 6]))
+    result = plc.interop.to_arrow(
+        plc.labeling.label_bins(
+            in_col, left_edges, left_inclusive, right_edges, right_inclusive
+        )
+    )
+    expected = pa.chunked_array([[0, 0, 0]], type=pa.int32())
+    assert result.equals(expected)

From 26a2ac5057db95092b3c51145c0543e81cf090cc Mon Sep 17 00:00:00 2001
From: Matthew Roeschke <10647082+mroeschke@users.noreply.github.com>
Date: Fri, 6 Sep 2024 14:21:01 -0700
Subject: [PATCH 2/4] int instead of bool

---
 python/pylibcudf/pylibcudf/libcudf/labeling.pxd | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/python/pylibcudf/pylibcudf/libcudf/labeling.pxd b/python/pylibcudf/pylibcudf/libcudf/labeling.pxd
index 026ea0df9ef..400c4282f7a 100644
--- a/python/pylibcudf/pylibcudf/libcudf/labeling.pxd
+++ b/python/pylibcudf/pylibcudf/libcudf/labeling.pxd
@@ -1,12 +1,12 @@
 # Copyright (c) 2021-2024, NVIDIA CORPORATION.
-from libcpp cimport bool
+from libcpp cimport int
 from libcpp.memory cimport unique_ptr
 from pylibcudf.libcudf.column.column cimport column
 from pylibcudf.libcudf.column.column_view cimport column_view
 
 
 cdef extern from "cudf/labeling/label_bins.hpp" namespace "cudf" nogil:
-    cpdef enum class inclusive(bool):
+    cpdef enum class inclusive(int):
         YES
         NO
 

From 22b9e547edc046a848a8631b33403e835e8989ea Mon Sep 17 00:00:00 2001
From: Matthew Roeschke <10647082+mroeschke@users.noreply.github.com>
Date: Fri, 6 Sep 2024 16:32:04 -0700
Subject: [PATCH 3/4] Add Inclusive enum, update enum docs

---
 docs/cudf/source/developer_guide/pylibcudf.md   | 17 +++++++++--------
 python/pylibcudf/pylibcudf/labeling.pxd         |  1 +
 python/pylibcudf/pylibcudf/labeling.pyx         | 10 ++++++++--
 .../pylibcudf/pylibcudf/libcudf/CMakeLists.txt  |  4 ++--
 python/pylibcudf/pylibcudf/libcudf/labeling.pyx |  0
 .../pylibcudf/pylibcudf/tests/test_labeling.py  |  5 +++++
 6 files changed, 25 insertions(+), 12 deletions(-)
 create mode 100644 python/pylibcudf/pylibcudf/libcudf/labeling.pyx

diff --git a/docs/cudf/source/developer_guide/pylibcudf.md b/docs/cudf/source/developer_guide/pylibcudf.md
index 4e10459fe2b..39840e72e21 100644
--- a/docs/cudf/source/developer_guide/pylibcudf.md
+++ b/docs/cudf/source/developer_guide/pylibcudf.md
@@ -186,7 +186,7 @@ Here is an example of appropriate enum usage.
 
 
 ```cython
-# cpp/copying.pxd
+# pylibcudf/libcudf/copying.pxd
 cdef extern from "cudf/copying.hpp" namespace "cudf" nogil:
     # cpdef here so that we export both a cdef enum class and a Python enum.Enum.
     cpdef enum class out_of_bounds_policy(bool):
@@ -194,8 +194,9 @@ cdef extern from "cudf/copying.hpp" namespace "cudf" nogil:
         DONT_CHECK
 
 
-# cpp/copying.pyx
-# This file is empty, but is required to compile the Python enum in cpp/copying.pxd
+# pylibcudf/libcudf/copying.pyx
+# This file is empty, but is required to compile the Python enum in pylibcudf/libcudf/copying.pxd
+# Ensure this file is included in pylibcudf/libcudf/CMakeLists.txt
 
 
 # pylibcudf/copying.pxd
@@ -203,21 +204,21 @@ cdef extern from "cudf/copying.hpp" namespace "cudf" nogil:
 # cimport the enum using the exact name
 # Once https://github.com/cython/cython/issues/5609 is resolved,
 # this import should instead be
-# from cudf._lib.cpp.copying cimport out_of_bounds_policy as OutOfBoundsPolicy
-from cudf._lib.cpp.copying cimport out_of_bounds_policy
+# from pylibcudf.libcudf.copying cimport out_of_bounds_policy as OutOfBoundsPolicy
+from pylibcudf.libcudf.copying cimport out_of_bounds_policy
 
 
 # pylibcudf/copying.pyx
 # Access cpp.copying members that aren't part of this module's public API via
 # this module alias
-from cudf._lib.cpp cimport copying as cpp_copying
-from cudf._lib.cpp.copying cimport out_of_bounds_policy
+from pylibcudf.libcudf cimport copying as cpp_copying
+from pylibcudf.libcudf.copying cimport out_of_bounds_policy
 
 # This import exposes the enum in the public API of this module.
 # It requires a no-cython-lint tag because it will be unused: all typing of
 # parameters etc will need to use the Cython name `out_of_bounds_policy` until
 # the Cython bug is resolved.
-from cudf._lib.cpp.copying import \
+from pylibcudf.libcudf.copying import \
     out_of_bounds_policy as OutOfBoundsPolicy  # no-cython-lint
 ```
 
diff --git a/python/pylibcudf/pylibcudf/labeling.pxd b/python/pylibcudf/pylibcudf/labeling.pxd
index 0415178b3a9..6f8797ae7d3 100644
--- a/python/pylibcudf/pylibcudf/labeling.pxd
+++ b/python/pylibcudf/pylibcudf/labeling.pxd
@@ -1,5 +1,6 @@
 # Copyright (c) 2024, NVIDIA CORPORATION.
 from libcpp cimport bool
+from pylibcudf.libcudf.labeling cimport inclusive
 
 from .column cimport Column
 
diff --git a/python/pylibcudf/pylibcudf/labeling.pyx b/python/pylibcudf/pylibcudf/labeling.pyx
index 8c269b51ca4..b5a7445df36 100644
--- a/python/pylibcudf/pylibcudf/labeling.pyx
+++ b/python/pylibcudf/pylibcudf/labeling.pyx
@@ -6,6 +6,8 @@ from pylibcudf.libcudf cimport labeling as cpp_labeling
 from pylibcudf.libcudf.column.column cimport column
 from pylibcudf.libcudf.labeling cimport inclusive
 
+from pylibcudf.libcudf.labeling import inclusive as Inclusive  # no-cython-lint
+
 from .column cimport Column
 
 
@@ -39,10 +41,14 @@ cpdef Column label_bins(
     """
     cdef unique_ptr[column] c_result
     cdef inclusive c_left_inclusive = (
-        inclusive.YES if left_inclusive else inclusive.NO
+        inclusive.YES
+        if left_inclusive
+        else inclusive.NO
     )
     cdef inclusive c_right_inclusive = (
-        inclusive.YES if right_inclusive else inclusive.NO
+        inclusive.YES
+        if right_inclusive
+        else inclusive.NO
     )
 
     with nogil:
diff --git a/python/pylibcudf/pylibcudf/libcudf/CMakeLists.txt b/python/pylibcudf/pylibcudf/libcudf/CMakeLists.txt
index b04e94f1546..2167616690f 100644
--- a/python/pylibcudf/pylibcudf/libcudf/CMakeLists.txt
+++ b/python/pylibcudf/pylibcudf/libcudf/CMakeLists.txt
@@ -12,8 +12,8 @@
 # the License.
 # =============================================================================
 
-set(cython_sources aggregation.pyx binaryop.pyx copying.pyx expressions.pyx reduce.pyx replace.pyx
-                   round.pyx stream_compaction.pyx types.pyx unary.pyx
+set(cython_sources aggregation.pyx binaryop.pyx copying.pyx expressions.pyx labeling.pyx reduce.pyx
+                   replace.pyx round.pyx stream_compaction.pyx types.pyx unary.pyx
 )
 
 set(linked_libraries cudf::cudf)
diff --git a/python/pylibcudf/pylibcudf/libcudf/labeling.pyx b/python/pylibcudf/pylibcudf/libcudf/labeling.pyx
new file mode 100644
index 00000000000..e69de29bb2d
diff --git a/python/pylibcudf/pylibcudf/tests/test_labeling.py b/python/pylibcudf/pylibcudf/tests/test_labeling.py
index 0a956eab70b..35f33a46837 100644
--- a/python/pylibcudf/pylibcudf/tests/test_labeling.py
+++ b/python/pylibcudf/pylibcudf/tests/test_labeling.py
@@ -18,3 +18,8 @@ def test_label_bins(left_inclusive, right_inclusive):
     )
     expected = pa.chunked_array([[0, 0, 0]], type=pa.int32())
     assert result.equals(expected)
+
+
+def test_Inclusive_enum():
+    assert plc.labeling.Inclusive.YES == 0
+    assert plc.labeling.Inclusive.NO == 1

From 0892610e4311f0aaf13787cc2343799e9542d13d Mon Sep 17 00:00:00 2001
From: Matthew Roeschke <10647082+mroeschke@users.noreply.github.com>
Date: Mon, 9 Sep 2024 09:57:19 -0700
Subject: [PATCH 4/4] Update python/pylibcudf/pylibcudf/tests/test_labeling.py

Co-authored-by: Lawrence Mitchell <wence@gmx.li>
---
 python/pylibcudf/pylibcudf/tests/test_labeling.py | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/python/pylibcudf/pylibcudf/tests/test_labeling.py b/python/pylibcudf/pylibcudf/tests/test_labeling.py
index 35f33a46837..f7fb7463b50 100644
--- a/python/pylibcudf/pylibcudf/tests/test_labeling.py
+++ b/python/pylibcudf/pylibcudf/tests/test_labeling.py
@@ -20,6 +20,6 @@ def test_label_bins(left_inclusive, right_inclusive):
     assert result.equals(expected)
 
 
-def test_Inclusive_enum():
+def test_inclusive_enum():
     assert plc.labeling.Inclusive.YES == 0
     assert plc.labeling.Inclusive.NO == 1