Add labeling APIs to pylibcudf (#16761)

Contributes to #15162

Authors:
  - Matthew Roeschke (https://github.com/mroeschke)

Approvers:
  - Lawrence Mitchell (https://github.com/wence-)

URL: #16761

docs/cudf/source/developer_guide/pylibcudf.md

@@ -186,38 +186,39 @@ 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):
         NULLIFY
         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
 
 # 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
 ```
 

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)

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

python/pylibcudf/pylibcudf/__init__.pxd

@@ -13,6 +13,7 @@ from . cimport (
     filling,
     groupby,
     join,
+    labeling,
     lists,
     merge,
     null_mask,

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",

python/pylibcudf/pylibcudf/labeling.pxd

@@ -0,0 +1,14 @@
+# Copyright (c) 2024, NVIDIA CORPORATION.
+from libcpp cimport bool
+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
+)

python/pylibcudf/pylibcudf/labeling.pyx

@@ -0,0 +1,65 @@
+# 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 pylibcudf.libcudf.labeling import inclusive as Inclusive  # no-cython-lint
+
+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))

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)

python/pylibcudf/pylibcudf/libcudf/labeling.pxd

@@ -1,14 +1,14 @@
 # Copyright (c) 2021-2024, NVIDIA CORPORATION.
-
+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:
-    ctypedef enum inclusive:
-        YES "cudf::inclusive::YES"
-        NO "cudf::inclusive::NO"
+    cpdef enum class inclusive(int):
+        YES
+        NO
 
     cdef unique_ptr[column] label_bins (
         const column_view &input,

python/pylibcudf/pylibcudf/tests/test_labeling.py

@@ -0,0 +1,25 @@
+# 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)
+
+
+def test_inclusive_enum():
+    assert plc.labeling.Inclusive.YES == 0
+    assert plc.labeling.Inclusive.NO == 1