qMerge remote-tracking branch 'upstream/feature/cudf-polars' into fea…

…t/manas_polars_docs

.github/workflows/pr.yaml

@@ -27,6 +27,7 @@ jobs:
       - wheel-tests-cudf
       - wheel-build-cudf-polars
       - wheel-tests-cudf-polars
+      - cudf-polars-polars-tests
       - wheel-build-dask-cudf
       - wheel-tests-dask-cudf
       - devcontainer
@@ -154,6 +155,17 @@ jobs:
       # This always runs, but only fails if this PR touches code in
       # pylibcudf or cudf_polars
       script: "ci/test_wheel_cudf_polars.sh"
+  cudf-polars-polars-tests:
+    needs: wheel-build-cudf-polars
+    secrets: inherit
+    uses: rapidsai/shared-workflows/.github/workflows/[email protected]
+    with:
+      # This selects "ARCH=amd64 + the latest supported Python + CUDA".
+      matrix_filter: map(select(.ARCH == "amd64")) | group_by(.CUDA_VER|split(".")|map(tonumber)|.[0]) | map(max_by([(.PY_VER|split(".")|map(tonumber)), (.CUDA_VER|split(".")|map(tonumber))]))
+      build_type: pull-request
+      # This always runs, but only fails if this PR touches code in
+      # pylibcudf or cudf_polars
+      script: "ci/test_cudf_polars_polars_tests.sh"
   wheel-build-dask-cudf:
     needs: wheel-build-cudf
     secrets: inherit

ci/run_cudf_polars_polars_tests.sh

@@ -0,0 +1,27 @@
+#!/bin/bash
+# Copyright (c) 2024, NVIDIA CORPORATION.
+
+set -euo pipefail
+
+# Support invoking run_cudf_polars_pytests.sh outside the script directory
+# Assumption, polars has been cloned in the root of the repo.
+cd "$(dirname "$(realpath "${BASH_SOURCE[0]}")")"/../polars/
+
+DESELECTED_TESTS=(
+    "tests/unit/test_polars_import.py::test_polars_import" # relies on a polars built in place
+    "tests/unit/streaming/test_streaming_sort.py::test_streaming_sort[True]" # relies on polars built in debug mode
+    "tests/unit/test_cpu_check.py::test_check_cpu_flags_skipped_no_flags" # Mock library error
+    "tests/docs/test_user_guide.py" # No dot binary in CI image
+)
+
+DESELECTED_TESTS=$(printf -- " --deselect %s" "${DESELECTED_TESTS[@]}")
+python -m pytest \
+       --import-mode=importlib \
+       --cache-clear \
+       -m "" \
+       -p cudf_polars.testing.plugin \
+       -v \
+       --tb=short \
+       ${DESELECTED_TESTS} \
+       "$@" \
+       py-polars/tests

ci/test_cudf_polars_polars_tests.sh

@@ -0,0 +1,68 @@
+#!/bin/bash
+# Copyright (c) 2024, NVIDIA CORPORATION.
+
+set -eou pipefail
+
+# We will only fail these tests if the PR touches code in pylibcudf
+# or cudf_polars itself.
+# Note, the three dots mean we are doing diff between the merge-base
+# of upstream and HEAD. So this is asking, "does _this branch_ touch
+# files in cudf_polars/pylibcudf", rather than "are there changes
+# between upstream and this branch which touch cudf_polars/pylibcudf"
+# TODO: is the target branch exposed anywhere in an environment variable?
+if [ -n "$(git diff --name-only origin/branch-24.08...HEAD -- python/cudf_polars/ python/cudf/cudf/_lib/pylibcudf/)" ];
+then
+    HAS_CHANGES=1
+    rapids-logger "PR has changes in cudf-polars/pylibcudf, test fails treated as failure"
+else
+    HAS_CHANGES=0
+    rapids-logger "PR does not have changes in cudf-polars/pylibcudf, test fails NOT treated as failure"
+fi
+
+rapids-logger "Download wheels"
+
+RAPIDS_PY_CUDA_SUFFIX="$(rapids-wheel-ctk-name-gen ${RAPIDS_CUDA_VERSION})"
+RAPIDS_PY_WHEEL_NAME="cudf_polars_${RAPIDS_PY_CUDA_SUFFIX}" RAPIDS_PY_WHEEL_PURE="1" rapids-download-wheels-from-s3 ./dist
+
+# Download the cudf built in the previous step
+RAPIDS_PY_WHEEL_NAME="cudf_${RAPIDS_PY_CUDA_SUFFIX}" rapids-download-wheels-from-s3 ./local-cudf-dep
+
+rapids-logger "Install cudf"
+python -m pip install ./local-cudf-dep/cudf*.whl
+
+rapids-logger "Install cudf_polars"
+python -m pip install $(echo ./dist/cudf_polars*.whl)
+
+TAG=$(python -c 'import polars; print(f"py-{polars.__version__}")')
+rapids-logger "Clone polars to ${TAG}"
+git clone https://github.com/pola-rs/polars.git --branch ${TAG} --depth 1
+
+# Install requirements for running polars tests
+rapids-logger "Install polars test requirements"
+python -m pip install -r polars/py-polars/requirements-dev.txt -r polars/py-polars/requirements-ci.txt
+
+function set_exitcode()
+{
+    EXITCODE=$?
+}
+EXITCODE=0
+trap set_exitcode ERR
+set +e
+
+rapids-logger "Run polars tests"
+./ci/run_cudf_polars_polars_tests.sh
+
+trap ERR
+set -e
+
+if [ ${EXITCODE} != 0 ]; then
+    rapids-logger "Running polars test suite FAILED: exitcode ${EXITCODE}"
+else
+    rapids-logger "Running polars test suite PASSED"
+fi
+
+if [ ${HAS_CHANGES} == 1 ]; then
+    exit ${EXITCODE}
+else
+    exit 0
+fi

ci/test_wheel_cudf_polars.sh

@@ -13,15 +13,21 @@ set -eou pipefail
 if [ -n "$(git diff --name-only origin/branch-24.08...HEAD -- python/cudf_polars/ python/cudf/cudf/_lib/pylibcudf/)" ];
 then
     HAS_CHANGES=1
+    rapids-logger "PR has changes in cudf-polars/pylibcudf, test fails treated as failure"
 else
     HAS_CHANGES=0
+    rapids-logger "PR does not have changes in cudf-polars/pylibcudf, test fails NOT treated as failure"
 fi
 
+rapids-logger "Download wheels"
+
 RAPIDS_PY_CUDA_SUFFIX="$(rapids-wheel-ctk-name-gen ${RAPIDS_CUDA_VERSION})"
 RAPIDS_PY_WHEEL_NAME="cudf_polars_${RAPIDS_PY_CUDA_SUFFIX}" RAPIDS_PY_WHEEL_PURE="1" rapids-download-wheels-from-s3 ./dist
 
 # Download the cudf built in the previous step
 RAPIDS_PY_WHEEL_NAME="cudf_${RAPIDS_PY_CUDA_SUFFIX}" rapids-download-wheels-from-s3 ./local-cudf-dep
+
+rapids-logger "Install cudf"
 python -m pip install ./local-cudf-dep/cudf*.whl
 
 rapids-logger "Install cudf_polars"

python/cudf/cudf/_lib/datetime.pyx

@@ -16,6 +16,8 @@ from cudf._lib.pylibcudf.libcudf.scalar.scalar cimport scalar
 from cudf._lib.pylibcudf.libcudf.types cimport size_type
 from cudf._lib.scalar cimport DeviceScalar
 
+import cudf._lib.pylibcudf as plc
+
 
 @acquire_spill_lock()
 def add_months(Column col, Column months):
@@ -37,43 +39,9 @@ def add_months(Column col, Column months):
 
 @acquire_spill_lock()
 def extract_datetime_component(Column col, object field):
-
-    cdef unique_ptr[column] c_result
-    cdef column_view col_view = col.view()
-
-    with nogil:
-        if field == "year":
-            c_result = move(libcudf_datetime.extract_year(col_view))
-        elif field == "month":
-            c_result = move(libcudf_datetime.extract_month(col_view))
-        elif field == "day":
-            c_result = move(libcudf_datetime.extract_day(col_view))
-        elif field == "weekday":
-            c_result = move(libcudf_datetime.extract_weekday(col_view))
-        elif field == "hour":
-            c_result = move(libcudf_datetime.extract_hour(col_view))
-        elif field == "minute":
-            c_result = move(libcudf_datetime.extract_minute(col_view))
-        elif field == "second":
-            c_result = move(libcudf_datetime.extract_second(col_view))
-        elif field == "millisecond":
-            c_result = move(
-                libcudf_datetime.extract_millisecond_fraction(col_view)
-            )
-        elif field == "microsecond":
-            c_result = move(
-                libcudf_datetime.extract_microsecond_fraction(col_view)
-            )
-        elif field == "nanosecond":
-            c_result = move(
-                libcudf_datetime.extract_nanosecond_fraction(col_view)
-            )
-        elif field == "day_of_year":
-            c_result = move(libcudf_datetime.day_of_year(col_view))
-        else:
-            raise ValueError(f"Invalid datetime field: '{field}'")
-
-    result = Column.from_unique_ptr(move(c_result))
+    result = Column.from_pylibcudf(
+        plc.datetime.extract_datetime_component(col.to_pylibcudf(mode="read"), field)
+    )
 
     if field == "weekday":
         # Pandas counts Monday-Sunday as 0-6

python/cudf/cudf/_lib/pylibcudf/datetime.pyx

@@ -4,6 +4,16 @@ from libcpp.utility cimport move
 
 from cudf._lib.pylibcudf.libcudf.column.column cimport column
 from cudf._lib.pylibcudf.libcudf.datetime cimport (
+    day_of_year as cpp_day_of_year,
+    extract_day as cpp_extract_day,
+    extract_hour as cpp_extract_hour,
+    extract_microsecond_fraction as cpp_extract_microsecond_fraction,
+    extract_millisecond_fraction as cpp_extract_millisecond_fraction,
+    extract_minute as cpp_extract_minute,
+    extract_month as cpp_extract_month,
+    extract_nanosecond_fraction as cpp_extract_nanosecond_fraction,
+    extract_second as cpp_extract_second,
+    extract_weekday as cpp_extract_weekday,
     extract_year as cpp_extract_year,
 )
 
@@ -31,3 +41,42 @@ cpdef Column extract_year(
     with nogil:
         result = move(cpp_extract_year(values.view()))
     return Column.from_libcudf(move(result))
+
+
+def extract_datetime_component(Column col, str field):
+
+    cdef unique_ptr[column] c_result
+
+    with nogil:
+        if field == "year":
+            c_result = move(cpp_extract_year(col.view()))
+        elif field == "month":
+            c_result = move(cpp_extract_month(col.view()))
+        elif field == "day":
+            c_result = move(cpp_extract_day(col.view()))
+        elif field == "weekday":
+            c_result = move(cpp_extract_weekday(col.view()))
+        elif field == "hour":
+            c_result = move(cpp_extract_hour(col.view()))
+        elif field == "minute":
+            c_result = move(cpp_extract_minute(col.view()))
+        elif field == "second":
+            c_result = move(cpp_extract_second(col.view()))
+        elif field == "millisecond":
+            c_result = move(
+                cpp_extract_millisecond_fraction(col.view())
+            )
+        elif field == "microsecond":
+            c_result = move(
+                cpp_extract_microsecond_fraction(col.view())
+            )
+        elif field == "nanosecond":
+            c_result = move(
+                cpp_extract_nanosecond_fraction(col.view())
+            )
+        elif field == "day_of_year":
+            c_result = move(cpp_day_of_year(col.view()))
+        else:
+            raise ValueError(f"Invalid datetime field: '{field}'")
+
+    return Column.from_libcudf(move(c_result))

python/cudf/cudf/pylibcudf_tests/test_datetime.py

@@ -1,16 +1,18 @@
 # Copyright (c) 2024, NVIDIA CORPORATION.
 
 import datetime
+import functools
 
 import pyarrow as pa
+import pyarrow.compute as pc
 import pytest
 from utils import assert_column_eq
 
 import cudf._lib.pylibcudf as plc
 
 
 @pytest.fixture
-def column(has_nulls):
+def date_column(has_nulls):
     values = [
         datetime.date(1999, 1, 1),
         datetime.date(2024, 10, 12),
@@ -22,9 +24,41 @@ def column(has_nulls):
     return plc.interop.from_arrow(pa.array(values, type=pa.date32()))
 
 
-def test_extract_year(column):
-    got = plc.datetime.extract_year(column)
+@pytest.fixture(scope="module", params=["s", "ms", "us", "ns"])
+def datetime_column(has_nulls, request):
+    values = [
+        datetime.datetime(1999, 1, 1),
+        datetime.datetime(2024, 10, 12),
+        datetime.datetime(1970, 1, 1),
+        datetime.datetime(2260, 1, 1),
+        datetime.datetime(2024, 2, 29, 3, 14, 15),
+        datetime.datetime(2024, 2, 29, 3, 14, 15, 999),
+    ]
+    if has_nulls:
+        values[2] = None
+    return plc.interop.from_arrow(
+        pa.array(values, type=pa.timestamp(request.param))
+    )
+
+
+@pytest.mark.parametrize(
+    "component, pc_fun",
+    [
+        ("year", pc.year),
+        ("month", pc.month),
+        ("day", pc.day),
+        ("weekday", functools.partial(pc.day_of_week, count_from_zero=False)),
+        ("hour", pc.hour),
+        ("minute", pc.minute),
+        ("second", pc.second),
+        ("millisecond", pc.millisecond),
+        ("microsecond", pc.microsecond),
+        ("nanosecond", pc.nanosecond),
+    ],
+)
+def test_extraction(datetime_column, component, pc_fun):
+    got = plc.datetime.extract_datetime_component(datetime_column, component)
     # libcudf produces an int16, arrow produces an int64
-    expect = pa.compute.year(plc.interop.to_arrow(column)).cast(pa.int16())
+    expect = pc_fun(plc.interop.to_arrow(datetime_column)).cast(pa.int16())
 
     assert_column_eq(expect, got)