Merge branch 'branch-24.12' into bug/fix-ewm-mean

.github/labeler.yml

@@ -12,7 +12,7 @@ cudf.polars:
   - 'python/cudf_polars/**'
 
 pylibcudf:
-  - 'python/cudf/pylibcudf/**'
+  - 'python/pylibcudf/**'
 
 libcudf:
   - 'cpp/**'

.github/workflows/build.yaml

@@ -62,7 +62,7 @@ jobs:
       arch: "amd64"
       branch: ${{ inputs.branch }}
       build_type: ${{ inputs.build_type || 'branch' }}
-      container_image: "rapidsai/ci-conda:latest"
+      container_image: "rapidsai/ci-conda:cuda12.5.1-ubuntu22.04-py3.11"
       date: ${{ inputs.date }}
       node_type: "gpu-v100-latest-1"
       run_script: "ci/build_docs.sh"

.github/workflows/pr.yaml

@@ -30,6 +30,7 @@ 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
@@ -49,6 +50,7 @@ jobs:
       test_java: ${{ steps.changed-files.outputs.java_any_changed == 'true' }}
       test_notebooks: ${{ steps.changed-files.outputs.notebooks_any_changed == 'true' }}
       test_python: ${{ steps.changed-files.outputs.python_any_changed == 'true' }}
+      test_cudf_pandas: ${{ steps.changed-files.outputs.cudf_pandas_any_changed == 'true' }}
     steps:
       - name: Get PR info
         id: get-pr-info
@@ -81,6 +83,7 @@ jobs:
               - '!java/**'
               - '!notebooks/**'
               - '!python/**'
+              - '!ci/cudf_pandas_scripts/**'
             java:
               - '**'
               - '!CONTRIBUTING.md'
@@ -89,11 +92,13 @@ jobs:
               - '!img/**'
               - '!notebooks/**'
               - '!python/**'
+              - '!ci/cudf_pandas_scripts/**'
             notebooks:
               - '**'
               - '!CONTRIBUTING.md'
               - '!README.md'
               - '!java/**'
+              - '!ci/cudf_pandas_scripts/**'
             python:
               - '**'
               - '!CONTRIBUTING.md'
@@ -102,6 +107,16 @@ jobs:
               - '!img/**'
               - '!java/**'
               - '!notebooks/**'
+              - '!ci/cudf_pandas_scripts/**'
+            cudf_pandas:
+              - '**'
+              - 'ci/cudf_pandas_scripts/**'
+              - '!CONTRIBUTING.md'
+              - '!README.md'
+              - '!docs/**'
+              - '!img/**'
+              - '!java/**'
+              - '!notebooks/**'
   checks:
     secrets: inherit
     uses: rapidsai/shared-workflows/.github/workflows/[email protected]
@@ -159,7 +174,7 @@ jobs:
       build_type: pull-request
       node_type: "gpu-v100-latest-1"
       arch: "amd64"
-      container_image: "rapidsai/ci-conda:latest"
+      container_image: "rapidsai/ci-conda:cuda12.5.1-ubuntu22.04-py3.11"
       run_script: "ci/test_java.sh"
   static-configure:
     needs: checks
@@ -180,7 +195,7 @@ jobs:
       build_type: pull-request
       node_type: "gpu-v100-latest-1"
       arch: "amd64"
-      container_image: "rapidsai/ci-conda:latest"
+      container_image: "rapidsai/ci-conda:cuda12.5.1-ubuntu22.04-py3.11"
       run_script: "ci/test_notebooks.sh"
   docs-build:
     needs: conda-python-build
@@ -190,7 +205,7 @@ jobs:
       build_type: pull-request
       node_type: "gpu-v100-latest-1"
       arch: "amd64"
-      container_image: "rapidsai/ci-conda:latest"
+      container_image: "rapidsai/ci-conda:cuda12.5.1-ubuntu22.04-py3.11"
       run_script: "ci/build_docs.sh"
   wheel-build-libcudf:
     needs: checks
@@ -244,6 +259,17 @@ 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
@@ -277,7 +303,7 @@ jobs:
     needs: [wheel-build-cudf, changed-files]
     secrets: inherit
     uses: rapidsai/shared-workflows/.github/workflows/[email protected]
-    if: needs.changed-files.outputs.test_python == 'true'
+    if: needs.changed-files.outputs.test_python == 'true' || needs.changed-files.outputs.test_cudf_pandas == 'true'
     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))]))
@@ -288,7 +314,7 @@ jobs:
     needs: [wheel-build-cudf, changed-files]
     secrets: inherit
     uses: rapidsai/shared-workflows/.github/workflows/[email protected]
-    if: needs.changed-files.outputs.test_python == 'true'
+    if: needs.changed-files.outputs.test_python == 'true' || needs.changed-files.outputs.test_cudf_pandas == 'true'
     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))]))

.github/workflows/test.yaml

@@ -41,7 +41,7 @@ jobs:
       sha: ${{ inputs.sha }}
       node_type: "gpu-v100-latest-1"
       arch: "amd64"
-      container_image: "rapidsai/ci-conda:latest"
+      container_image: "rapidsai/ci-conda:cuda12.5.1-ubuntu22.04-py3.11"
       run_script: "ci/test_cpp_memcheck.sh"
   static-configure:
     secrets: inherit
@@ -81,7 +81,7 @@ jobs:
       sha: ${{ inputs.sha }}
       node_type: "gpu-v100-latest-1"
       arch: "amd64"
-      container_image: "rapidsai/ci-conda:latest"
+      container_image: "rapidsai/ci-conda:cuda12.5.1-ubuntu22.04-py3.11"
       run_script: "ci/test_java.sh"
   conda-notebook-tests:
     secrets: inherit
@@ -93,7 +93,7 @@ jobs:
       sha: ${{ inputs.sha }}
       node_type: "gpu-v100-latest-1"
       arch: "amd64"
-      container_image: "rapidsai/ci-conda:latest"
+      container_image: "rapidsai/ci-conda:cuda12.5.1-ubuntu22.04-py3.11"
       run_script: "ci/test_notebooks.sh"
   wheel-tests-cudf:
     secrets: inherit

build.sh

@@ -17,13 +17,14 @@ ARGS=$*
 # script, and that this script resides in the repo dir!
 REPODIR=$(cd $(dirname $0); pwd)
 
-VALIDARGS="clean libcudf pylibcudf cudf cudfjar dask_cudf benchmarks tests libcudf_kafka cudf_kafka custreamz -v -g -n --pydevelop -l --allgpuarch --disable_nvtx --opensource_nvcomp  --show_depr_warn --ptds -h --build_metrics --incl_cache_stats --disable_large_strings"
-HELP="$0 [clean] [libcudf] [pylibcudf] [cudf] [cudfjar] [dask_cudf] [benchmarks] [tests] [libcudf_kafka] [cudf_kafka] [custreamz] [-v] [-g] [-n] [-h] [--cmake-args=\\\"<args>\\\"]
+VALIDARGS="clean libcudf pylibcudf cudf cudf_polars cudfjar dask_cudf benchmarks tests libcudf_kafka cudf_kafka custreamz -v -g -n --pydevelop -l --allgpuarch --disable_nvtx --opensource_nvcomp  --show_depr_warn --ptds -h --build_metrics --incl_cache_stats --disable_large_strings"
+HELP="$0 [clean] [libcudf] [pylibcudf] [cudf] [cudf_polars] [cudfjar] [dask_cudf] [benchmarks] [tests] [libcudf_kafka] [cudf_kafka] [custreamz] [-v] [-g] [-n] [-h] [--cmake-args=\\\"<args>\\\"]
    clean                         - remove all existing build artifacts and configuration (start
                                    over)
    libcudf                       - build the cudf C++ code only
    pylibcudf                     - build the pylibcudf Python package
    cudf                          - build the cudf Python package
+   cudf_polars                   - build the cudf_polars Python package
    cudfjar                       - build cudf JAR with static libcudf using devtoolset toolchain
    dask_cudf                     - build the dask_cudf Python package
    benchmarks                    - build benchmarks
@@ -239,11 +240,6 @@ if hasArg --pydevelop; then
     PYTHON_ARGS_FOR_INSTALL="${PYTHON_ARGS_FOR_INSTALL} -e"
 fi
 
-# Append `-DFIND_CUDF_CPP=ON` to EXTRA_CMAKE_ARGS unless a user specified the option.
-if [[ "${EXTRA_CMAKE_ARGS}" != *"DFIND_CUDF_CPP"* ]]; then
-    EXTRA_CMAKE_ARGS="${EXTRA_CMAKE_ARGS} -DFIND_CUDF_CPP=ON"
-fi
-
 if hasArg --disable_large_strings; then
     BUILD_DISABLE_LARGE_STRINGS="ON"
 fi
@@ -358,6 +354,12 @@ if buildAll || hasArg cudf; then
         python ${PYTHON_ARGS_FOR_INSTALL} .
 fi
 
+# Build and install the cudf_polars Python package
+if buildAll || hasArg cudf_polars; then
+
+    cd ${REPODIR}/python/cudf_polars
+    python ${PYTHON_ARGS_FOR_INSTALL} .
+fi
 
 # Build and install the dask_cudf Python package
 if buildAll || hasArg dask_cudf; then

ci/cudf_pandas_scripts/pandas-tests/job-summary.py

@@ -67,20 +67,33 @@ def emoji_failed(x):
 # convert pr_results to a pandas DataFrame and then a markdown table
 pr_df = pd.DataFrame.from_dict(pr_results, orient="index").sort_index()
 main_df = pd.DataFrame.from_dict(main_results, orient="index").sort_index()
-diff_df = pr_df - main_df
-total_usage = pr_df['_slow_function_call'] + pr_df['_fast_function_call']
-pr_df['CPU Usage'] = ((pr_df['_slow_function_call']/total_usage)*100.0).round(1)
-pr_df['GPU Usage'] = ((pr_df['_fast_function_call']/total_usage)*100.0).round(1)
+total_usage = main_df["_slow_function_call"] + main_df["_fast_function_call"]
+main_df["CPU Usage"] = ((main_df["_slow_function_call"] / total_usage) * 100.0).round(1)
+main_df["GPU Usage"] = ((main_df["_fast_function_call"] / total_usage) * 100.0).round(1)
+
+total_usage = pr_df["_slow_function_call"] + pr_df["_fast_function_call"]
+pr_df["CPU Usage"] = ((pr_df["_slow_function_call"] / total_usage) * 100.0).round(1)
+pr_df["GPU Usage"] = ((pr_df["_fast_function_call"] / total_usage) * 100.0).round(1)
+
+cpu_usage_mean = pr_df["CPU Usage"].mean().round(2)
+gpu_usage_mean = pr_df["GPU Usage"].mean().round(2)
+
+gpu_usage_rate_change = abs(pr_df["GPU Usage"].mean() - main_df["GPU Usage"].mean())
+pr_df["CPU Usage"] = pr_df["CPU Usage"].fillna(0)
+pr_df["GPU Usage"] = pr_df["GPU Usage"].fillna(0)
+main_df["CPU Usage"] = main_df["CPU Usage"].fillna(0)
+main_df["GPU Usage"] = main_df["GPU Usage"].fillna(0)
 
-cpu_usage_mean = pr_df['CPU Usage'].mean().round(2)
-gpu_usage_mean = pr_df['GPU Usage'].mean().round(2)
+diff_df = pr_df - main_df
+diff_df["CPU Usage"] = diff_df["CPU Usage"].round(1).fillna(0)
+diff_df["GPU Usage"] = diff_df["GPU Usage"].round(1).fillna(0)
 
-# Add '%' suffix to 'CPU Usage' and 'GPU Usage' columns
-pr_df['CPU Usage'] = pr_df['CPU Usage'].fillna(0).astype(str) + '%'
-pr_df['GPU Usage'] = pr_df['GPU Usage'].fillna(0).astype(str) + '%'
+# Add '%' suffix to "CPU Usage" and "GPU Usage" columns
+pr_df["CPU Usage"] = pr_df["CPU Usage"].astype(str) + "%"
+pr_df["GPU Usage"] = pr_df["GPU Usage"].astype(str) + "%"
 
-pr_df = pr_df[["total", "passed", "failed", "skipped", 'CPU Usage', 'GPU Usage']]
-diff_df = diff_df[["total", "passed", "failed", "skipped"]]
+pr_df = pr_df[["total", "passed", "failed", "skipped", "CPU Usage", "GPU Usage"]]
+diff_df = diff_df[["total", "passed", "failed", "skipped", "CPU Usage", "GPU Usage"]]
 diff_df.columns = diff_df.columns + "_diff"
 diff_df["passed_diff"] = diff_df["passed_diff"].map(emoji_passed)
 diff_df["failed_diff"] = diff_df["failed_diff"].map(emoji_failed)
@@ -99,13 +112,36 @@ def emoji_failed(x):
         "passed_diff": "Passed delta",
         "failed_diff": "Failed delta",
         "skipped_diff": "Skipped delta",
+        "CPU Usage_diff": "CPU Usage delta",
+        "GPU Usage_diff": "GPU Usage delta",
     }
 )
-df = df.sort_values(by=["Failed tests", "Skipped tests"], ascending=False)
-
+df = df.sort_values(by=["CPU Usage delta", "Total tests"], ascending=False)
+df["CPU Usage delta"] = df["CPU Usage delta"].map(emoji_failed)
+df["GPU Usage delta"] = df["GPU Usage delta"].map(emoji_passed)
+df = df[
+    [
+        "Total tests",
+        "CPU Usage delta",
+        "GPU Usage delta",
+        "Passed tests",
+        "Failed tests",
+        "Skipped tests",
+        "CPU Usage",
+        "GPU Usage",
+        "Total delta",
+        "Passed delta",
+        "Failed delta",
+        "Skipped delta",
+    ]
+]
 print(comment)
 print()
-print(f"Average CPU and GPU usage for the tests: {cpu_usage_mean}% and {gpu_usage_mean}%")
+print(
+    f"Average GPU usage: {gpu_usage_mean}% {'an increase' if gpu_usage_rate_change > 0 else 'a decrease'} by {gpu_usage_rate_change}%"
+)
+print()
+print(f"Average CPU usage: {cpu_usage_mean}%")
 print()
 print("Here are the results of running the Pandas tests against this PR:")
 print()

ci/cudf_pandas_scripts/run_tests.sh

@@ -56,10 +56,10 @@ else
 
     echo "" > ./constraints.txt
     if [[ $RAPIDS_DEPENDENCIES == "oldest" ]]; then
-        # `test_python` constraints are for `[test]` not `[cudf-pandas-tests]`
+        # `test_python_cudf_pandas` constraints are for `[test]` not `[cudf-pandas-tests]`
         rapids-dependency-file-generator \
             --output requirements \
-            --file-key test_python \
+            --file-key test_python_cudf_pandas \
             --matrix "cuda=${RAPIDS_CUDA_VERSION%.*};arch=$(arch);py=${RAPIDS_PY_VERSION};dependencies=${RAPIDS_DEPENDENCIES}" \
         | tee ./constraints.txt
     fi

ci/release/update-version.sh

@@ -25,9 +25,9 @@ NEXT_PATCH=$(echo $NEXT_FULL_TAG | awk '{split($0, a, "."); print a[3]}')
 NEXT_SHORT_TAG=${NEXT_MAJOR}.${NEXT_MINOR}
 
 # Need to distutils-normalize the versions for some use cases
-CURRENT_SHORT_TAG_PEP440=$(python -c "from setuptools.extern import packaging; print(packaging.version.Version('${CURRENT_SHORT_TAG}'))")
-NEXT_SHORT_TAG_PEP440=$(python -c "from setuptools.extern import packaging; print(packaging.version.Version('${NEXT_SHORT_TAG}'))")
-PATCH_PEP440=$(python -c "from setuptools.extern import packaging; print(packaging.version.Version('${NEXT_PATCH}'))")
+CURRENT_SHORT_TAG_PEP440=$(python -c "from packaging.version import Version; print(Version('${CURRENT_SHORT_TAG}'))")
+NEXT_SHORT_TAG_PEP440=$(python -c "from packaging.version import Version; print(Version('${NEXT_SHORT_TAG}'))")
+PATCH_PEP440=$(python -c "from packaging.version import Version; print(Version('${NEXT_PATCH}'))")
 
 echo "Preparing release $CURRENT_TAG => $NEXT_FULL_TAG"
 
@@ -45,6 +45,8 @@ sed_runner "s/branch-.*/branch-${NEXT_SHORT_TAG}/g" ci/test_wheel_dask_cudf.sh
 DEPENDENCIES=(
   cudf
   cudf_kafka
+  cugraph
+  cuml
   custreamz
   dask-cuda
   dask-cudf
@@ -57,7 +59,7 @@ DEPENDENCIES=(
   rmm
 )
 for DEP in "${DEPENDENCIES[@]}"; do
-  for FILE in dependencies.yaml conda/environments/*.yaml; do
+  for FILE in dependencies.yaml conda/environments/*.yaml python/cudf/cudf_pandas_tests/third_party_integration_tests/dependencies.yaml; do
     sed_runner "/-.* ${DEP}\(-cu[[:digit:]]\{2\}\)\{0,1\}==/ s/==.*/==${NEXT_SHORT_TAG_PEP440}.*,>=0.0.0a0/g" "${FILE}"
   done
   for FILE in python/*/pyproject.toml; do
@@ -80,6 +82,7 @@ for FILE in .github/workflows/*.yaml .github/workflows/*.yml; do
   sed_runner "s/dask-cuda.git@branch-[^\"\s]\+/dask-cuda.git@branch-${NEXT_SHORT_TAG}/g" "${FILE}"
 done
 sed_runner "s/branch-[0-9]\+\.[0-9]\+/branch-${NEXT_SHORT_TAG}/g" ci/test_wheel_cudf_polars.sh
+sed_runner "s/branch-[0-9]\+\.[0-9]\+/branch-${NEXT_SHORT_TAG}/g" ci/test_cudf_polars_polars_tests.sh
 
 # Java files
 NEXT_FULL_JAVA_TAG="${NEXT_SHORT_TAG}.${PATCH_PEP440}-SNAPSHOT"

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=native \
+       ${DESELECTED_TESTS} \
+       "$@" \
+       py-polars/tests