Merge remote-tracking branch 'upstream/main' into fix-memory-errors

betatim · betatim · commit 7531432cd1e3 · 2025-12-05T09:34:16.000+01:00
diff --git a/.github/workflows/pr_assign_author.yml b/.github/workflows/pr_assign_author.yml
@@ -20,4 +20,10 @@ jobs:
           REPOSITORY:  ${{ github.repository }}
         run: |
           # use gh pr edit to add the PR author as assignee
-          gh pr edit "$PR_NUMBER" --add-assignee "$LOGIN" --repo "$REPOSITORY"
+          # fail gracefully if the author cannot be assigned (e.g., not a collaborator)
+          if gh pr edit "$PR_NUMBER" --add-assignee "$LOGIN" --repo "$REPOSITORY"; then
+            echo "Successfully assigned PR #$PR_NUMBER to $LOGIN"
+          else
+            echo "Could not assign PR #$PR_NUMBER to $LOGIN (user may not be a collaborator)"
+            echo "This is expected for bot accounts or external contributors without write access"
+          fi
diff --git a/ci/build_docs.sh b/ci/build_docs.sh
@@ -10,6 +10,9 @@ PYTHON_CHANNEL=$(rapids-download-conda-from-github python)
 rapids-logger "Create test conda environment"
 . /opt/conda/etc/profile.d/conda.sh
 
+rapids-logger "Configuring conda strict channel priority"
+conda config --set channel_priority strict
+
 RAPIDS_VERSION_MAJOR_MINOR="$(rapids-version-major-minor)"
 export RAPIDS_VERSION_MAJOR_MINOR
 
diff --git a/ci/check_style.sh b/ci/check_style.sh
@@ -7,6 +7,9 @@ set -euo pipefail
 rapids-logger "Create checks conda environment"
 . /opt/conda/etc/profile.d/conda.sh
 
+rapids-logger "Configuring conda strict channel priority"
+conda config --set channel_priority strict
+
 ENV_YAML_DIR="$(mktemp -d)"
 
 rapids-dependency-file-generator \
diff --git a/ci/run_clang_tidy.sh b/ci/run_clang_tidy.sh
@@ -1,12 +1,15 @@
 #!/bin/bash
-# SPDX-FileCopyrightText: Copyright (c) 2023-2024, NVIDIA CORPORATION.
+# SPDX-FileCopyrightText: Copyright (c) 2023-2025, NVIDIA CORPORATION.
 # SPDX-License-Identifier: Apache-2.0
 
 set -euo pipefail
 
 rapids-logger "Create clang_tidy conda environment"
 . /opt/conda/etc/profile.d/conda.sh
 
+rapids-logger "Configuring conda strict channel priority"
+conda config --set channel_priority strict
+
 rapids-dependency-file-generator \
   --output conda \
   --file-key clang_tidy \
diff --git a/ci/test_notebooks.sh b/ci/test_notebooks.sh
@@ -5,6 +5,9 @@ set -euo pipefail
 
 . /opt/conda/etc/profile.d/conda.sh
 
+rapids-logger "Configuring conda strict channel priority"
+conda config --set channel_priority strict
+
 rapids-logger "Downloading artifacts from previous jobs"
 CPP_CHANNEL=$(rapids-download-conda-from-github cpp)
 PYTHON_CHANNEL=$(rapids-download-conda-from-github python)
diff --git a/ci/test_python_integration.sh b/ci/test_python_integration.sh
@@ -13,7 +13,7 @@ trap "EXITCODE=1" ERR
 set +e
 
 rapids-logger "pytest cuml integration"
-./ci/run_cuml_integration_pytests.sh \
+timeout 1h ./ci/run_cuml_integration_pytests.sh \
   --numprocesses=8 \
   --dist=worksteal \
   --junitxml="${RAPIDS_TESTS_DIR}/junit-cuml.xml" \
diff --git a/ci/test_python_singlegpu.sh b/ci/test_python_singlegpu.sh
@@ -21,7 +21,7 @@ trap "EXITCODE=1" ERR
 set +e
 
 rapids-logger "pytest cuml single GPU"
-timeout 2h ./ci/run_cuml_singlegpu_pytests.sh \
+timeout 1h ./ci/run_cuml_singlegpu_pytests.sh \
   --numprocesses=8 \
   --dist=worksteal \
   --junitxml="${RAPIDS_TESTS_DIR}/junit-cuml.xml" \
@@ -30,7 +30,7 @@ timeout 2h ./ci/run_cuml_singlegpu_pytests.sh \
   --cov-report=xml:"${RAPIDS_COVERAGE_DIR}/cuml-coverage.xml"
 
   rapids-logger "pytest cuml accelerator"
-timeout 2h ./ci/run_cuml_singlegpu_accel_pytests.sh \
+timeout 15m ./ci/run_cuml_singlegpu_accel_pytests.sh \
   --numprocesses=8 \
   --dist=worksteal \
   --junitxml="${RAPIDS_TESTS_DIR}/junit-cuml-accel.xml" \
diff --git a/ci/test_wheel.sh b/ci/test_wheel.sh
@@ -55,7 +55,7 @@ timeout 1h ./ci/run_cuml_singlegpu_pytests.sh \
   --junitxml="${RAPIDS_TESTS_DIR}/junit-cuml.xml"
 
 # Run test_sparse_pca_inputs separately
-timeout 1h ./ci/run_cuml_singlegpu_pytests.sh \
+timeout 10m ./ci/run_cuml_singlegpu_pytests.sh \
   -k 'test_sparse_pca_inputs' \
   --junitxml="${RAPIDS_TESTS_DIR}/junit-cuml-sparse-pca.xml"
 
diff --git a/conda/environments/all_cuda-129_arch-aarch64.yaml b/conda/environments/all_cuda-129_arch-aarch64.yaml
@@ -71,7 +71,7 @@ dependencies:
 - rmm==26.2.*,>=0.0.0a0
 - scikit-build-core>=0.10.0
 - scikit-learn>=1.4
-- scipy>=1.8.0
+- scipy>=1.11.0
 - seaborn
 - sphinx
 - sphinx-copybutton
diff --git a/conda/environments/all_cuda-129_arch-x86_64.yaml b/conda/environments/all_cuda-129_arch-x86_64.yaml
@@ -71,7 +71,7 @@ dependencies:
 - rmm==26.2.*,>=0.0.0a0
 - scikit-build-core>=0.10.0
 - scikit-learn>=1.4
-- scipy>=1.8.0
+- scipy>=1.11.0
 - seaborn
 - sphinx
 - sphinx-copybutton
diff --git a/conda/environments/all_cuda-130_arch-aarch64.yaml b/conda/environments/all_cuda-130_arch-aarch64.yaml
@@ -71,7 +71,7 @@ dependencies:
 - rmm==26.2.*,>=0.0.0a0
 - scikit-build-core>=0.10.0
 - scikit-learn>=1.4
-- scipy>=1.8.0
+- scipy>=1.11.0
 - seaborn
 - sphinx
 - sphinx-copybutton
diff --git a/conda/environments/all_cuda-130_arch-x86_64.yaml b/conda/environments/all_cuda-130_arch-x86_64.yaml
@@ -71,7 +71,7 @@ dependencies:
 - rmm==26.2.*,>=0.0.0a0
 - scikit-build-core>=0.10.0
 - scikit-learn>=1.4
-- scipy>=1.8.0
+- scipy>=1.11.0
 - seaborn
 - sphinx
 - sphinx-copybutton
diff --git a/conda/recipes/cuml/recipe.yaml b/conda/recipes/cuml/recipe.yaml
@@ -99,6 +99,7 @@ requirements:
     - numba-cuda >=0.19.1,<0.20.0a0
     - numpy >=1.23,<3.0a0
     - scikit-learn >=1.4
+    - scipy >=1.11.0
     - packaging
     - pylibraft =${{ minor_version }}
     - python
diff --git a/cpp/CMakeLists.txt b/cpp/CMakeLists.txt
@@ -514,8 +514,7 @@ if(BUILD_CUML_CPP_LIBRARY)
 
   if(all_algo OR spectralclustering_algo)
     target_sources(
-      cuml_objs PRIVATE src/spectral/spectral.cu src/spectral/spectral_clustering.cu
-                        src/spectral/spectral_embedding.cu
+      cuml_objs PRIVATE src/spectral/spectral_clustering.cu src/spectral/spectral_embedding.cu
     )
   endif()
 
diff --git a/cpp/include/cuml/cluster/spectral.hpp b/cpp/include/cuml/cluster/spectral.hpp
diff --git a/cpp/src/pca/pca_mg.cu b/cpp/src/pca/pca_mg.cu
@@ -78,6 +78,9 @@ void fit_impl(raft::handle_t& handle,
                            n_streams,
                            true);
   } else {
+    for (std::uint32_t i = 0; i < n_streams; i++) {
+      handle.sync_stream(streams[i]);
+    }
     signFlipComponents(handle,
                        input_data[0]->ptr,
                        components,
@@ -191,6 +194,9 @@ void fit_impl(raft::handle_t& handle,
                              n_streams,
                              true);
     } else {
+      for (std::uint32_t i = 0; i < n_streams; i++) {
+        handle.sync_stream(streams[i]);
+      }
       signFlipComponents(h,
                          input_data[0]->ptr,
                          vMatrix.data(),
diff --git a/cpp/src/spectral/spectral.cu b/cpp/src/spectral/spectral.cu
diff --git a/dependencies.yaml b/dependencies.yaml
@@ -155,6 +155,7 @@ files:
       - py_run_cuml
       - py_version
       - test_notebooks
+      - test_python_xgboost
   py_build_cuml:
     output: pyproject
     pyproject_dir: python/cuml
@@ -327,7 +328,7 @@ dependencies:
           - joblib>=0.11
           - numba>=0.60.0,<0.62.0a0
           - &numpy numpy>=1.23,<3.0a0
-          - scipy>=1.8.0
+          - scipy>=1.11.0
           - packaging
           - rich
           - &scikit_learn scikit-learn>=1.4
@@ -506,10 +507,39 @@ dependencies:
           # included in the "devcontainers" key. The libxgboost package depends
           # on librmm but we do not want to have a package depending on librmm
           # in devcontainers since it should be built from source.
-          - &rapids-xgboost rapids-xgboost==26.2.*,>=0.0.0a0
+          - rapids-xgboost==26.2.*,>=0.0.0a0
+    specific:
       - output_types: [requirements, pyproject]
-        packages:
-          - &xgboost xgboost>=2.1.0
+        matrices:
+          - matrix:
+              cuda: "12.*"
+              cuda_suffixed: "true"
+            packages:
+              - xgboost-cu12>=2.1.0
+          - matrix:
+              arch: aarch64
+              cuda: "13.*"
+              cuda_suffixed: "true"
+            packages:
+              # TODO: always install xgboost-cu13 once there are xgboost-cu13 aarch64 wheels
+              #
+              # xgboost-cu13 originally shipped without aarch64 wheels
+              # Using the unsuffixed name pulls in CUDA 12 packages for some versions and may
+              # at some point in the future pull in CUDA 13 packages.
+              #
+              # ref: https://github.com/rapidsai/cuml/pull/7550#issuecomment-3603279934
+              - &xgboost_unsuffixed xgboost>=2.1.0
+          - matrix:
+              arch: x86_64
+              cuda: "13.*"
+              cuda_suffixed: "true"
+            packages:
+              - xgboost-cu13>=2.1.0
+          # this unsuffixed dependency is useful in environments where xgboost is built from source
+          # (like RAPIDS DLFW builds)
+          - matrix:
+            packages:
+              - *xgboost_unsuffixed
   test_notebooks:
     common:
       - output_types: [conda, requirements]
@@ -523,11 +553,9 @@ dependencies:
       - output_types: conda
         packages:
           - matplotlib-base
-          - *rapids-xgboost
       - output_types: requirements
         packages:
           - matplotlib
-          - *xgboost
   depends_on_cuda_python:
     specific:
       - output_types: [conda, requirements, pyproject]
diff --git a/python/cuml/cuml/decomposition/pca_mg.pyx b/python/cuml/cuml/decomposition/pca_mg.pyx
@@ -2,10 +2,7 @@
 # SPDX-FileCopyrightText: Copyright (c) 2019-2025, NVIDIA CORPORATION.
 # SPDX-License-Identifier: Apache-2.0
 #
-
 import numpy as np
-import sklearn
-from packaging.version import Version
 
 import cuml.internals
 from cuml.decomposition import PCA
@@ -94,7 +91,7 @@ class PCAMG(BaseDecompositionMG, PCA):
         cdef uintptr_t noise_variance_ptr = noise_variance.ptr
         cdef bool use_float32 = (dtype == np.float32)
         cdef handle_t* handle_ = <handle_t*><size_t>self.handle.getHandle()
-        cdef bool flip_signs_based_on_U = (Version(sklearn.__version__) < Version("1.5.0"))
+        cdef bool flip_signs_based_on_U = self._u_based_sign_flip
 
         # Perform fit
         with nogil:
diff --git a/python/cuml/cuml/decomposition/tsvd_mg.pyx b/python/cuml/cuml/decomposition/tsvd_mg.pyx
@@ -2,10 +2,7 @@
 # SPDX-FileCopyrightText: Copyright (c) 2019-2025, NVIDIA CORPORATION.
 # SPDX-License-Identifier: Apache-2.0
 #
-
 import numpy as np
-import sklearn
-from packaging.version import Version
 
 import cuml.internals
 from cuml.decomposition import TruncatedSVD
@@ -100,7 +97,7 @@ class TSVDMG(BaseDecompositionMG, TruncatedSVD):
         cdef uintptr_t singular_values_ptr = singular_values.ptr
         cdef bool use_float32 = dtype == np.float32
         cdef handle_t* handle_ = <handle_t*><size_t>self.handle.getHandle()
-        cdef bool flip_signs_based_on_U = (Version(sklearn.__version__) < Version("1.5.0"))
+        cdef bool flip_signs_based_on_U = self._u_based_sign_flip
 
         # Perform Fit
         with nogil:
diff --git a/python/cuml/pyproject.toml b/python/cuml/pyproject.toml
@@ -93,7 +93,7 @@ dependencies = [
     "rich",
     "rmm==26.2.*,>=0.0.0a0",
     "scikit-learn>=1.4",
-    "scipy>=1.8.0",
+    "scipy>=1.11.0",
     "treelite>=4.6.1,<5.0.0a0",
 ] # This list was generated by `rapids-dependency-file-generator`. To make changes, edit ../../dependencies.yaml and run `rapids-dependency-file-generator`.
 classifiers = [
diff --git a/python/cuml/tests/dask/test_dask_pca.py b/python/cuml/tests/dask/test_dask_pca.py
diff --git a/python/cuml/tests/dask/test_dask_tsvd.py b/python/cuml/tests/dask/test_dask_tsvd.py
