iter

azure-pipelines.yml

@@ -144,7 +144,7 @@ jobs:
         THREADPOOLCTL_VERSION: 'min'
         COVERAGE: 'false'
       # Linux + Python 3.8 build with OpenBLAS and without SITE_JOBLIB
-      py38_conda_defaults_openblas:
+      py38_conda_conda_forge_openblas:
         DISTRIB: 'conda'
         CONDA_CHANNEL: 'conda-forge'
         PYTHON_VERSION: '3.8'
@@ -283,13 +283,15 @@ jobs:
         CPU_COUNT: '3'
         TEST_DOCS: 'true'
       # TODO: re-enable when we find out why MKL on defaults segfaults
+      # It seems that scikit-learn from defaults channel is built with LLVM/CLANG OMP
+      # while we use MKL OMP. This could be the cause of the segfaults.
       pylatest_conda_mkl_no_openmp:
         DISTRIB: 'conda'
         BLAS: 'mkl'
         SKLEARN_SKIP_OPENMP_TEST: 'true'
         CPU_COUNT: '3'
         TEST_DOCS: 'true'
-      conda_defaults_openblas:
+      conda_conda_forge_openblas:
         DISTRIB: 'conda'
         CONDA_CHANNEL: 'conda-forge'
         BLAS: 'openblas'

build_tools/azure/install.sh

@@ -18,7 +18,7 @@ make_conda() {
 }
 
 # imports get_dep
-source build_tools/shared.sh
+source build_tools/shared.shpython= pip 'blas[build=mkl]' numpy scipy scikit-learn joblib pandas Pillow matplotlib
 
 if [[ "$DISTRIB" == "conda" || "$DISTRIB" == *"mamba"* ]]; then
 

build_tools/azure/test_script.sh

@@ -12,10 +12,10 @@ mkdir -p $TEST_DIR
 cp setup.cfg $TEST_DIR
 cd $TEST_DIR
 
-python -c "import joblib; print(f'Number of cores (physical): \
-{joblib.cpu_count()} ({joblib.cpu_count(only_physical_cores=True)})')"
+# python -c "import joblib; print(f'Number of cores (physical): \
+# {joblib.cpu_count()} ({joblib.cpu_count(only_physical_cores=True)})')"
+# python -c "import sklearn; sklearn.show_versions()"
 python -c "import imblearn; imblearn.show_versions()"
-python -c "import sklearn; sklearn.show_versions()"
 
 if ! command -v conda &> /dev/null
 then