name: Unit tests

permissions:

contents: read

on:

push:

pull_request:

schedule:

# Nightly build at 02:30 UTC

- cron: "30 2 * * *"

# Manual run

workflow_dispatch:

concurrency:

group: ${{ github.workflow }}-${{ github.head_ref || github.run_id }}

cancel-in-progress: true

env:

VIRTUALENV: testvenv

TEST_DIR: ${{ github.workspace }}/tmp_folder

CCACHE_DIR: ${{ github.workspace }}/ccache

COVERAGE: 'true'

jobs:

lint:

name: Lint

runs-on: ubuntu-latest

if: github.repository == 'scikit-learn/scikit-learn'

steps:

- name: Checkout

uses: actions/checkout@v5

- uses: actions/setup-python@v6

with:

python-version: '3.12'

cache: 'pip'

- name: Install linters

run: |

source build_tools/shared.sh

# Include pytest compatibility with mypy

pip install pytest $(get_dep ruff min) $(get_dep mypy min) cython-lint

- name: Run linters

run: ./build_tools/linting.sh

- name: Run Meson OpenMP checks

run: |

pip install ninja meson scipy

python build_tools/check-meson-openmp-dependencies.py

retrieve-commit-message:

name: Retrieve the latest commit message

runs-on: ubuntu-latest

if: github.repository == 'scikit-learn/scikit-learn'

outputs:

message: ${{ steps.git-log.outputs.message }}

steps:

- uses: actions/checkout@v5

with:

ref: ${{ github.event.pull_request.head.sha }}

- id: git-log

name: Retrieve the latest commit message

shell: bash

run: |

set -eu

message=$(git log --format=%B -n 1)

{

echo 'message<<EOF'

echo "${message}"

echo EOF

} >> "${GITHUB_OUTPUT}"

retrieve-selected-tests:

# Parse the commit message to check if `build_tools/azure/test_script.sh` should run

# only specific tests.

#

# If so, selected tests will be run with SKLEARN_TESTS_GLOBAL_RANDOM_SEED="all".

#

# The commit message must take the form:

# <title> [all random seeds]

# <test_name_1>

# <test_name_2>

# ...

name: Retrieve the selected tests

runs-on: ubuntu-latest

if: github.repository == 'scikit-learn/scikit-learn'

outputs:

tests: ${{ steps.selected-tests.outputs.tests }}

needs: [retrieve-commit-message]

steps:

- id: selected-tests

name: Retrieve the selected tests

shell: python

env:

COMMIT_MESSAGE: ${{ needs.retrieve-commit-message.outputs.message }}

run: |

import os

commit_message = os.environ["COMMIT_MESSAGE"]

# Retrieve selected tests from commit message

if "[all random seeds]" in commit_message:

selected_tests = commit_message.split("[all random seeds]")[1].strip()

selected_tests = selected_tests.replace("\n", " or ")

# quote 'selected_tests' to cover the case of multiple selected tests

selected_tests = f"{selected_tests!r}"

else:

selected_tests = ""

# Write selected tests to `GITHUB_OUTPUT`

with open(os.environ["GITHUB_OUTPUT"], "a") as file:

file.write(f"tests={selected_tests}\n")

unit-tests:

name: ${{ matrix.name }}

runs-on: ${{ matrix.os }}

if: github.repository == 'scikit-learn/scikit-learn'

needs: [lint, retrieve-commit-message, retrieve-selected-tests]

strategy:

# Ensures that all builds run to completion even if one of them fails

fail-fast: false

matrix:

include:

- name: Linux pymin_conda_forge_arm

os: ubuntu-24.04-arm

DISTRIB: conda

LOCK_FILE: build_tools/github/pymin_conda_forge_arm_linux-aarch64_conda.lock

# Linux environment to test the latest available dependencies.

# It runs tests requiring lightgbm, pandas and PyAMG.

- name: Linux pylatest_pip_openblas_pandas

os: ubuntu-24.04

DISTRIB: conda

LOCK_FILE: build_tools/azure/pylatest_pip_openblas_pandas_linux-64_conda.lock

SKLEARN_TESTS_GLOBAL_RANDOM_SEED: 3 # non-default seed

SCIPY_ARRAY_API: 1

CHECK_PYTEST_SOFT_DEPENDENCY: true

SKLEARN_WARNINGS_AS_ERRORS: 1

# disable pytest-xdist to have 1 job where OpenMP and BLAS are not single

# threaded because by default the tests configuration (sklearn/conftest.py)

# makes sure that they are single threaded in each xdist subprocess.

PYTEST_XDIST_VERSION: none

PIP_BUILD_ISOLATION: true

- name: macOS pylatest_conda_forge_arm

os: macOS-15

DISTRIB: conda

LOCK_FILE: build_tools/azure/pylatest_conda_forge_osx-arm64_conda.lock

SKLEARN_TESTS_GLOBAL_RANDOM_SEED: 5 # non-default seed

SCIPY_ARRAY_API: 1

PYTORCH_ENABLE_MPS_FALLBACK: 1

CHECK_PYTEST_SOFT_DEPENDENCY: true

env: ${{ matrix }}

steps:

- name: Checkout

uses: actions/checkout@v5

- name: Create cache for ccache

uses: actions/cache@v4

with:

path: ${{ env.CCACHE_DIR }}

key: ccache-v1-${{ matrix.name }}-${{ hashFiles('**/*.pyx*', '**/*.pxd*', '**/*.pxi*', '**/*.h', '**/*.c', '**/*.cpp', format('{0}', matrix.LOCK_FILE)) }}

restore-keys: ccache-${{ matrix.name }}

- name: Set up conda

uses: conda-incubator/setup-miniconda@v3

with:

miniforge-version: latest

auto-activate-base: true

activate-environment: ""

- name: Build scikit-learn

run: bash -l build_tools/azure/install.sh

- name: Set random seed for nightly/manual runs

if: github.event_name == 'schedule' || github.event_name == 'workflow_dispatch'

run: echo "SKLEARN_TESTS_GLOBAL_RANDOM_SEED=$((RANDOM % 100))" >> $GITHUB_ENV

- name: Run tests

env:

COMMIT_MESSAGE: ${{ needs.retrieve-commit-message.outputs.message }}

SELECTED_TESTS: ${{ needs.retrieve-selected-tests.outputs.tests }}

COVERAGE: ${{ env.COVERAGE == 'true' && needs.retrieve-selected-tests.outputs.tests == ''}}

run: bash -l build_tools/azure/test_script.sh

- name: Run doctests in .py and .rst files

run: bash -l build_tools/azure/test_docs.sh

if: ${{ needs.retrieve-selected-tests.outputs.tests == ''}}

- name: Run pytest soft dependency test

run: bash -l build_tools/azure/test_pytest_soft_dependency.sh

if: ${{ env.CHECK_PYTEST_SOFT_DEPENDENCY == 'true' && needs.retrieve-selected-tests.outputs.tests == ''}}

- name: Combine coverage reports from parallel test runners

run: bash -l build_tools/azure/combine_coverage_reports.sh

if: ${{ env.COVERAGE == 'true' && needs.retrieve-selected-tests.outputs.tests == ''}}

- name: Upload coverage report to Codecov

uses: codecov/codecov-action@v5

if: ${{ env.COVERAGE == 'true' && needs.retrieve-selected-tests.outputs.tests == ''}}

with:

files: ./coverage.xml

token: ${{ secrets.CODECOV_TOKEN }}

disable_search: true

- name: Update tracking issue

if: ${{ always() && (github.event_name == 'schedule' || github.event_name == 'workflow_dispatch')}}

run: |

set -ex

if [[ ${{ job.status }} == "success" ]]; then

TESTS_PASSED=true

else

TESTS_PASSED=false

fi

pip install defusedxml PyGithub

python maint_tools/update_tracking_issue.py \

${{ secrets.BOT_GITHUB_TOKEN }} \

"$GITHUB_WORKFLOW ${{ matrix.name }}" \

"$GITHUB_REPOSITORY" \

https://github.com/$GITHUB_REPOSITORY/actions/runs/$GITHUB_RUN_ID \

--tests-passed $TESTS_PASSED \

--auto-close false \

--job-name "${{ matrix.name }}"