Skip to content

actions/setup-python@v4 and pyyaml==6.0.1 (#172) #80

actions/setup-python@v4 and pyyaml==6.0.1 (#172)

actions/setup-python@v4 and pyyaml==6.0.1 (#172) #80

Workflow file for this run

## Run benchmarks on the three OSs
## We run only on one python version (a recent one)
## I think that different python versions might share hardware resources so results might have a much larger fluctuation
name: Performance benchmarks
on:
push:
branches:
- main
jobs:
benchmarks:
runs-on: ${{ matrix.os }}
strategy:
fail-fast: false
matrix:
os: [ubuntu-latest, macos-latest, windows-latest]
# Do not run on forks
if: github.repository_owner == 'aiidateam'
steps:
- uses: actions/checkout@v2
- name: Set up Python
uses: actions/setup-python@v5
with:
python-version: 3.8
- name: Install dependencies (including dev dependencies at frozen version)
# I'm using pip install -e to make sure that the coverage properly traces the runs
# also of the concurrent tests (maybe we can achieve this differently)
run: |
python -m pip install --upgrade pip
pip install -e .[optionaltests]
pip install -r requirements.lock
- name: Run benchmarks
run: pytest --benchmark-only --benchmark-json output.json
- name: Store benchmark result
## Run only on push on main! Otherwise people (or other branches) might access to the github-actions branch
## This is currently disabled because we are in a workflow that has the correct 'on' settings (only push,
## and only 'main'). Otherwise, enable this
#if: "github.event_name == 'push' && github.ref == 'refs/heads/main'"
uses: rhysd/github-action-benchmark@v1
with:
name: "Benchmark on ${{ matrix.os }}"
tool: "pytest"
output-file-path: output.json
github-token: ${{ secrets.GITHUB_TOKEN }}
auto-push: true
# Show alert with commit comment on detecting possible performance regression
alert-threshold: '200%'
comment-on-alert: true
# For now I don't fail on alert since there is quite some oscillation on the results
# fail-on-alert: true