Firstly, you need to create a virtual environment, activate it, and install opengen using

pip install .

from within python.

To run the benchmarks, you need to install:

pip install pytest-benchmark[histogram]

Then, firstly, create some necessary optimizers to be benchmarked

python prepare_benchmarks.py

Run this from within python/. This will create a number of solvers to be benchmarked; these will be stored in python/.python_test_build/benchmarkable.

Then benchmark them with

py.test test/benchmark_open.py --benchmark-warmup-iterations=20

To generate a fancy histogram (box plots) run

py.test test/benchmark_open.py --benchmark-warmup-iterations=20 --benchmark-histogram=out
# to convert the SVG file to PNG:
qlmanage -t -s 1000 -o . out.svg

The generated benchmark looks like this:

Run

python -W ignore test/test_constraints.py -v
python -W ignore test/test_ros2.py -v
python -W ignore test/test.py -v