Evaluating Power Consumption of Python Test Generation using Pynguin

Prerequisites

• Docker
• Python 3.10.12

Setup Instructions

1. Install EnergiBridge
2. Build the pynguin docker container: docker build -t pynguin-cli .
3. Use ./pynguin.sh or everywhere instead of the normal pynguin command
4. In case of write permission errors, you may need to give permissions to the Docker appuser: sudo chown -R 10001:10001 ./tests/
5. Set up the Python environment using venv or uv

Python Environment Setup w/venv

note: make sure to use Python 3.10.12

6. Create a python virtual environment:

python -m venv .venv

7. Activate the virtual environment:

source .venv/bin/activate

8. Install requirements:

pip install -r requirements.txt

9. Done, run the experiment

Python Environment Setup w/uv

6. Generate a venv:

uv venv --seed

7. Install dependencies:

uv pip install -r requirements.txt

Running the experiment

1. Rename pyenergibridge_config.json.example to pyenergibridge_config.json:

mv pyenergibridge_config.json{.example,}

2. Update the binary path in pyenergibridge_config.json to the installation path of the EnergiBridge binary
3. Create experiment configs in pynguing_configs.json, passing a config name and the pynguin CLI parameters. For example:

[
    //...
    {
        "name": "dynamosa-42",
        "params": {
            "seed": 42,
            "algorithm": "DYNAMOSA"
        }
    }
    //...
]

5. Run the main.py script:

   • with uv:

   uv run python main.py --allowed-projects codetiming_local,docstring_parser_local --iterations 10

   • with uv from the zen script on macos:

   ./run_zen.sh

   • from .venv:

   .venv/bin/python main.py --allowed-projects codetiming_local,docstring_parser_local --iterations 10

For more information on available command-line options:

python main.py --help

Analysis

To collect the results from the json files and run the analysis, use:

python analysis.py