How-To Guides

This part of the project documentation focuses on a problem-oriented approach. You'll tackle common tasks that you might have, with the help of the code provided in this project.

How is it structured?

Clone the code from this GitHub repository on your preferred directory and following structure will be downloaded:

power-consumption-app/
├───app.py
├── docs/
│   |── explanation.md
│   |── index.md
│   ├── how-to-guides.md
|   └── reference.md
└── src/
    │
    ├── backend/
    │   ├── bash.py
    │   ├── engine.py
    │   └── __init__.py
    └── pyproject.toml

How to start

After cloning the project run the following command to install the necessary dependencies.

    cd power-consumption-app
    pip install power-consumption

How to configure

In the app.py file you will find the follow function:

@measure_energy(handler=csv_handler)
def application():
    """Replace code inside with the script you want to run"""
    solution = 0
    for _ in range(50):
        solution = ((1 + 2 + 3) ** 2) ** 2
    print(solution)

Replace the code inside the function block with your prefer code.

How to use it

Run

    python3 app.py

The console inside the function application() will run 100 times and the results will be saved in a file called result.csv or the name you have chosen.

First column: timestamp -> Timestamp of the start of the iteration Second column: tag -> Name of the function running Third column: duration -> Duration of the iteration Forth column: package_0 -> Total energy in uJ consumed by CPU during the iteration Fifth column: nvidia_gpu_0 -> Total energy in uJ consumed by GPU during the iteration

How to read the final result

The application will print in the console: - How much energy was consume in total - Total duration of all iterations - How much power the application consumes