This directory contains benchmark data collected during the experiments of the paper: "Python Under the Microscope: A Comparative Energy Analysis of Execution Methods".
It includes raw measurement logs, aggregated performance metrics, and normalized values across five Python execution methods, covering 15 benchmark algorithms.
The complete dataset is also publicly available on Kaggle:
The Kaggle version is ideal for data analysis, reproducibility testing, and direct import into notebooks or visual tools.
data/
└── collection_1/
├── analysis/ # Final normalized values and GreenScore summaries
├── combine/ # Aggregated energy, time, and carbon CSVs (15 algorithms × 5 methods)
├── cpython/ # Raw CSV logs for CPython runs (energy + time)
├── ctypes/ # Raw CSV logs for ctypes runs
├── cython/ # Raw CSV logs for Cython runs
├── pycompile/ # Raw CSV logs for py_compile runs
├── pypy/ # Raw CSV logs for PyPy runs
└── input.py # Utility script for generating controlled input data
- Raw CSVs: Contain timestamped logs of 50 repeated runs per algorithm–method pair.
- Metrics: Energy (μJ), Execution Time (s), and Estimated Carbon Emissions (gCO₂eq).
- Normalized Data: Used to calculate GreenScore via min–max normalization per algorithm.
- Analysis Folder: Final processed results, including method-wise averages and rankings.
- All files are formatted for pandas-friendly import.
- Folder names match execution method names (
cpython,cython, etc.). - Reproducibility ensured with fixed inputs, consistent machine, and versioned code.
If you use this dataset in your work, please cite:
@misc{md__fatin_shadab_turja_2025,
title={Python Energy Microscope: Benchmarking 5 Execution},
url={https://www.kaggle.com/dsv/12207710},
DOI={10.34740/KAGGLE/DSV/12207710},
publisher={Kaggle},
author={Md. Fatin Shadab Turja},
year={2025}
}