Skip to content

README.md

Latest commit

 

History

History
104 lines (70 loc) · 3.79 KB

README.md

File metadata and controls

104 lines (70 loc) · 3.79 KB

Scripts: Data Processing & GreenScore Computation

This directory contains utility scripts used for data aggregation, energy/time/carbon processing, and final GreenScore computation for the study:

"Python Under the Microscope: A Comparative Energy Analysis of Execution Methods."

These scripts enable seamless batch processing and reproducibility of benchmark results.

Directory Overview

scripts/
├── avg_combine.py         # Merge mean energy, time, and carbon files into one
├── carbon.py              # Convert energy (μJ) to carbon emissions (gCO₂e)
├── combine_energy.py      # Combine energy results from all methods into one CSV
├── combine_time.py        # Combine time results from all methods into one CSV
├── energy_avg.py          # Compute per-file average energy usage
├── time_avg.py            # Compute per-file average execution time
├── greenscore.py          # Normalize metrics, compute mean scores, and rank methods by GreenScore
├── std.py                 # Compute standard deviation of energy, time, carbon across methods
└── README.md              # This file

Script Descriptions

Script Description
avg_combine.py Combines final means from energy, time, and carbon CSVs into one unified summary_df.
carbon.py Converts energy consumption (in μJ) to estimated carbon emissions (gCO₂e) using global average.
combine_energy.py Merges energy_avg.csv files from each execution method into one energy comparison table.
combine_time.py Merges time_avg.csv files from each execution method into one time comparison table.
energy_avg.py Averages package (uJ) values across repeated runs in each method/algorithm folder.
time_avg.py Averages execution time from multiple benchmark runs.
greenscore.py Full pipeline for min–max normalization, per-method averaging, and GreenScore computation.
std.py Calculates per-method standard deviation for energy, time, and carbon emissions.

GreenScore Workflow

  1. Run:

    • combine_energy.py
    • combine_time.py
    • carbon.py

  2. Use avg_combine.py to merge the three mean files.

  3. Run greenscore.py to:

    • Normalize each metric (row-wise per algorithm).

    • Average scores across all algorithms per method.

    • Compute GreenScore using the formula:

      $$ \text{GreenScore} = 0.4 \cdot \text{Energy}_{\text{norm}} + 0.4 \cdot \text{Carbon}_{\text{norm}} + 0.2 \cdot \text{Time}_{\text{norm}} $$

  4. Review output files:

    • green_score_components_means.csv
    • green_score_ranking.csv

Example Usage

To generate normalized sustainability rankings:

python greenscore.py

To combine time CSVs from each method:

python combine_time.py combined_time.csv

To calculate carbon footprint from energy:

python carbon.py

Output Format

All scripts produce CSV files with standard field names:

  • algorithm, method, energy_mean_μJ, carbon_mean_gCO2eq, execution_time (s)
  • green_score, energy_std, carbon_std, time_std

These are designed for direct use in:

  • 📊 Visualization scripts
  • 📄 LaTeX tables in the final paper
  • 📁 Kaggle dataset uploads

Notes

  • Ensure consistent folder structure when running combine_*.py scripts.
  • Scripts use relative paths. Adjust paths as needed for your environment.
  • Carbon intensity is set to 0.000475 gCO₂e/J based on global averages.