Replication Package - Measuring Software Resilience Using Socially Aware Truck Factor Estimation

Alexis Butler, Dan O'Keeffe, Santanu Kumar Dash

Contents

• ./data_collection → Script to collate VCS data for the target projects
• ./analysis → Script to run baseline and proposed estimators against the full dataset
• ./post_processing → Script to clean estimator output - catch contributor duplication that arises from alises missed by automated alias grouping
• ./indepth_eval → Script to evaluate the performance of the estimators
• ./shared_models → Collection of datamodels used by the various scripts

Requirements

• Python 3.8
• A Python Virtual Environment manager (conda etc.)

Setup

Dataset

• Download the zipped dataset from Zenodo: https://zenodo.org/records/15223467
• Un-zip the dataset (result will contain a folder called 'output' and a json file truck_factors.json)
• move the folder output into data_collection
• move the json file truck_factors.json to the root of this repo

Scripts

• Create a Python3.8 virtual environment
• Install dependencies from requirements.txt

Usage Notes

• Scripts are inter-dependant:
  • data_collection.main → analysis.main → post_processing.main → indepth_eval.main
• assuming dataset is downloaded execution can start from analysis.main
• post_processing and indepth_eval scripts require human input as directed in the console

Contact

Please raise any issues or questions using the built-in GitHub Issue system, Alexis will address them in due course.

Paper

Raw Bibtex cite to paper - TBC