Replication Package

Welcome to the replication package of the paper entitled "Are We Welcome Here? A Preliminary Study of Newcomer Onboarding in the ROS Ecosystem"

Repository Structure

• scripts: Complete data processing pipeline
  • 00-11: Core data collection pipeline (download ROS index, build mappings, extract features, apply exclusion criteria)
  • inflow_plot: Descriptive statistics and visualization generation
  • Utility scripts: generate_all_commits_spreadsheet.py, generate_ros_packages_statistics.py
• data: Per-repository data collected from GitHub (metadata, commits, contributors, issues, community files)
• tables: Summarized results and statistics (inflow metrics, repository overlap, cluster assignments)
• figs: Generated visualizations (inflow patterns, documentation metrics)
• config.py: Centralized configuration (paths, colors, visualization styles)
• run_pipeline.py: Main entry point to execute the complete pipeline

Running the Replication Pipeline

To replicate our data collection and analysis:

pip install -r requirements.txt
echo "GITHUB_TOKEN=your_github_token_here" > .env
python run_pipeline.py

The pipeline analyzes ROS packages across three distributions (ROS 2 Humble, Jazzy, and Kilted) and generates the final repository dataset at out/filtered_repo_dataset.csv. Individual scripts can also be run independently if you want to execute specific steps of the pipeline.

Contact

If you have any questions or are interested in contributing to this project, please don't hesitate to contact us:

• Juliana Freitas (jfreit4@lsu.edu)
• Elijah Phifer (ephife3@lsu.edu)
• Nabila Fairuz (nfairu1@lsu.edu)
• Felipe Fronchetti (ffronchetti@lsu.edu)