FUT — Papers I–X (auto-figures bootstrap)

This is a clean scaffold of your repo with:

• Papers I–X as standalone LaTeX (each has a self-contained .bib, robust figure handling, references last).
• A workflow that runs your pipeline (make bench, make phase, make figs) and generates the exact paper figures the TeX expects, committing them back into paperN/figs/.
• A starter scripts/generate_paper_figs.py that reads run_report.json and writes the expected filenames for Papers I–III now (extend similarly for IV–X as their data exist).

Quick start (local)

python3 -m venv .venv && source .venv/bin/activate
pip install -r requirements.txt

# Your pipeline (replace with your real steps)
make bench
make phase
make figs

# Generate paper figures from run_report.json
python scripts/generate_paper_figs.py

# Compile a paper (needs LaTeX on your system)
cd paper1 && pdflatex paper_I.tex && bibtex paper_I && pdflatex paper_I.tex && pdflatex paper_I.tex

GitHub Actions (automatic figures)

On any push, the workflow .github/workflows/fig-autobuild.yml will:

1. Set up Python and install deps from requirements.txt.
2. Run make bench && make phase && make figs (replace with your real pipeline).
3. Run python scripts/generate_paper_figs.py to overwrite placeholder images with real ones.
4. Commit updated paper1/figs/*.png, paper2/figs/*.png, paper3/figs/*.png back to the repo.

Extend scripts/generate_paper_figs.py with your outputs for Papers IV–X, then add those paths to the workflow's commit step.

Where to drop your real code

• Put your actual simulation/analysis code under bench/, src/, figs/, and write numeric summaries to run_report.json.
• The figure generator expects keys as shown in the example run_report.json in repo root (edit to match your outputs).

Important

• Figures currently include placeholders generated from run_report.json. Once your real pipeline writes that file (and/or additional CSVs), rerun the workflow and the figures will be replaced automatically.