NASA Human Spaceflight Risk Network

Interactive analysis of NASA's human spaceflight risk DAGs with a FastAPI backend and a React + Cytoscape frontend. The app merges 29 risk graphs, computes structural metrics and DRM-aware rankings, and presents them in a searchable, filterable UI.

Repository layout

• backend/ - FastAPI service (graph parsing/merging, centrality and scoring endpoints).
• frontend/ - React + Vite client for visualization and ranking exploration.
• data/ - All required inputs (*.dag, NODE definition.json, LC score.csv).
• START_BACKEND.bat / START_FRONTEND.bat - Convenience launchers for Windows.
• requirements.txt - Backend Python dependencies (installed into a virtual environment).

Prerequisites

• Python 3.10+ with pip
• Node.js 18+ with npm
• Windows users can rely on the .bat files; macOS/Linux users can follow the manual steps below.

Quick start (Windows)

1. START_BACKEND.bat
   • Creates backend/venv if missing.
   • Installs requirements.txt.
   • Runs backend/test_startup.py to confirm imports and data presence in data/.
   • Starts Uvicorn on http://localhost:8000.
2. In a new terminal, START_FRONTEND.bat
   • Installs frontend dependencies (if node_modules is absent).
   • Starts Vite dev server on http://localhost:5173 with API pointed to localhost:8000.

Quick start (macOS/Linux)

Backend:

cd backend
python -m venv venv
source venv/bin/activate
pip install -r ../requirements.txt
cd src
uvicorn main:app --reload --host 0.0.0.0 --port 8000

Frontend:

cd frontend
npm install
npm run dev -- --host

Configure the API base (if not default) via frontend/.env (VITE_API_BASE=https://human-space-flight-risk-network.onrender.com by default in the example).

Data folder

All required inputs live in data/:

• *.dag - 29 risk DAGs in GraphViz format.
• NODE definition.json - Node definitions and notes.
• LC score.csv - Likelihood x Consequence scores used for DRM-aware ranking.

The backend loads directly from repo_root/data at startup; no database setup is needed.

Backend notes

• Framework: FastAPI + Uvicorn.
• Key modules: DAG parsing/merging (graph/), LxC parsing (data/), metrics and scoring (metrics/), CSV export helpers (utils/export.py).
• Primary endpoints: /api/graph, /api/risks, /api/drms, /api/centrality, /api/score, /api/top10, /api/export/*.
• Recommended runtime: Python 3.11 (Render config pins pythonVersion: 3.11.7 to avoid pandas build issues on 3.13).

Frontend notes

• Stack: React 18, TypeScript, Vite, Cytoscape.js, Axios.
• Entry points: frontend/src/App.tsx with supporting components under frontend/src/components/.
• Expects the API base URL from VITE_API_BASE (defaults to https://human-space-flight-risk-network.onrender.com in .env.example).
• GitHub Pages ready: npm run deploy from frontend/ builds to dist/ and publishes to the gh-pages branch (Vite base is set to /Human-Space-Flight-Risk-Network/ and homepage is configured).

Hosted backend (Render, free tier)

• Push this repo (including data/) to GitHub.
• In Render, create a New Web Service from the repo; it will read render.yaml.
  • Build command: pip install -r requirements.txt
  • Start command: cd backend/src && uvicorn main:app --host 0.0.0.0 --port $PORT
  • Plan: Free (service may sleep when idle).
• After deploy, note the public URL (e.g., https://human-space-flight-risk-network.onrender.com).
• Set VITE_API_BASE in frontend/.env to that URL (the example already matches), then from frontend/ run npm run deploy to republish GitHub Pages so the static site calls the hosted backend.

Housekeeping

• Temporary environments are ignored via .gitignore (backend/venv, frontend/node_modules, caches, etc.).
• No additional documentation is required beyond this README; data artifacts are all in data/.