Akiba So
fc468464b2
Context: Build a spatial risk assessment system correlating air quality data with children's respiratory disease incidence across Wuhan. Approach: FastAPI backend serving PostGIS spatial queries, React frontend with Deck.gl maps, and a PyTorch SpatialTemporalGCN pipeline for multi-day (1d/3d/7d) risk prediction. Changes: - backend/ — FastAPI API with auth (JWT), alerts, risk analysis, geocoded case data, grid statistics, and report endpoints - frontend/ — React dashboard with interactive risk maps, alert monitoring, district comparison charts, and timeline player - models/ — SpatialTemporalGCN model with trained weights and ONNX export for inference - scripts/ — ETL pipeline for weather + medical data, grid generation, feature engineering, training, and daily inference - deploy/ — Docker Compose configs for backend, frontend, and MLflow - docs/ — API docs, deployment guide, user guide, and code review Impact: Enables spatial risk visualization, alert monitoring, and ML-driven health risk forecasting for environmental health teams.
1.3 KiB
1.3 KiB
Backend — FastAPI + PostGIS
Stack
- FastAPI (async), asyncpg connection pool, Pydantic v2 settings
- PostGIS via GeoAlchemy2, spatial queries with Shapely
- Auth: python-jose + passlib (JWT/bcrypt)
Structure
backend/
main.py # App entry, CORS, router registration
database.py # asyncpg pool, Settings from .env
models.py # Pydantic response/request models
routers/ # One file per domain (risk, alerts, cases, grid, etc.)
app/ # Legacy code (routers/cases.py, routers/grid.py, performance.py)
Patterns
- Routers:
APIRouter()with prefix, registered inmain.pyviaapp.include_router() - DB access:
async with db.get_connection()context manager (globaldbsingleton) - Settings:
pydantic_settings.BaseSettingsloaded from.envat module level - Endpoints return Pydantic models, not raw dicts
Running
cd backend
source venv/bin/activate
uvicorn main:app --reload --port 8000
Anti-Patterns
- Don't use sync database drivers — always asyncpg
- Don't put business logic in routers — delegate to service functions
- Don't hardcode DB credentials — use Settings from environment
- Don't skip Pydantic validation on request/response bodies
- Don't import from
app/— it's legacy, prefer top-level modules