Atmosphere Ledger

AtmosLedger Atmospheric data platform for ingesting, aggregating, and analyzing weather time-series data.
It provides a clean API, background ingestion workers, and a modern chart-driven UI for exploration. Simply put, pulls and detects anomalous weather.

---

Badges

---

Features

• Ingestion – Async ingestion from Open-Meteo (hourly observations)
• Aggregation – Daily rollups (mean / min / max temperature, precipitation)
• Anomaly Detection – Rolling z-score–based daily anomaly detection
• Scalable Architecture – FastAPI + PostgreSQL + Redis + RQ workers
• Tested – Pytest coverage for core API flows
• UI Demo – React + Recharts frontend for visual exploration

---

Tech Stack

Backend

• Python 3.12
• FastAPI
• SQLAlchemy 2.x, Alembic
• PostgreSQL
• Redis + RQ

Frontend

• React (Vite)
• Recharts

---

Quick Start (Development)

# Start infrastructure + worker
docker compose up -d

# Install backend dependencies
poetry install

# Run API
poetry run uvicorn atmosledger.main:app --reload

# FrontEnd

npm install
npm run dev

Example API Endpoints
http

POST /ingestions/open-meteo/run
POST /series/daily/aggregate
GET  /series/daily
POST /anomalies/daily/detect
GET  /anomalies/daily