RADAR-API

This is the backend API for RADAR-Frontend, a system for interactive exploration of anomaly detection algorithms in different types of data.

The API exposes endpoints to run pipelines, manage data, and execute algorithms provided by the RADAR library.

---

🧩 Features

• Modular pipeline execution
• Integration with the RADAR anomaly detection library
• Support for time series and tabular data
• RESTful API built with FastAPI

---

⚙️ Installation

Prerequisites

• Python: 3.10 (tested with 3.10.18)
• Pytorch: 2.7.1 (tested with with CUDA 11.8 support)
• conda (optional but recommended)

Clone the API (backend) repository

git clone https://github.com/ari-dasci/RADAR-API.git

Clone the RADAR library

git clone https://github.com/ari-dasci/RADAR.git

Install RADAR library + API (conda environment)

conda create --prefix ./envs/radar-env python=3.10.18

conda activate ./envs/radar-env 

conda env update --prefix /path/to/env --file environment.yml --prune   # Replace /path/to/env with the location of your Conda environment.

#Make sure Pytorch is installed now, if you have CUDA 11.8
pip3 install torch torchvision torchaudio --index-url https://download.pytorch.org/whl/cu118

pip install pytorch-lightning

Now install the API with

pip install "fastapi[standard]"
pip install "uvicorn[standard]"

Export path to RADAR library

export PYTHONPATH=«route_to_RADAR»

Run the FastAPI server

uvicorn main:app --reload --host 0.0.0.0 

API will be available at: http://localhost:8000

Swagger documentation: http://localhost:8000/docs

---

🐳 Docker — Full System (Frontend + Backend)

The easiest way to get the complete RADAR system running is via Docker. A single image bundles the Angular frontend, the FastAPI backend, and the RADAR library.

Option 1: Pull from Docker Hub

docker pull beatrizbellogarcia/radar:latest
docker run -d -p 80:80 --name radar beatrizbellogarcia/radar:latest

Then open http://localhost in your browser.

Option 2: Build from source

git clone https://github.com/ari-dasci/RADAR-API.git
cd RADAR-API
docker build -f docker/Dockerfile -t radar-unified .
docker run -d -p 80:80 --name radar radar-unified

This will automatically:

1. Clone and compile the RADAR-Frontend Angular app.
2. Clone the RADAR library (if not already present locally).
3. Install all Python dependencies and bundle everything into a single lightweight image.

Useful commands

# Stop the container
docker stop radar

# Start it again
docker start radar

# View logs
docker logs -f radar

# Remove the container
docker rm -f radar

---

📘 Full Documentation & Frontend Access

Please visit the frontend repository for:

• Full documentation

• Usage examples

• UI for interacting with the API

👉 RADAR-Frontend Repository