Getting Started with Python

Python is the fast path to building with HORUS. You get the same shared-memory IPC, the same message types, and the same scheduler — with Python's ecosystem of ML frameworks, data tools, and rapid prototyping.

Your Learning Path

Step 1: Install HORUS

curl -fsSL https://raw.githubusercontent.com/softmata/horus/release/install.sh | bash

The installer sets up both the CLI and the Python bindings. See Installation for details.

Step 2: Build Your First App

Follow the Quick Start (Python) — build a publisher and subscriber in 10 minutes.

Step 3: Learn the Concepts

• Nodes: The Building Blocks — how nodes work
• Topics: How Nodes Talk — pub/sub communication
• Scheduler: Running Your Nodes — execution model
• Choosing Your Configuration — progressive complexity guide

Step 4: Tutorials

• Tutorial 1: IMU Sensor Node — read sensor data and publish it
• Tutorial 2: Motor Controller — subscribe to commands, simulate physics
• Tutorial 3: Full Robot System — combine sensors, controller, and estimator
• Tutorial 4: Custom Messages — send typed data between nodes
• Tutorial 5: Hardware & Real-Time — drivers and scheduling

Step 5: Go Deeper

• Python API Reference — complete Node, Scheduler, Topic API
• Python Message Library — 55+ standard robotics types
• Recipes — copy-paste patterns for common tasks

Why Python for HORUS?

• ML/AI integration — use PyTorch, ONNX, TensorFlow, OpenCV directly in your nodes
• Rapid prototyping — iterate on behavior logic without compile cycles
• Same IPC — Python nodes share the same zero-copy topics as Rust nodes
• NumPy interop — Image, PointCloud, and Tensor types expose data as NumPy arrays
• One-liner setup — horus.Node(name, tick, rate) + horus.run(node) and you're running

Python-Specific Guides

• Python Bindings Architecture — how the two-layer Python API works
• ML Integration — PyTorch, ONNX, HuggingFace in HORUS nodes
• Python CV Node Recipe — OpenCV computer vision pipeline
• Async Nodes — async/await support for non-blocking I/O

Need Help?

• Troubleshooting — fix installation and runtime errors
• Choosing a Language — comparing Rust and Python for your project