Strip ROS2/Docker infrastructure, restructure as standalone Python library

[Copilot]

Converts ROS2 package to standalone Python library containing three SVDD anomaly detection methods.

Changes

Removed infrastructure:

• ROS2 package structure (package.xml, setup.py, setup.cfg, colcon build files)
• ROS2 nodes and utilities (monitor.py, record_features.py, ms_svdd_node.py, rosbag_to_csv.py)
• Docker/devcontainer configuration (.devcontainer/, .dockerignore)
• GitHub Actions workflows (ros-ci.yml, build-docker.yml, dev-container.yml, pytest.yml)
• ROS-specific directories (config/, resource/, test/, rosbags/)

Restructured to root level:

.
├── ae_autoencoder/          # 3 Python files (model, train, eval)
├── ms_svdd/                 # 2 Python files + 2 .pt models + config.yaml + experiment.md
├── svm_svdd/                # 4 Python files (model, train, eval)
├── features.py              # Feature extraction utilities
├── features*.npz            # Pre-extracted datasets
├── svdd_model.py            # Base SVDD model
└── train.py                 # Training utilities

Updated:

• README.md rewritten for Python-only usage (removed all ROS2 instructions)

Net: -2,097 lines of infrastructure, kept 14 Python scripts + 4 model/data files organized by method.

Original prompt

can you take my repo https://github.com/LCAS/ros2_svdd_monitor/tree/copilot/create-ros2-svdd-monitor-package-again and remove all the ROS, devcontainer and docker and just leave folders of python scripts and model files in the folders for each methods tested ae_autoencoder, ms_svdd and svm_svdd

---

✨ Let Copilot coding agent set things up for you — coding agent works faster and does higher quality work when set up for your repo.