scikit-mobility is a Python library designed for comprehensive human mobility analysis. It provides specialized data structures, TrajDataFrame and FlowDataFrame, for representing and manipulating trajectory and origin-destination flow data, respectively. The library caters to researchers and data scientists working with diverse mobility datasets, enabling them to extract mobility metrics, generate synthetic trajectories and flows, and assess privacy risks.

How It Works

The library leverages custom pandas extensions, TrajDataFrame and FlowDataFrame, to structure mobility data. TrajDataFrame requires latitude, longitude, and datetime, with optional user and trajectory IDs, facilitating individual movement analysis. FlowDataFrame represents origin-destination flows and is associated with a spatial tessellation (a GeoDataFrame defining geographic regions). This approach allows for efficient data handling and the application of various mobility models and analytical techniques.

Quick Start & Requirements

• Installation: Recommended via conda install -c conda-forge scikit-mobility. Pip installation is possible but may lead to dependency issues, especially on Windows/Mac.
• Prerequisites: Python >= 3.8. GeoPandas dependencies may require specific installation steps.
• Documentation: https://scikit-mobility.github.io/scikit-mobility/
• Tutorials: https://github.com/scikit-mobility/tutorials

Highlighted Details

• Supports trajectory preprocessing: noise filtering, stop detection, stop clustering, and compression.
• Implements individual mobility measures (e.g., radius of gyration, jump lengths) and collective measures.
• Includes generative models for synthetic trajectories (e.g., Exploration and Preferential Return) and flows (e.g., Gravity, Radiation).
• Offers privacy risk assessment tools using location attack models.
• Provides utilities for downloading and loading various mobility datasets.

Maintenance & Community

The project is actively maintained, with contributions welcome. Further community engagement details are not explicitly provided in the README.

Licensing & Compatibility

The README does not explicitly state the license. Compatibility for commercial use or closed-source linking is not specified.

Limitations & Caveats

The README strongly advises against pip installation due to potential dependency issues, particularly on Windows and macOS. Some mobility measures require the TrajDataFrame to be sorted by datetime.