The Trajectools plugin adds trajectory analysis algorithms to the QGIS Processing toolbox.
Trajectools requires MovingPandas (a Python library for movement data analysis) and optionally integrates scikit-mobility (for privacy tests), stonesoup (for smoothing), and gtfs_functions (for GTFS data support).
The recommended way to install these dependencies is through conda/mamba:
(base) conda create -n qgis -c conda-forge python=3.9
(base) conda activate qgis
(qgis) mamba install -c conda-forge qgis movingpandas scikit-mobility stonesoup
(qgis) pip install gtfs_functions
(More details: https://anitagraser.com/2023/01/21/pyqgis-jupyter-notebooks-on-windows-using-conda/)
If you cannot use conda, you may try installing from the QGIS Python Console:
import pip
pip.main(['install', 'movingpandas'])
pip.main(['install', 'scikit-mobility'])
pip.main(['install', 'stonesoup'])
pip.main(['install', 'gtfs_functions'])
The Trajectools plugin can be installed directly in QGIS using the built-in Plugin Manager:
Figure 1: QGIS Plugin Manager with Trajectools plugin installed.
Figure 2: Trajectools (v2.3) algorithms in the QGIS Processing toolbox
The individual Trajectools algorithms are flexible and modular and can therefore be used on a wide array on input datasets, including, for example, the open Microsoft Geolife dataset a sample of which is included in the plugin repo:
Please cite [0] when using Trajectools in your research and reference the appropriate release version using the Zenodo DOI.
@inproceedings{graser2024trajectools,
title = {Trajectools Demo: Towards No-Code Solutions for Movement Data Analytics},
author = {Graser, Anita and Dragaschnig, Melitta},
booktitle = {2024 25th IEEE International Conference on Mobile Data Management (MDM)},
pages = {235--238},
year = {2024},
organization = {IEEE},
doi = {10.1109/MDM61037.2024.00048},
}