movekit
is an open-source software package for the processing and analysis of movement data.
movekit
supports different tasks:
- Data pre-processing
- Clean data (remove duplicates, drop missing values, etc.)
- Normalize and filter the data
- ...
- Feature extraction:
- Extract different features such as the distance covered, average speed, the average acceleration, etc.
- Apply time series analysis on these extracted features
- Check different distances (euclidean & hausdorff) between movers
- Detect outliers in data
- ...
- Group-level analysis
- Calculate centroids and medoids of the group of movers for different time steps
- Compute polarization of movers
- Identify different clusters/groups
- Obtain dynamic time warping of all mover trajectories
- ...
- Spatial data analysis:
- Create convex hull, voronoi diagram and delaunay triangulation for all movers at each time step
- Extract areas of the created objects
- Network analysis with networkX
- Create networks created for each time step and examine their attributes (centroid, polarization, total distance, mean speed, ...)
- Investigate individual nodes of each time steps network graph
- Investigate individual edges of each time steps network graph
- Track development of network graphs over time
- ...
- Plotting analysis results:
- Create basic plots for features such as acceleration or speed
- Plot movement of movers in static or animated images
- Create interactive map to plot geo data
movekit
provides support for movement data and trajectories in different format:
Data:
- 2-dimensional data in the Euclidean space
- 3-dimensional data in the Euclidean space
- GPS coordinates (latitude and longitude)
- Data with different time formats
- Data in (Geo)JSON format
- Data from Movebank data base
The easiest way to install movekit is by using pip
:
pip install movekit
The following website contains the documentation
You can view a demo of common features here: Jupyter Notebooks.
Released under a GNU General Public License. See the LICENSE file for details. List of Authors
The package is funded by the Deutsche Forschungsgemeinschaft (DFG, German Research Foundation) under Germany's Excellence Strategy – EXC 2117 – 422037984.