Skip to content

Latest commit

 

History

History
82 lines (62 loc) · 2.71 KB

README.md

File metadata and controls

82 lines (62 loc) · 2.71 KB
Raw

MOVEKIT

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

Installation

The easiest way to install movekit is by using pip :

pip install movekit

Docs & Demo

The following website contains the documentation

You can view a demo of common features here: Jupyter Notebooks.

License

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.