/T-DBSCAN

Python implementation of T-DBSCAN clustering algorithm for Spatiotemporal Density Clustering. Original paper:

Primary LanguagePython

T-DBSCAN - Spatiotemporal Density Clustering for GPS Trajectory Segmentation

This is a pythonic implementation of the T-DBSCAN algorithm. Original paper: "T-DBSCAN: A Spatiotemporal Density Clustering for GPS Trajectory Segmentation".

Steps

  1. Download the tdbscan.py file and place it inside the Lib/site_packages folder
  2. To call the algorithm use df_clustered = tdbscan.T_DBSCAN(df, CEps, Eps, MinPts). Make sure the original data (df) is sorted based on timestamp. CEps is the outer search radius, Eps is the inner search radius and MinPts is the minimum number of points.