/Spatiotemporal-DBSCAN-using-kdTrees

Storing GIS data with temporal information (events data) in k-dimensional trees for efficient querying, designed to aid clustering algorithms such as DBSCAN.

Primary LanguagePythonMIT LicenseMIT

spatiotemporal-kdTree

  1. Imports a file with first 4 columns [uniqueid, longitude, latitude, date (in YYYYMMDD), ...]
  2. forms a kd-Tree based on the lat, lon and date (treats all as numeric)
  3. performs fixed-radius neighborhood search
  4. distance must be a tuple: distance = (spatial distance in km, temporal distance in days)