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