This is the code repository for the Python implementation of the TRip dIversity Measure - TRIM - which determines the degree of regularity in fine-grained movement behaviour. TRIM effectively quantify the diversity of an individual's movement behaviour between an origin-destination pair. This measure can categorize users based on the diversity of their movement, including in scenarios where the number of individuals' trips is small.
The level of spatial granularity, i.e., size of the grid cells, can be specified in the params.py, where the default granularity is set to 2 km by 2 km cells. The maximum height of the prefix tree, i.e., budget, can also be set in the same file (MAX_B).
The code is using a subset of GeoLife data expressed as a set of grid cell IDs (geoLife-mapped-level*.txt). The distance between every pair of grid cells is also arealy computed and stored in pairDist-level*.csv.