Measuring people and populations
Question. Which measure answers my research question, and what does it mean?
Individual measures describe each person; collective measures describe where and when the whole population moves. To run the individual measures on real data, open Measure individual mobility.
Which measure for which question
| Question | Start with | Unit |
|---|---|---|
| How widely does each person range? | radius_of_gyration, maximum_distance |
km |
| How long are their trips and waits? | jump_lengths, waiting_times |
km, seconds |
| How many places, how evenly used? | number_of_locations, uncorrelated_entropy |
count, bits |
| Which places matter? | frequency_rank, home_location, work_location |
places |
| When do visits occur? | visits_per_time_unit (collective) |
visits per interval |
| Which origins and destinations connect? | od_matrix (collective) |
trips |
Choosing the input
Point-based measures such as the radius of gyration use raw trajectories. Visit-based measures (diversity, regularity, motif measures) need visit tables, such as staypoints. OD measures need trips or an origin-destination layout. Do not average a per-person result by hand when a collective function already defines the population.
Assumptions
Distances are straight lines in kilometres, times are seconds, and hour-of-day measures treat hours in the data's own time zone, so convert to local time first.
Related API
Individual measures · Collective measures · Evaluation measures
Worked examples: Measure individual mobility, Routine, predictability and mobility profiles and Population measures, Flows between places and Occupancy and phone records.