About GPS Synth 🗂️

This Python codebase generates synthetic GPS data based on the properties of a network and users. The code source allows for flexible specification of unique characteristics for a network (e.g., a place of interest) and user parameters (e.g., speed limit).

The basic foundation of the approach is as follows:

• Each use case is denoted as a profile. For each profile, network and user parameters should be specified
• Profiles can share the identical network (if network parameters are the same), but will always have a different movement history for their users (meaning that even if users' parameters are the same, the output trajectories will be different)
• A network is a combination of anchor points/locations and roads/ways that connect locations to each other (essentially, a graph with nodes and edges)
• All locations are derived from OpenStreetMap's conceptual data model of the physical world using the osmnx library
• There are three main types of locations: home, work and event (bars, parks, museums, etc.). Event could be either one of the regular locations of a user or random location (amenity) in the area
• The entire movement history of a user can be described as a consecutive process of stay and movement activities happen in some locations
• The combination of these two types of activities and the type of locations in which they occur are specified by the plot. Each unique movement plot should automatically produce a new child class of the Parent Class User (e.g., User_employed_walk class in ./gps_synth/user/user_employed_walk.py)