GPS Mobility Coding Tasks

This folder contains several take-home tasks related to GPS mobility data analysis, along with local reference assets for the Garden City tasks.

Task instructions

• instructions_flows.md: analyze SafeGraph origin-destination flows for Philadelphia.
• instructions_geolife_task.md: analyze GeoLife trajectories and identify candidate home locations in Beijing.
• instructions_ping_and_stop_summarization.md: summarize raw Garden City pings and precomputed stops provided in this repository.
• instructions_synthetic_mobility_task.md: generate synthetic Garden City trajectories with NOMAD, detect stops, attribute them to places, and analyze user patterns. This task intentionally uses Garden City assets packaged with the NOMAD library rather than repo-hosted input data.

Local assets

• gc_data/: raw synthetic Garden City ping data partitioned by date.
• gc_data_stops/: precomputed Garden City stop tables partitioned by user.
• garden_city.geojson: building layer for visualization in the Garden City task.
• garden-city.gpkg: Garden City geopackage included as a local reference copy.
• garden-city-buildings-mercator.parquet: projected Garden City building layer included as a local reference copy.
• generate_synthetic_trajectories.ipynb and generate_synthetic_trajectories.py: optional local starter materials for the synthetic NOMAD task.

Notes

The synthetic mobility task is different from the others: candidates are expected to generate their own synthetic data with NOMAD and to use NOMAD's packaged Garden City resources directly. The local copies in this folder are provided only as reference or optional starter material.