Data Access Platform for Human Mobility in Epidemiology
DAPHME will provide a repository of processing and analysis methods for GPS mobility data, centralizing tools necessary for large-scale analyses. This repository will serve the dual function of facilitating research for first-time users and enhancing the replicability and robustness of existing methodologies. By collecting many such methods in a single place and documenting their assumptions and robustness metrics, DAPHME will enhance methodological transparency and improve commensurability of alternative approaches.
Objective 3. DAPHME will form the core around which a community of researchers can coalesce. Although we will be the primary developers of initial assets, over time the relevant research community will play an increasing role in refining practices and tackling data challenges.
DAPHME offers PySpark and Python implementations of spatial data processing, human mobility metrics and epidemic modeling including:
- Clustering of GPS data
- Home attribution
- Normalization