COMPASS - Software Releases

# VERSION 1.0 – currently under peer-review #Synthetic populations are increasingly used across research and policy applications, particularly in domains where modelling relies on detailed representations of populations and their spatial distribution. Spatially explicit synthetic populations have gained particular importance for analysing place-based interventions, including urban planning, climate mitigation strategies, and the assessment of health inequalities, all of which often require granular spatial resolutions.Despite their growing relevance, the generation of robust real-world data inputs for such applications remains challenging. Existing approaches—most notably those based on simulated annealing—face limitations in terms of robustness, representativeness, and transparency of uncertainty.To address these challenges, we present COMPASS, an open-source software package for the creation of spatially explicit synthetic populations. COMPASS enables the generation of high-resolution population datasets while (1) ensuring aggregate-level representativeness across multiple scales (e.g. individuals and households), (2) preserving harmonised household and kinship structures, and (3) systematically monitoring uncertainty arising from the population synthesis process.COMPASS is provided as a performance-optimised, multi-platform software framework and is designed for seamless integration with common statistical environments, including R and Python.