Synthetic Housing and Homelessness System Administrative Data

This is a synthetic housing and homelessness administrative data set that approximates system level and shelter level interactions. The data set is designed to be statistically similar to a real North American housing and homelessness system of care. System Level: The data in the system level CSV file contains synthetic records of the housing episodes, shelter episodes and system intake interactions of a synthetic population accessing housing and homelessness services over several years. Shelter Level: The shelter level CSV file contains replaces the shelter episodes in the system level data with daily shelter access records that include the type of shelter services accessed and keyword category counts that represent staff log entries. The software linked below is python code that can be reconfigured to generate data sets that represent housing and homelessness systems for communities of different sizes. This reconfiguration uses parameters that are typically recorded by community based agencies: number of programs, average number of people using each program per day, percentage of people using shelter who go through the system service intake process, etc. The paper entitled " A Parametric Model for Generating Synthetic Homelessness and Housing Administrative Data" can be accessed at https://doi.org/10.48550/arXiv.2611.00260 and describes the synthetic data model implementation.