Data for: "Simulating nationwide coupled disease and fear spread in an agent based model"

Abstract This record contains the data to be publicly released in order to support the publication of the paper "Simulating nationwide coupled disease and fear spread in an agent based model". These data consist of binary output files from EpiCast, an agent-based model (ABM) of infectious respiratory disease spread, for a realistic synthetic population representing the contiguous United States (lower 48 states + Washington DC). Each binary file contains counters tracking changes in agent states (such as transitions between disease compartments or changes in behavior) in specific locations over the course of a given simulation run. Simulation Results This data set contains the results of 65 runs of EpiCast on a synthetic population encompassing the contiguous United States. This population is generated using UrbanPop [1], a population synthesis framework from Oak Ridge National Laboratory, which assigns over 322 million agents to nighttime locations in specific US Census tracts. At runtime, EpiCast partitions agents in a tract across a series of ~2000 agent communities to indicate where each resides, and then assigns each agent a potentially seperate community to work at during the day. Simulation files in the scenarios directory are provided in a binary format custom to EpiCast (the .bin files). For the remaining runs, we instead provide comma-seperated value (CSV) files which contain summary statistics for each timestep in the simulation. There are four components to the top-level file: Run name: a string identifying the run saved to the file FIPS: a vector containing the FIPS codes of all n simulated tracts for this run Demographics: a table summarizing the demographics of each simulated tract in the run Data: a table containing counters describing the progression of the simulation for each day in each simulated tract, as represented by changes in agent states Note that all counters described as "cumulative" count the total number of unique agents which have fallen into the specified category over the course of the simulation up to and including the timestep corresponding to the row containing the data, while all counters described as "current" count only those agents in the specified category in a given timestep. This means that "cumulative" counters for a given tract are guaranteed to be monotonically increasing. Run name: a string identifying the run saved to the file FIPS: a vector containing the FIPS codes of all n simulated tracts for this run Demographics: a table summarizing the demographics of each simulated tract in the run Data: a table containing counters describing the progression of the simulation for each day in each simulated tract, as represented by changes in agent states The demographics table has n rows, each describing one simulated tract. The columns are: Total: the total population of the tract Age Group (0-4): the number of agents ≤ 5, 6-17, 18-29, 30-64, and ≥ 65 years old, respectively Household (1-7 Members): the number of agents living in households containing the specified number of members (i.e. 1-7) Race (White, Black, Native, Asian, Native Hawaiian/Pacific Islander, Other, and Multiple): number of agents with the specified race Ethnicity (non-Hispanic, Hispanic): number of agents who are non-Hispanic or Hispanic, respectively Broadcasters: number of broadcasters (radio and television stations) located in this tract The data table has n × t rows, where t is the number of simulated days in the run, and each row is identified by the combination of census tract FIPS code and a simulation day. The columns are: Demographic column (e.g. Age 0, or Race Native): the cumulative number of infected agents belonging to the specified demographic column. This can be any column in the demographics table other than broadcasters. Treatment status (Hospitalized, ICU, Ventilated, and Dead) in Age Group (0-4): cumulative number of agents who have had their treatment progress to the specified status (including failure; i.e. the death of the patient) in a given age group Infection source (Family, Playgroup, Daycare, School, Work, Work at a School, Neighborhood Cluster, Bar, Neighborhood, Business): cumulative number of agents who have been infected in the specified context within the simulation Disease status (Susceptible, Exposed, Pre-symptomatic Infectious, Symptomatic Infectious, Asymptomatic Infectious, Recovered): number of agents who currently have the specified disease status Mental state (Neutral, Afraid): number of agents who currently have the specified mental state Behavior (Default, Withdrawn due to hospitalization, Withdrawn due purely to fear, Withdrawn due to being afraid and symptomatic): number of agents who are currently behaving in the specified manner Broadcaster position (Neutral, Spreading fear, Countering fear spread): number of broadcasters which currently hold the specified position Analysis Scripts See README.md within epicast-coupled-main.zip for details on how to run the analysis scripts on this dataset. Acknowledgments This work was supported by the U.S. Department of Energy, Office of Science, Office of Advanced Scientific Computing Research and by cooperative agreement CDC-RFA-FT-23-0069 from the CDC's Center for Forecasting and Outbreak Analytics. Its contents are solely the responsibility of the authors and do not necessarily represent the official views of the Centers for Disease Control and Prevention. This work was performed at Los Alamos National Laboratory (LANL), an equal opportunity employer, which is operated by Triad National Security, LLC, for the National Nuclear Security Administration (NNSA) of the U.S. Department of Energy (DOE) under contract \#19FED1916814CKC. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. The findings and conclusions in this report are those of the authors and do not necessarily represent the official position of LANL. This research used resources provided by the Darwin testbed at LANL which is funded by the Computational Systems and Software Environments subprogram of LANL's Advanced Simulation and Computing program (NNSA/DOE). This material is based upon work supported by the U.S. Department of Energy, Office of Science, Office of Advanced Scientific Computing Research, Department of Energy Computational Science Graduate Fellowship under Award Number DE-SC0021. References 1. Tuccillo, J. et al. Urbanpop: A spatial microsimulation framework for exploring demographic influences on human dynamics. Appl. Geogr. 151, 102844 (2023).