EpiModel Extensions for High-Performance Computing

Extension package to EpiModel to run large-scale stochastic network models on modern high-performance computing systems. Provides scenario-based batch simulation, helpers for building 'slurmworkflow' job workflows and 'swfcalib' calibrations, cluster configuration presets, and shell tooling for detecting and recovering degenerate SLURM tasks.

Details

EpiModel provides tools for the mathematical modeling of infectious diseases. Supported model classes include stochastic network models, which rely on the statistical framework of exponential-family random graph models (ERGMs) that evolve over time. This allows for modeling of disease-related contacts with duration, such as ongoing sexual partnerships.

The level of statistical complexity of these models, based in Markov-chain Monte Carlo (MCMC) simulation, results in computationally intensive simulation processes. The goal of EpiModelHPC is to provide a standardized framework for extending EpiModel to run on modern high-performance computing (HPC) systems.

References

The main website for EpiModel is at https://www.epimodel.org/. The source code for this extension package is hosted on GitHub at https://github.com/EpiModel/EpiModelHPC. Bug reports and feature requests may be filed there.

See also

Useful links:

• https://www.epimodel.org/

• https://epimodel.github.io/EpiModelHPC/

• Report bugs at https://github.com/EpiModel/EpiModelHPC/issues

Author

Maintainer: Samuel Jenness samuel.m.jenness@emory.edu

Authors:

• Samuel Jenness samuel.m.jenness@emory.edu