EMOD input files

Every EMOD simulation is driven by a set of input files. Understanding what each file controls will help you follow the tutorial scripts.

Configuration file

config.json contains the simulation-wide settings: how long to run, which disease model to use, transmission parameters, and hundreds of other options. In the tutorials, build_config() is the callback that builds this file. The built-in country models provide a pre-validated configuration for the modeled country that you can then modify.

Demographics file

The demographics file describes the human population: initial population size, age and sex structure, mortality, fertility, and relationship formation parameters. In the tutorials, build_demographics() builds this file. The built-in country models provide demographics calibrated to a specific country's epidemic.

Campaign file

The campaign file defines interventions — what to deploy, to whom, and when. Each intervention event specifies a trigger, a target population, and an intervention class such as ART initiation, voluntary medical male circumcision (VMMC), or condom distribution. In the tutorials, build_campaign() builds this file. The built-in country models include a default cascade of care that can be customized or replaced.

Reports and output files

Reports are configured in the script before the experiment runs. They tell EMOD what to measure and at what frequency — for example, annual HIV prevalence by age and sex, or new infections per timestep. EMOD writes each report's results to an output file in the output/ subdirectory of each simulation. InsetChart.json is a time-series summary of key channels that must be explicitly enabled before it is produced. Additional reports can be added to the task via add_reports(). Tutorial 2 introduces reports and shows how to download and plot the output files.