tutorial_usage_guide.md

Tutorial: Using EMOD

Running a Simulation

1. Log on to BigPurple - either through SSH or VS Code's SSH function

2. Navigate to the tutorial directory in the command line

3. Run the simulation:

sbatch script.sbatch

The command sbatch is a Slurm command that runs script.sbatch. Within that script is an un-commented line which tells python to run your EMOD simulation.

How do I know it is working?

Within your tutorial directory, the sbatch command will generate a file with a name like slurm_[numbers].out. This is a text file that catches all of the output from the simulation.

Wait about a minute, then open that file. If an error has occurred, you will see the error messages there.

If you see a very long message with the line Waiting 60 seconds for experiments to complete... at the bottom, that means the simulation is running.

What is EMOD doing when we run scenarios?

• Initializes an ensemble of simulation runs. Each simulation run takes its parameters from one of the lines from resampled_parameter_sets_short.py. That file contains 100 lines, each of which represents one set of parameters for EMOD. Each simulation run spins up a job on BigPurple, so 100 simulation runs will require 100 separate threads on BigPurple, which uses Slurm to manage all of the jobs.
• For each simulation run
  • EMOD builds a synthetic population
  • Seeds the HIV epidemic
  • Allows the epidemic to proceed forward in time
  • Writes to output certain reports documents which are formatted according to the specifications given in the config.json file

Postprocessing

After running the simulation, the next step will postprocessing