Skip to content

Spatially explicit individual based model that simulates Coffee Leaf Rust epidemics on a coffee farm

License

Notifications You must be signed in to change notification settings

comses-education/spatialrust-model

Repository files navigation

SpatialRust: Coffee Leaf Rust Epidemic Model

fair-software.eu Docker Build Apptainer Build

SpatialRust is an individual based model that simulates Coffee Leaf Rust epidemics within a coffee farm.

Installing dependencies

Move to this project's directory and run:

$ julia install.jl

Running the model

There are two script files available. You can run a single simulation using a fixed parameter set using scripts/OneRun.jl as follows:

$ julia scripts/OneRun.jl

Results from this single run will be saved in a results folder as singlerun.csv.

The second option, scripts/ParameterRuns.jl, lets you run a parameter exploration experiment. The default setup of this experiment will run 880 simulations. To modify the parameter values to be evaluated or the replicates for each combination, open scripts/ParameterRuns.jl and edit lines 7 to 11. Like the first option, you can run the script from bash:

$ julia scripts/ParameterRuns.jl

Results from this experiment will be saved in a results folder as parameterexp.csv. Both scripts take care of creating the results folder if it has not been created yet.

Running on Open Science Grid

  1. Establish an account on Open Science Grid https://osg-htc.org/research-facilitation/accounts-and-projects/general/index.html
  2. Create a host alias for your OSG account (https://github.com/comses-education/fair-osg-template#set-up-your-user-account-on-the-open-science-grid)
  3. Build a singularity image and deploy it to your OSG /public/<username> directory via $ make OSG_USERNAME=<your-osg-username> deploy
  4. ssh into the OSG login node, cd into the spatialrust directory and submit the generated spatialrust.submit via $ condor_submit spatialrust.submit
  5. this runs the ParameterRuns.jl on OSG and should drop off a results.zip file with the data in the same directory you submitted the job script.