The docker container with the dependencies for the R code we're using:
docker pull shahsam/covidscenariopipeline:latest
From that container, the code to run the simulations (single threaded) is:
Rscript R/hosp_run.R model_output/unifiedNPI/ high 1
The output of the run will be written to hospitalization/model_output/unifiedNPI/
and the expected outputs are in hospitalization/model_output/goldenNPI.
There are a few tweaks to the hosp_run.R code-- we use %do% instead of %dopar%
for the loops in order to make debugging easier, and I used a set.seed call
at the top of the file to make the randomization parts of the run reproducible.