SIMAGRI Agricultural Simulator

Built using Docker, Python 3.8, Dash Plotly, and DSSAT (https://dssat.net/), How to compile DSSAT on Linux (https://dssat.net/210/).

Three versions of SIMAGRI have been developed for the following countries:

• Ethiopia
• Senegal
• Colombia

For each country SIMAGRI enables crop simulation either based on historical weather data or as a forecast.

SIMAGRI Ethiopia: http://simagri-ethiopia1.iri.columbia.edu/historical

Instructions to run SIMAGRI locally:

1. Install Docker. The installer can be found here: [WIN] [OSX] [LINUX].

2. Clone this repo:

git clone git@github.com:Agro-Climate/ET_DSS_hist.git

cd ET_DSS_hist

git checkout ET_DSS_hist_Linux

Due to the similarities of the different localizations of SIMAGRI the app is structured so files unique to each country are stored in apps/<country>

Steps 3-5 will allow building Docker images for each country and running them as Docker containers

3. Build a Docker image for desired country:

docker build -f ./apps/<country>/Dockerfile -t simagri_<country>_img:latest .

4. Run a Docker container for desired country:

docker run --name=simagri_<country> -e PYTHONUNBUFFERED=1 --rm -dp <port>:5000 simagri_<country>_img:latest

After running this command the app may be viewed at localhost:<port> (when deploying to production on a server port 80 should be used)

The following command may also be used to allow tracking of the command line output of the docker container:

docker logs --follow simagri_<country>

5. Kill a running container and clear unused resources (optional):

docker kill simagri_<country>

docker image prune -af