This project uses data on influenza virus loads in wastewater that are generated by EAWAG and the State Laboratory of the canton of Basel-Stadt.
It also uses data on laboratory-confirmed influenza cases from the obligatory notification system "Sentinella" in Switzerland. These data are provided by the Federal Office of Public Health.
The relevant data files are expected to be in data/raw_data.
The pre-processing scripts in R/helper_scripts generate generate clean versions of these raw data files and save them to data.
The Re estimation scripts estimate_wastewater_re.R and estimate_confirmed_case_re.R run automatically via Github actions when new clean versions of the data files are pushed.
These scripts save Re estimates to app/data.
Estimates will only be generated for the current influenza season, this is hardcoded in the Re estiamte scripts and in the data preprocessing scripts here.
Cached estimates from previous seasons are stored here.
The shiny app can then be run to visualize the Re estimates. You can run the shiny app locally from RStudio by opening app/server.R, setting your working directory to the app directory, and clicking "Run App".
The shiny app can also be run locally using the docker image specified by Dockerfile. Once you have docker installed and the docker daemon running:
docker build -t ww-shiny .
docker run --rm -p 8080:3838 ww-shiny
# Open app in browser: http://localhost:8080
The app is actually deployed using this docker image. Another Github action workflow re-builds the image using the latest data in the repository every Monday, Wednesday, and Friday morning. This workflow can also be manually triggered from Github. The workflow pushes the re-built image to the Github package registry associated with this repository.
The server that hosts the online dashboard continuously monitors the pacakge repository using watchtower and will update the online version of the dashboard whenever a new image is available.
To pull the production image and run it locally, follow the steps here to log in to the Github package registry. Then run:
docker pull ghcr.io/wise-ch/wastewater-influenza-dashboard:master
docker run --rm -p 8080:3838 ghcr.io/wise-ch/wastewater-influenza-dashboard:master
# Open app in browser: https://localhost:8080
The dashboard is currently hosted on an server at ETH Zurich.
The app is currently running at https://wise.ethz.ch/influenza/.
The dashboard was developed by Sarah Nadeau. It is based on work by Taru Singhal and Jana Huisman for SARS-CoV-2. Chaoran Chen helped set up and maintains the infrastructure. Adrian Lison is maintaining and further developing the dashboard.
Build the image, then run a container locally in interactive mode:
docker run --rm -it -p 8080:3838 ww-shiny bash
# You should be inside the container now
/usr/bin/shiny-server # Try lauching the app
# CTRL+C to stop the running app
cat /var/log/shiny-server/shiny-server-shiny-*.log # Inspect the log to see what went wrong
exit # Exit the container