Template for R docker image for climate/health predictions

This is a template for creating a Docker image with custom R packages.

This template is based on a base image that has the INLA packaged already installed, which saves a lot of time compared to installing it from source.

How to use

1. Make a fork of this repository, and give it a suitable name (e.g. docker_for_some_model). Clone the repository locally. Make sure to enable Github actions in the cloned repo (Go to Settings->Actions->General in the new repo). You might also need to enable then in the actions tab (from the bar at the top of the screen) by confirming that you know the workflows and actions.
2. Decide on a name for the image you want to build. Edit the file .githbub/workflows/build.yml and make sure IMAGE_NAME points to your name. Note that upper case letters will be converted to lower case.
3. If you want, change how the Github action is being run. Currently, it runs on every push to the master branch (NB: Change this to main if your branch is called main).
4. Edit Dockerfile. The file contains an example of how to add an R package
5. Push the repository to Github. You should create a repository on Github first, add the correct remote and then push.

If everything works, a Github action will start running after you push. If nothing fails, an Image will be pushed to the Github package registry. The name will be ghcr.io/dhis2-chap/docker_r_inla.

In order to make it possible for anyone to use the image without authentication, you have to make the image public. On the right side in the front page of the repository you created, you will see the package. Click Package settings and edit visibility.

To verify that the image exists and is available to the public, you can run docker pull ghcr.io/ORGANIZATION/IMAGE_NAME:master.

Tip: It can be a good idea to add new lines for every package you want to install in the Dockerfile, to make use of Docker's caching.

Tip: The current Dockerfile is based on an image that has R with INLA. If you don't need INLA, feel free to base the image on just R instead, e.g. by using FROM.