Bioconductor provides an excellent set of docker images containing R, RStudioServer, and the sets of Bioconductor packages appropriate for certain use cases.
Here we have a version of the Bioconductor bioconductor/release_core which adds on R packages containing training material such as ISB CGC Examples, gcloud, and some configuration to .Rprofile.
The following instructions assume that Docker is installed and the current user can run docker commands. For help with that one-time initial setup, see https://cloud.google.com/container-registry/docs/#install_docker.
(1) Get the latest Dockerfile, etc., from this repository via a git clone https://github.com/isb-cgc/examples-R.git or git pull.
(2) Change into the directory where the Dockerfile reside.
cd examples-R/inst/docker
(3) Make sure your build machine has the latest Bioconductor Docker image.
docker pull bioconductor/release_core
(4) Build the image using a tag indicating today's date. Always specify a tag.
docker build -t b.gcr.io/isb-cgc-public-docker-images/r-examples:2015-10-30 .
(5) Push the new version to the public image repository.
gcloud docker push b.gcr.io/isb-cgc-public-docker-images/r-examples:2015-10-30
(6) Also tag the new version as 'latest'. Always explicity mark as 'latest' a particular tagged version.
docker tag -f b.gcr.io/isb-cgc-public-docker-images/r-examples:2015-10-30 \
b.gcr.io/isb-cgc-public-docker-images/r-examples:latest
(7) And push 'latest'. (This will be really quick since its just updating metadata about 'latest'.)
gcloud docker push b.gcr.io/isb-cgc-public-docker-images/r-examples:latest