Build the Docker image
NOTE: This guide is for building the Docker image from source. If you just want to run gds_env, see the install guide instead.
Requirements
- Docker installed and running
- A clone of the
gds_envrepository make
Build process
From the root of the repository, run:
make build
This builds the image and tags it as gds:latest (and gds:<date>_<arch>). The build log is saved to env/build_<arch>.log.
The build installs the full conda environment defined in env/gds_<arch>.yml, plus a set of additional tools (Quarto, Jekyll, tippecanoe, decktape, LaTeX tools, Vim, GPQ). It can take a while — especially on first run.
Once complete, verify the image was built:
docker image ls | grep gds
Testing the build
To run the test notebooks against a built image:
make test image=gds:<your-tag>
This runs both the Python and R check notebooks inside the container and saves HTML output locally.
Building gds_code
To build the gds_code variant (VS Code in the browser) on top of an existing gds image:
make build_code image=gds:<your-tag>
See the gds_code stack page for more details.