Docker¶
We use Docker to build our respective integrated development environments (IDE) for working with dpl software.
RStudio containers are based on the Rocker R and RStudio builds. CyVerse rehosts the Rocker images in its data science workbench, the Discovery Environment.
CyVerse container builds are maintained on a GitHub Organization, and hosted on its private Harbor Registry.
JupyterLab containers are based on the Project Jupyter image stacks.
Testing¶
Run Docker for testing the code with Jupyter Lab or RStudio-Server.
Install the Docker Desktop for Windows or Mac OS X, or command line for Linux.
Pull pre-existing Docker images for RStudio-Server or JupyterLab:
docker pull jupyter/datascience-notebook:latest
To run RStudio-Server (authentication user: rstudio, password: set it yourself below):
$ git clone https://github.com/opendendro/dplPy
$ cd dplPy
$ docker run -it --rm -p 8787:8787 -e PASSWORD=new_password -v $PWD:/home/rstudio/dplPy -e REDIRECT_URL=http://localhost:8787 rocker/geospatial:latest
Open your browser and navigate to http://localhost:8787
If you're running remotely, use the DNS of the virtual machine service with the :8787 or allow it to open the tab for you, e.g., with CodeSpaces or GitPod.
From the R Console:
> install.packages("dplR", dependencies=TRUE)
The Project Jupyter DataScience Notebook runs both Python and R:
$ git clone https://github.com/opendendro/dplPy
$ cd dplPy
$ docker run -it --rm -p 8888:8888 -v $PWD:/home/jovyan/dplPy -e REDIRECT_URL=http://localhost:8888 jupyter/datascience-notebook:latest jupyter lab --no-browser --NotebookApp.token=''
Open your browser and navigate to http://localhost:8888
- Note: that we're disabling the Notebook Token so you don't have to authenticate; remove
--NotebookApp.token=''to re-enable.