Docker provides a robust, lightweight way to ship applications within a small, self-sufficient operating system, which includes all the dependencies and tools required for the app the run smoothly. This structure is called a container, and different versions of a container are referred to as images.
The major advantage of using this system to deploy a set of applications is that applications are guaranteed to work the same for every user, regardless of their operating system or system settings. For RECON packages, this means that you do not need to worry about using the right R version, or installing packages and their dependencies, because the RECON docker image will contain all this already.
Several RECON docker images will be developed. The default one will always be
reconhub/recon
.
First of all, you need docker installed on your computer. Follow this link for guidelines to install Docker Community Edition.
Then you need to choose the RECON docker image you want to use (see next sections on available images), and choose how you want to use the image. There are two recommended ways to use RECON docker images: starting a Rstudio server, or starting a bash session.
This approach is aimed at regular users, is the use the image to start a Rstudio server, which enable the user to start a Rstudio session within a web browser, with all currently functional RECON packages and their dependencies pre-installed. In short: you will be able to use the RECON packages without even installing R or Rstudio. See an example of such session using the package epicontacts:
Importantly, you can also start the RECON GUI from this session. For instance,
to start the incidence GUI, you only need typing incidence.ui::incidence_ui()
:
This image is started by typing:
docker run -d -p 8787:8787 reconhub/recon
where reconhub/recon
can be replaced by your preferred RECON docker image. A
Rstudio session can then be started by opening a web browser and going to the
URL 127.0.0.1:8787
(or localhost:8787
). The login and password are
'rstudio'.
This approach is recommended for developers. It will open a bash
shell logged
as guest
within the /home/guest
. A folder /home/guest/dev
will contain git
clones of the various installed packages. A R session can be started (without
support for graphics) simply by typing R
.
To start the docker image with a bash
shell, type:
docker run --rm -it --user guest reconhub/recon /bin/bash
where [image]
is the docker image you want to use.
A list of available images can be found from the RECON docker repository.
This image contains all functional RECON packages. CRAN versions are installed when available, otherwise github versions are used. It also contains package sources for testing. It is based on rocker/verse
-
CRAN packages: outbreaks, incidence, epicontacts
-
github packages: epicontacts, distcrete, outbreaker2, vimes, shinyHelpers, recon.ui, incidence.ui, epicontacts.ui
-
users: 'guest' (for bash session) and 'rstudio' (for Rstudio session)
-
package sources: git-cloned in
/home/guest/dev/
This image designed for testing is based on daily snapshots of R-devel from rocker/drd. It also contains all the sources of the packages cloned from github.
Note: unless you need to test a package on R-devel, prefer using the more reliable default image (reconhub/recon).
-
github packages: outbreaks, incidence, epicontacts, distcrete, outbreaker2, vimes, shinyHelpers, recon.ui, incidence.ui, epicontacts.ui
-
users: 'guest' (for bash session)
-
package sources: git-cloned in
/home/guest/dev/
Contributors (by alphabetic order):
See details of contributions on:
https://github.com/reconhub/docker/graphs/contributors
Contributions are welcome via pull requests.
Please note that this project is released with a Contributor Code of Conduct. By participating in this project you agree to abide by its terms.
Maintainer: Thibaut Jombart ([email protected])