6.1 Installing Package Manager
In this session, you will:
- Understand the problems associated with R package management in enterprise environments
- Learn how RStudio Package Manager (RSPM) solves these problems
- Install RSPM on a server
- configure RStudio Server Pro and RStudio Connect as consumers of the packages
What role does RStudio Package Manager play?#
The problems of R package management#
Any R administrator is concerned about distributing R packages to their users.
CRAN has more than 15,000 packages and is updated more than 30 times per day and in addition your users may be creating packages of their own, or using development versions of packages from Github.
Some typical things you might consider include;
How can I:
- ensure reproducibility of packages for all of my users?
- centralize package management for R users across teams?
- serve R packages in an airgapped environment?
- allow R users not to worry about package versions and availability?
- see which packages and licenses are in use in the organization?
- speed up package installation across a large number of users?
Package Manager can help solve these problems#
RStudio Package Manager allows you to organize and centralize R package management across your team, department, or entire organization.
RStudio Package Manager helps:
- Administrators maintain R servers without breaking code
- IT centralize packages that meet standards
- Data scientists reproduce results without worrying about packages
- Developers share their code as private packages
- Introduce CI/CD for internally developed packages
- Save Linux and Windows users from compiling packages from source
RStudio Package Manager can serve packages from various sources:
- Curated CRAN
- Internal Packages
- Packages from Git
You can read more about the benefits of RStudio Package Manager on our website.
RSPM documentation and resources#
The source of truth about RStudio Package Manager is the admin guide:
The course also touches upon the remaining sections so you know where to go to get the information you need after the course.
View a live instance of Package Manager#
Let's look at the basic functionality of RStudio Package Manager:
RStudio maintains a public instance of Package Manager which is free to use. The only difference between this and the version that you install youself if that you're not able to use any of the customizations mentioned above. We can use this to view in the interface and see how it looks to an end user. Visit https://packagemanager.rstudio.com and explore the following:
- View different repositories
- View package usage statistics
- Get more details on a specific package in a repository
- Review steps to configure and install packages from a repository
RStudio Package Manager:
- Organizes packages in a central repository
- Performs automatic versioning of the entire repository
- Tracks package usage statistics and metrics
- Integrates with with R, RStudio Server Pro, and RStudio Connect
Data Science with the RStudio toolchain#
RStudio Package Manager supports all other uses of R within your organization, from development to publishing and deployment.
RSPM can be used with direct access to CRAN or in air-gapped / offline installations.
- RHEL / CentOS 6+ and 7+
- SUSE 12 SP3+
- Ubuntu 14.04, 16.04, 18.04
- Minimum of 2GB of RAM
- Recommended 50 GB to 200 GB of disk storage
- Package sources can be lazily cached from CRAN or eagerly downloaded
View more information on Configuration and Sizing Requirements.
Support for various network setups:
- Direct online access to https://rspm-sync.rstudio.com
- Proxied network access to https://rspm-sync.rstudio.com
- Airgapped / offline systems
- By default, root privileges are required to install, start, stop RSPM and activate license, but there is a
Minimal Root Useinstallation path
- By default, RSPM creates and runs under a service account,
Other useful information:
- R version 3.0+ is required if you want to build packages from Git
- RSPM currently serves R source packages (All platforms) and binary packages (Linux and Windows only)
In the following exercise, you'll install RSPM to your server.
Signs of success:
- RStudio Package Manager is up and running!