2.1 Installing R

In this session#

• Understand the different mechanisms for installing R

• Installing R from source and pre-compiled binaries

Three mechanisms for installing R#

You have three different mechanisms for installing R on a Linux machine:

• Install from a Linux package manager

  • For example yum (Red Hat) or apt (Ubuntu)

• Compile from source

  • Download the R source code and compile using the relevant GCC compiler

• Install from pre-compiled binaries provided by RStudio

  • Download and install the pre-compiled binaries available from RStudio

Method 1: Installing R using yum or apt#

Installing from yum or apt (or similar) is traditionally the mechanism that most administrators would use.

For data science work in production, it is important to provide code stability by installing multiple versions of R side-by-side.

This means that using your Linux package manager for installing R in production is not a good idea.

Method 2: Compile R from source#

For data science work in production, it is important to provide code stability by installing multiple versions of R side-by-side.

You can achieve this by installing R from source.

• Download the source code
• Create a "make" configuration
• Run make
• Run make install

This offers maximum flexibility, because you can specify the location of the installed package and you can optimize for your hardware.

Instructions for installing R from source

The downside of installing R from source is that it takes some time for the compilation to complete.

If you want to learn more about make you can read about it here: https://swcarpentry.github.io/make-novice/

Method 3: Install from pre-compiled binaries#

RStudio recommends that in most cases you install R from pre-compiled binaries.

This gives you the benefit of:

• Side-by-side installation of multiple versions of R
• It is quick, since the compilation step has previously been completed

To install from pre-compiled binaries, follow the instructions at Install R

Summary#

Recommendation#

Our suggested order for making a choice of R installation is:

1. Install from pre-compiled binaries

   • to get multiple R versions side-by-side

2. Compile from source

   • gives more flexibility but takes longer

3. Use yum or apt

   • Does not allow side-by-side install
   • Do not use for long-term data science stability

Your turn#

Next complete the exercise.

• R is running