Installing RRO 3.2.2 [Archived Documentation] · MRAN Skip to main content

Installing RRO 3.2.2 [Archived Documentation]

Archive Notice

These are not the latest instructions.

The following documentation describes how to install and uninstall Revolution R Open 3.2.2, which has been archived.

Take advantage of the latest version by downloading and installing Microsoft R Open 3.4.0.

Visit our Previous Releases page to learn about other past versions.


Due to the fact that these instructions have been archived, certain links on this page may no longer work over time.

Introduction

Revolution R Open (RRO), now called Microsoft R Open, is the enhanced distribution of R from Microsoft Corporation. Microsoft R Open is a complete open source platform for statistical analysis and data science, which is free to download and use.

The current version, RRO 3.2.2, is based on (and 100% compatible with) the statistical language, R-3.2.2, and includes additional capabilities for performance, reproducibility and platform support. Learn more...

System Requirements

Before you begin installing RRO 3.2.2, make sure the machine on which you will install meets the minimum hardware and software requirements defined below.

Supported Platforms. Revolution R Open is built and tested on the following 64-bit platforms:

  • Windows® 7.0 (SP 1), 8.1, 10, Windows Server® 2008 R2 (SP1) and 2012
  • Mac OS X Mavericks (10.9), Yosemite (10.10)
  • Ubuntu 12.04, 14.04, 15.04
  • CentOS / Red Hat Enterprise Linux 5.8, 6.5, 7.0
  • SUSE Linux Enterprise Server 10 & 11 & OpenSUSE 13.1

Processor. Processor 64-bit processor with x86-compatible architecture (such as AMD64, Intel 64, x86-64, IA-32e, EM64T, or x64 chips). ARM chips, Itanium-architecture chips (also known as IA-64), and non-Intel Macs are not supported. Multiple-core chips are recommended.

Free disk space. 250 MB

RAM. 1 GB required, 2 GB recommended.

Installing on Windows

Before installing RRO 3.2.2, please review the following system requirements.

Side-by-Side RRO Installs

On Windows, RRO 3.2.2 can work side-by-side with any other version of CRAN R, including past version of RRO. In order to continue working with a different version of R from a terminal window, you must specify the full path to the R executable you want to use.

After the installation of RRO 3.2.2, you can launch Rgui.exe to start interacting with R.

For example, if you installed RRO 8.0.3 and then upgraded to RRO 3.2.2, here is how you could work with them side-by-side:

  • Start RRO 3.2.2 by executing C:\Program Files\RRO\R-3.2.2\bin\x64\Rgui.exe if that is where you installed RRO 3.2.2.
  • Start RRO 8.0.3 by executing C:\Program Files\RRO\R-3.1.3\bin\x64\Rgui.exe if that is where you installed RRO 8.0.3.

How to Install RRO & MKL

Important! Any package dependencies will be installed automatically during the installation of RRO as long as your machine has access to the internet.

To install RRO 3.2.2 (and MKL) on Windows:

  1. Make sure R is not running. Stop every R process prior to installing RRO. You may need to close your IDE.

  2. Go to the directory where you downloaded the RRO installer.

  3. Launch the RRO installer and follow the prompts.

    Note: After installing RRO, the default R path is updated to point to R installed with RRO 3.2.2.

    Important!

    • If you choose not to write to the registry keys during the installation of RRO, then you may be missing some functionality.
  4. To configure RRO for reproducible research, you can disable the workspace save and reload feature. Learn more...

  5. To enable the optional multi-threaded performance for RRO, you must:

    Warning! RRO must be installed before you install MKL. If you uninstall RRO and reinstall RRO, your MKL would no longer work. In this case, you must first uninstall RevoMath from within the Add or Remove Programs or Uninstall a program dialog before proceeding with the reinstall of MKL.

    A. Download the Math Kernel Library (MKL) for this release.

    B. Open a Command Prompt window as Administrator.

    C. Go to the directory into which you downloaded RevoMath-3.2.2.msi, the MKL utility.

    D. Launch RevoMath-3.2.2.msi and follow the prompts.
    Important! You must have administrator privileges to install MKL properly on Windows.
    Note: This MKL installer will not be able to detect RRO if RRO was installed with the option to "Save version number in registry” unchecked.

    D. Learn how to set the number of threads here.

  6. When the RRO installation is complete, launch Rgui.exe to start interacting with R. On some platforms, you can find Rgui.exe under C:\Program Files\RRO\R-3.2.2\bin\x64\Rgui.exe.

What's Next After Install

1 Download and configure an IDE, such as RStudio.
2 If you're new to the R language, try this datacamp tutorial.
3 Check out some of these other resources.
 

Note: Revolution R Open also works with DeployR (installed separately). DeployR enables your application developers to deploy the R scripts, models, and data your R programmers have created in their applications.

Installing on Mac OS X

Before installing RRO 3.2.2, please review the following system requirements.

Side-by-Side RRO Installs

At this time, side-by-side installations of RRO 3.2.2 and other versions of R are not supported on Mac OS X.

How to Install RRO

Important! Any package dependencies will be installed automatically during the installation of Microsoft R Open as long as your machine has access to the internet.

RRO can be installed on Mac OS X using the standard Apple installer package.

To install RRO 3.2.2 on Mac OS X:

  1. Make sure R is not running. Stop every R process prior to installing RRO. You may need to close your IDE. Note: At this time, side-by-side installations of RRO 3.2.2 and other versions of R are not supported on Mac OS X.

  2. Go to the directory where you downloaded the RRO installer.

  3. Launch the RRO installer by double-clicking the filename.

  4. In the dialog box, click Continue.

  5. Follow the steps in the RRO installer.

  6. Click Install to begin the installation process.

  7. To configure RRO for reproducible research, you can disable the workspace save and reload feature. Learn more...

  8. When the RRO installation is complete, start interacting with R in the R console. Launch the R console by clicking on the Revolution R Open program icon in the Applications folder.

What's Next After Install

1 Download and configure an IDE, such as RStudio.
2 If you're new to the R language, try this datacamp tutorial.
3 Check out some of these other resources.
 

Note: Revolution R Open also works with DeployR (installed separately). DeployR enables your application developers to deploy the R scripts, models, and data your R programmers have created in their applications.

Installing on Linux

Before installing RRO 3.2.2, please review the following system requirements.

Side-by-Side RRO Installs

On Linux machines, RRO 3.2.2 can work side-by-side with any other version of CRAN R, including past version of RRO. During the installation of RRO, a symbolic link is created under /usr/bin/ that points to the last version of R that was installed, which in this case would be RRO 3.2.2.

After installation, you can start RRO 3.2.2 from a terminal window by typing R since it is the last one installed and is therefore under /usr/bin/.

You can also continue to work from a terminal window with other versions of R that were installed prior to RRO 3.2.2. To do so, specify the full path to the R executable you want to use.

For example, if you installed RRO 8.0.0 and then upgraded to RRO 3.2.2, you could work with both as follows:

  • Start RRO 3.2.2 by executing R since it was the last one installed.
  • Start RRO 8.0.0 by executing /usr/lib64/RRO/R-3.1.1/bin/R.
Tip: You can also manage multiple side-by-side installations of any application using the alternatives command. This command allows you create and manage symbolic links to the different installations, and thus easily refer to the installation of your choice. Example.

How to Install RRO & MKL

Important! Any package dependencies will be installed automatically during the installation of RRO provided your machine has access to the internet.

To install RRO 3.2.2 (and MKL) on Ubuntu:

Alert! RRO installations from a live CD are not supported on Ubuntu.
  1. Log in as root or a user with sudo privileges. If the latter, precede commands requiring root privileges with sudo.

  2. Make sure R is not running. Stop every R process prior to installing. You may need to close your IDE.

  3. Make sure the system repositories are up-to-date prior to installing RRO.  

    Note: This command will not install anything, but rather ensures that the repositories contain the latest software.

    sudo apt-get update
  4. Download the RRO installer for your platform.

  5. Change to the directory where you downloaded the RRO installer.

  6. Install RRO as follows:

    For Ubuntu 12.04:

    dpkg -i RRO-3.2.2-Ubuntu-12.4.x86_64.deb

    For Ubuntu 14.04:

    dpkg -i RRO-3.2.2-Ubuntu-14.4.x86_64.deb

    For Ubuntu 15.04:

    dpkg -i RRO-3.2.2-Ubuntu-15.4.x86_64.deb
  7. To configure RRO for reproducible research, you can disable the workspace save and reload feature. Learn more...

  8. To enable the optional multi-threaded performance for RRO, do the following:

    A. Log in as root or a user with sudo privileges. If the latter, precede commands requiring root privileges with sudo.

    B. Download the Math Kernel Library (MKL) for your platform.

    C. Go to the directory where you downloaded the MKL installer.

    D. Unzip the RevoMath-3.2.2.tar.gz file and run the MKL install script.

    tar -xzf RevoMath-3.2.2.tar.gz
    cd RevoMath
    ./RevoMath.sh

    E. Choose option 1 to install MKL and follow the onscreen prompts.

    F. Learn how to set the number of threads here.

  9. When the RRO installation is complete, run R in a terminal to begin interacting with R.

    Note: After installing, the default R path is updated to point to R installed with RRO 3.2.2.

To install RRO 3.2.2 (and MKL) on CentOS or Red Hat:

  1. Log in as root or a user with sudo privileges. If the latter, precede commands requiring root privileges with sudo.

  2. Make sure R is not running. Stop every R process prior to installing. You may need to close your IDE.

  3. Make sure the system repositories are up-to-date prior to installing RRO.  

    Note: This command will not install anything, but rather ensures that the repositories contain the latest software.

    sudo yum clean all
  4. Download the RRO installer for your platform.

  5. Change to the directory where you downloaded the RRO installer.

  6. Install RRO as follows:

    For CentOS / Red Hat 5.8:

    yum install RRO-3.2.2.el5.x86_64.rpm

    For CentOS / Red Hat 6.5:

    yum install RRO-3.2.2.el6.x86_64.rpm

    For CentOS / Red Hat 7.0:

    yum install  RRO-3.2.2.el7.x86_64.rpm
  7. To enable the optional multi-threaded performance for RRO, do the following:

    A. Log in as root or a user with sudo privileges. If the latter, precede commands requiring root privileges with sudo.

    B. Download the Math Kernel Library (MKL) for your platform.

    C. Go to the directory where you downloaded the MKL installer.

    D. Unzip the RevoMath-3.2.2.tar.gz file and run the MKL install script.

    tar -xzf RevoMath-3.2.2.tar.gz
    cd RevoMath
    ./RevoMath.sh

    E. Choose option 1 to install MKL and follow the onscreen prompts.

    F. Learn how to set the number of threads here.

  8. When the RRO installation is complete, run R in a terminal to begin interacting with R.

    Note: After installing, the default R path is updated to point to R installed with Microsoft R Open 3.2.2.

To install RRO 3.2.2 (and MKL) on SLES or openSUSE:

  1. Log in as root or a user with sudo privileges. If the latter, precede commands requiring root privileges with sudo.

  2. Make sure R is not running. Stop every R process prior to installing RRO. You may need to close your IDE.

  3. Make sure the system repositories are up-to-date prior to installing RRO.  

    Note: This command will not install anything, but rather ensures that the repositories contain the latest software.

    sudo zypper clean --all
  4. For openSUSE only: In order to be able to install packages with RRO, you must first install the package gcc-c++, if not already installed:

    # check for package
    sudo zypper search -i gcc-c++
    
    # if missing, install package  
    sudo zypper install gcc-c++
  5. Download the RRO installer for your platform.

  6. Change to the directory where you downloaded the RRO installer.

  7. Install RRO as follows:

    For SLES 10:

     zypper install RRO-3.2.2-SLES10.x86_64.rpm

    For SLES 11:

     zypper install RRO-3.2.2-SLES11.x86_64.rpm

    For openSUSE 13.1:

     zypper install RRO-3.2.2-openSUSE-13.1.x86_64.rpm
  8. To enable the optional multi-threaded performance for RRO, do the following:

    A. Log in as root or a user with sudo privileges. If the latter, precede commands requiring root privileges with sudo.

    B. Download the Math Kernel Library (MKL) for your platform.

    C. Go to the directory where you downloaded the MKL installer.

    D. Unzip the RevoMath-3.2.2.tar.gz file and run the MKL install script.

         For SLES 10, unzip as follows:

       tar -xzf RevoMath-3.2.2-SLES10.tar.gz
       cd RevoMath
       ./RevoMath.sh

         For SLES 11 or openSUSE 13.1, unzip as follows:

       tar -xzf RevoMath-3.2.2.tar.gz
       cd RevoMath
       ./RevoMath.sh

    E. Choose option 1 to install MKL and follow the onscreen prompts.

    F. Learn how to set the number of threads here.

  9. When the RRO installation is complete, run R in a terminal to begin interacting with R.

    Note: After installing, the default R path is updated to point to R installed with Microsoft R Open 3.2.2.

What's Next After Install

1 Download and configure an IDE, such as RStudio.
2 If you're new to the R language, try this datacamp tutorial.
3 Check out some of these other resources.
 

Note: Revolution R Open also works with DeployR (installed separately). DeployR enables your application developers to deploy the R scripts, models, and data your R programmers have created in their applications.

Uninstalling RRO

To uninstall RRO 3.2.2, select the tab for the relevant operating system and follow the instructions provided.

Warning: Always uninstall MKL before uninstalling RRO.

First, uninstall MKL Math Libraries:

This will revert to the original Rblas.dll and Rlapack.dll files.

  1. Remove the RevoMath package using the Windows instructions for uninstalling a program specific to your version of Windows. For example, on Windows 8, choose Control Panel > Programs & Features > Uninstall.

  2. Find RevoMath in the list and uninstall.

Then, uninstall RRO :

  1. Remove RRO using the Windows instructions for uninstalling a program specific to your version of Windows. For example, on Windows 8, choose Control Panel > Programs & Features > Uninstall.

  2. Find RRO for Windows 3.2.2 in the list and uninstall.

  3. Remove the directory in which you installed RRO, by default C:/Program Files/RRO.

To uninstall RRO on Mac OS X:

Run the following commands to uninstall:

   cd /Applications
   sudo rm -rf "Revolution R Open".app
   cd /Library/Frameworks
   sudo rm -rf R.framework
   sudo rm /usr/bin/R
   sudo rm /usr/bin/Rscript

First, uninstall MKL (used for multi-threaded performance):

  1. Go to the RevoMath directory and run the MKL utility:

    cd RevoMath
    ./RevoMath.sh
  2. Choose option 2 to uninstall MKL.

  3. Remove the library itself:

    rm –rf `R RHOME`/library/RevoUtilsMath.sh

Then, uninstall RRO on Ubuntu:

  1. Log in as root or a user with sudo privileges. If the latter, precede commands requiring root privileges with sudo.

  2. Run the following command to uninstall:

    dpkg --purge rro
    rm –rf /usr/lib64/RRO-3.2.2

First, uninstall MKL (used for multi-threaded performance):

  1. Go to the RevoMath directory and run the MKL utility:

    cd RevoMath
    ./RevoMath.sh
  2. Choose option 2 to uninstall MKL.

  3. Remove the library itself:

    rm –rf `R RHOME`/library/RevoUtilsMath.sh

Then, uninstall RRO on CentOS or Red Hat:

  1. Log in as root or a user with sudo privileges. If the latter, precede commands requiring root privileges with sudo.

  2. Run the following command to uninstall:

    yum erase RRO 
    rm –rf /usr/lib64/RRO-3.2.2

First, uninstall MKL (used for multi-threaded performance):

  1. Go to the RevoMath directory and run the MKL utility:

    cd RevoMath
    ./RevoMath.sh
  2. Choose option 2 to uninstall MKL.

  3. Remove the library itself:

    rm –rf `R RHOME`/library/RevoUtilsMath.sh

Then, uninstall RRO on openSUSE:

  1. Log in as root or a user with sudo privileges. If the latter, precede commands requiring root privileges with sudo.

  2. Run the following command to uninstall:

    zypper remove RRO
    rm –rf /usr/lib64/RRO-3.2.2

Warning: Depending on how R packages were installed, you may need to manually remove any R packages that remain in your personal R library and/or the global R library after uninstalling RRO.