Benchmarks Online

Skip Navigation Links


Page One

Campus Computing News

Winter Break Hours

UNT Web Presence Harvested

Today's Cartoon

RSS Matters

The Network Connection

Link of the Month

WWW@UNT.EDU

Short Courses

IRC News

Staff Activities

Subscribe to Benchmarks Online
    

 Research and Statistical Support - University of North Texas

RSS Matters

Link to the last RSS article here: Part I: From The Source - R 2.2.0 on OS-X Tiger 10.4.3 - Ed.

Part II: From The Source - R 2.2.0 on OS-X Tiger 10.4.3

By Patrick McLeod, ACS Research Consultant

Introduction

In Part I of this Benchmarks Online article, we walked through installing R on OS-X Tiger 10.4.3 from source. While I would hope anyone reading this column would be at least equally comfortable with working in both a command line and GUI environment, most computer users do prefer a GUI interface. Thankfully, R has a new Cocoa-built GUI interface on OS-X. In Part II of “From the Source” we will build R.app using XCode tools and get this GUI interface up and running.

Before we begin, there is a bit of clarification that will help keep what we’re working on in this article delineated from other R GUIs on OS-X. R.app is the most current GUI for R on OS-X, but it is not the only GUI project. There is also an Aqua interface to R and a Cocoa GUI bundle. R.app is written in Cocoa, but is not a bundle; it’s a whole new Cocoa GUI for R on OS-X and is current for all versions of R on OS-X after R 2.0.

As before, be sure to reference the official R for Max OS-X FAQ:

http://cran.r-project.org/bin/macosx/RMacOSX-FAQ.html .

Building R.app

The hard part of building R for OS-X is, as previous described, making sure all of the dependencies you need are present and properly installed. We are reminded of this once more in building R.app, which relies on a specific set of configuration flags in the compilation of R 2.2.0 that I did not mention in the previous article and the most current version of XCode installed (a version as or more recent than XCode 1.5 should work).

First, when you issue your ./configure command to configure and make R 2.2.0 from source, you need to set the following flags to insure the proper environment for building R.app as the next step:

./configure –with-blas=’-framework vecLib’ –with-lapack –with-aqua
make
sudo make install

BLAS is a facility for matrix computations, LAPACK is a facility for numerical linear algebra, the vecLib framework is a collection of processing and operations facilities, including BLAS and LAPACK and aqua is an older flag, a still-necessary holdover from the aqua GUI days.

After R has been installed with the above flags, the R.app source can be downloaded in tar ball from the CRAN website (follow the hyperlink to R for Mac 10.2.x and above). Once this file has been downloaded, you’ll want to uncompress it (tar –xvf) and then issue the following command from within the uncompressed Mac-GUI directory:

xcodebuild –target R –buildstyle Deployment

R.app in Pictures

Now that we’ve successfully compiled R.app, let’s look at our creation. In the /Applications file path, find the R icon and double-click on it. It should launch something that looks like the following:

R screen

So what’s so special about this? It doesn’t look that different! Well, once we take a timed screen grab with action (using the package installer from the drop down menu to download and install R packages), you can begin to see the usability advantages offered by R.app:

R.app screen

Once correctly compiled, R.app displays an R Console (I changed the background color preferences to light green from the default white) and nine GUI drop-down menus for various aspects of working with R (R, File, Edit, Format, Workspace, Packages & Data, Misc, Window and Help).

The Disk Image Buzzkill: R-2.2.0

If you’ve read this article and the preceding article through and would like to bring R-2.2.0 with the Cocoa GUI up but you’re not comfortable with doing all of this work from the command line, there is an alternative: Installing R-2.2.0 from a disk image available from CRAN and can be done via point and click.

While folks who are new to using command line tools to configure, compile and/or install would find it easier to download the disk image file and point and click their way through the installation, you will not get the same kind of control over detail from the GUI installation that you will from the command line. Regardless, it’s worth noting that a command line alternative does exist.

On a closing note, Rich, Mike and I would like to wish all of our readers "Happy Holidays" and we look forward to you visiting us in 2006!

 


Please note that information published in Benchmarks Online is likely to degrade over time, especially links to various Websites. To make sure you have the most current information on a specific topic, it may be best to search the UNT Website - http://www.unt.edu . You can also search Benchmarks Online - http://www.unt.edu/benchmarks/archives/back.htm as well as consult the UNT Helpdesk - http://www.unt.edu/helpdesk/ Questions and comments should be directed to
benchmarks@unt.edu

 

Return to top