Transcript Slide 1

Seven good reasons why
everyone should be using R
Why use R?
There are a huge number of statistics packages available, including
SPSS, Stata, Minitab, SAS, GenStat, etc., so why use something
scary like R (see below)? Here are 7 excellent reasons...
Why use R?
Reason 1 of 7:
It’s FREE, and always will be!
• For comparison, a single user licence for SPSS costs around £3,000.
• There’s no guarantee that a particular commercial piece of stats software will be
available at your next job – if you’re going to put in the effort to learn a stats
programme, you might as well choose one that you will always have access to!
Why use R?
Reason 2 of 7:
R is used by the majority of academic
statisticians.
• Statisticians develop new statistics in R, therefore:
• R code will generally be available for published statistical techniques;
• It contains advanced statistical routines not yet available in other packages;
• Collaborations with statisticians will be most fruitful if you share a common
language.
• R packages are constantly updated to fix bugs and include new routines – tests in
R therefore contain less errors and are more feature-rich than in other programs.
Why use R?
Reason 3 of 7:
R is platform independent.
• If you use Windows, this may not be such a big deal but it is a tremendous
advantage if you collaborate with Mac or Linux users.
Why use R?
Reason 4 of 7:
R has unrivalled help resources.
• There are a large number of superb books and online resources dedicated R,
aimed at both new and advanced users (more of this at the end) – far, far more
than for any other statistics package.
• Because R is community constructed, free software, advanced users and the
developers themselves are more willing to provide help.
• The quality and quantity of help for R is particularly relevant when trying to teach
yourself a new technique or statistical method.
Why use R?
Reason 5 of 7:
R has state-of-the-art graphics capabilities.
• Put simply, R produces beautiful, publication-quality figures.
• R gives you a high level of control over all aspects of a figure’s appearance.
• You can produce figure types typically only available in a small number of
specialist commercial packages (e.g. Matlab, Mathematica).
Why use R?
Reason 6 of 7:
The command line interface!
• The command line interface – perhaps counter intuitively – is much, much better
for learning about stats.
• It is easy to share code with colleagues or download it from discussion
forums, statistics websites, etc.
• You can tinker with the code to gain a really good understanding of what it
actually does and whether it will work for you.
• It gives you complete control over all aspects of the stats you are running – you
are not constrained by the options available in a graphical user interface.
• If you must, graphical user interfaces are available (e.g. http://sciviews.org/_rgui)
Why use R?
Reason 7 of 7:
R is far more than a statistics application.
• It is a full programming language, and so provides an unparalleled platform for
programming new statistical methods , modifying existing ones and working with
data (e.g. to automatically process large datasets, access data stored in databases)
in an easy and straightforward manner.
• People have written packages for just about everything, from optimization to
bioinformatics.
• It can link directly to other programming languages, such as Matlab , C, C++ and
Java.
To find out more:
• Staff at Lincoln, e.g. Tom Pike, Biological Sciences ([email protected])
• Dedicated discussion forums, e.g. www.r-project.org
• Dedicated websites, e.g. www.statmethods.net
• Many great books, including: