1 Resources
The home page for R is at http://www.r-project.org/. This contains links to obtain the software itself, gives access to the wealth of online documentation, and gives details of further resources, e.g. books.
There are some recommended online tutorials at
http://www.macs.hw.ac.uk/actuarial/R/
including how to download and install R on your computer.
Starting and quitting R
Notes on R
Start R by double-clicking on the R shortcut. This will start the R GUI (graphical user interface). Type commands in the R console (the command window within the GUI). Exit via the File menu, or the q() command. R prompts you to save an image of your current workspace (currently defined objects, etc). This is automatically reloaded when R is next started.
3 Running R
Within the command console, R issues its characteristic prompt: >
The user types expressions which are then evaluated. The results are either displayed immediately, or may be assigned (using either of the assignment operators = or <- ) to R objects.
In the examples below input typed by the user is preceded by the above prompt—all else is the computer’s response:
[1] 1 2 3 4 5 6 7 8 9 10
> x = 1:10 >x
[1] 1 2 3 4 5 6 7 8 9 10 > y = log(x)
[1] 0.00000 0.69315 1.09861 1.38629 1.60944 1.79176 1.94591 2.07944 2.19722
[10] 2.30259
[1] 0.0000 1.3863 3.2958 5.5452 8.0472 10.7506 13.6214 16.6355 19.7750 [10] 23.0259
[1] 0.0000 1.3863 3.2958 5.5452 8.0472 10.7506 13.6214 16.6355 19.7750
[10] 23.0259
[1] 10.208
> objects()
> objects()
> objects()
(a built-in R function with, on this occasion, no supplied arguments) produces a list of all user-defined objects (variables, datasets, functions) in the current workspace.
Objects may be selectively removed with the function rm. All the user-defined objects in the current workspace may be removed at once via the Misc menu.
There are other objects stored elsewhere on the system. Some are available for immediate use; these are the in-built functions (e.g. log, mean, var, plot) and constants (e.g. pi). Others are available via add-on R packages—see Section 8.
Previous command lines may be recalled for editing via the up-arrow and down-arrow keys. Long calculations may be interrupted via (on Windows) the
4 Managing data in R
Small datasets may be typed directly into the R workspace, e.g.
> marks = c(53, 67, 81, 25, 72, 40)
[1] 53 67 81 25 72 40
> mean(marks)
[1] 56.3333
Alternatively the function scan may be used:
> marks = scan()
1: 53 67 81 25 72 40
Read 6 items
[1] 53 67 81 25 72 40
Data already stored elsewhere is either available for immediate use, or may usually be loaded into the workspace via the function data. For example
displays Anderson’s famous iris data. Type
to see a both a list of immediately available datasets and also instructions on how to find all datasets already on the system.
The functions data and load will also load data from external files which use R’s own storage format. The function save can be used to save objects in this format. The function save.image can be used to save the entire current workspace (see below).
The function read.table can be used to read a rectangular array of data stored in an external ASCII (text) file (for files with tricky formatting it may be necessary to use the function scan). Similarly the function write.table can be used to write a rectangular array of data to an external ASCII (text) file
For help on any of these functions see Section 10.
5 Managing your work in R
The R workspace is the current working environment, primarily the user-defined objects (variables, datasets, functions).
On exiting R, you are prompted to save an image of your current workspace. This is automatically reloaded when R is next started.
In a lengthy R session you may wish to save the workspace periodically to protect against system crashes. Use the File menu, or the function save.image.
It is sensible (i.e. essential) to keep a separate R workspace for each of your projects. The easiest way to do this is to install a separate desktop shortcut to R for each project. Give this a name which corresponds to the project.
On Windows systems copy an existing shortcut, change its name, and use the shortcut Properties, Start in entry to correspondingly change the name of the R working directory.
There are other ways to keep workspaces separate, but the above is the simplest.
6 R object types
Here is a list of a few of the more important object types.
Vector: a one-dimensional array of arbitrary length. Subsets of the vector may be refer- enced. All elements of the vector must be of the same type—numerical, character, etc.
Matrix: a two-dimensional array with an arbitrary number of rows and columns. Subsets of the matrix may be referenced, and individual rows and columns of the matrix may be handled as vectors. Again all elements of the matrix must be of the same type.
Array: as a matrix, but of arbitrary dimension.
Data frame: a set of data organised similarly to a matrix. However each column of the data frame may contain its own type of data. Columns typically correspond to variables in a statistical study, while rows correspond to observations of these variables. A data frame may be handled similarly to a matrix, and individual columns of the data frame may be handled as vectors.
Function: see Section 7.
List: an arbitrary collection of other R objects (which may include other lists).
7 R functions
A function is one type of R object. R has a vast number of ‘built-in’ functions. Examples are mean, var and plot, in addition to those functions already mentioned above.
You may write your own functions. These enable you to reduce complex procedures—which you might wish to run repeatedly with differing inputs—to single commands. Examples of such procedures are experimental data analyses and simulations.
To create and edit functions, use the (built-in) R function edit. (As usual, type ?edit at the command prompt for details—see Section 10.)
Alternatively the function source—also available via the File menu—can be used to input lengthy function definitions from an externally edited file. (The function dump will write an existing function definition to a file.)
For details of how to get help on any function, see Section 10.
8 R packages
R is built from packages of datasets and functions. A number of packages are loaded by default and contain everything necessary for basic statistical analysis. Other packages may be loaded on demand, either via the Packages menu, or via the R function library.
Once a package is loaded, the functions and datasets within it are automatically available. To make available a dataset from within a package, use the function data
Of particular interest to advanced statistical users is the package MASS, which contains the functions and datasets from the book Modern Applied Statistics with S by W N Venables and B D Ripley. This package can be loaded with
> library(MASS)
Documentation on any package is available via the R help system—see Section 10.
A useful package containing various actuarial functions is the package actuar. This, and other packages that are not on your computer, can be downloaded and installed in R using the menu option
Packages → Install package(s)… and then Load package…
9 R graphics
Graphs are created automatically by the appropriate R functions, e.g. plot.
A graph may be resized. It may also be printed via the menu obtained by clicking on it
with the right mouse button.
On Windows a graph may be copied and pasted (or dragged and dropped) directly into a Microsoft Word document, Excel spreadsheet, or other Windows application. To copy, use the menu obtained by clicking on the right mouse button in the graph (it is probably best to use the Windows metafile format). To paste, e.g. in Word, again use the right mouse button menu.
Note the very different final effects between resizing a graph before and after copying and pasting it.
Alternatively, and perhaps more safely, a graph may be saved to an external file in any one of various formats. Again, the Windows metafile format is probably the most useful if the file is subsequently to be inserted into another Windows application. Insert a graphics file in Word via the Insert, Picture, From File menu.
10 R help and documentation
A very comprehensive help system is available via the Help menu. In particular:
• Html help starts an HTML version of the entire help system (equivalent to typing
help.start() at the R prompt);
• R functions (text) gives immediate help on a particular function or dataset (equivalent
to typing, for example, ?plot at the R prompt);
The standard manual for R is An Introduction to R by W. N. Venables, D. M. Smith and the R Development Core Team. It is available in the following formats: as HTML via the R Help, R language (html) menu; as PDF (best for printing), via R Help, Manuals menu, or from the R homepage, where you can always get the latest version.
Further resources on R are available from the R homepage at http://www.r-project.org/.
11 Useful reading
• Verzani, J. (2005) Using R for Introductory Statistics. Chapman & Hall/CRC.
• Schumacker, R.E. (2014) Learning Statistics Using R. Sage.
• Faraway, J. (2006) Extending the Linear Model with R :Generalized Linear, Mixed Effects and Nonparametric Regression Models. Chapman & Hall/CRC.