g Easier!

Making Everythin

R
Learn to:
Use R for data analysis and processing Write functions and scripts for repeatable analysis Create high-quality charts and graphics Perform statistical analysis and build models

Andrie de Vries Joris Meys

Chapter 1

Introducing R: The Big Picture

In This Chapter
Discovering the benefits of R Identifying some programming concepts that make R special

ith an estimated worldwide user base of more than 2 million people, the R language has rapidly grown and extended since its origin as an academic demonstration language in the 1990s. Some people would argue and we think theyre right that R is much more than a statistical programming language. Its also: A very powerful tool for all kinds of data processing and manipulation A tool that makes all kinds of publication-quality graphics and data visualizations A collection of freely distributed add-on packages A toolbox with tremendous versatility In this chapter, we fill you in on the benefits of R, as well as its unique features and quirks.

You can download R at www.r-project.org. This website also provides more information on R and links to the online manuals, mailing lists, conferences and publications.

CO

PY

RI

GH

A community of programmers, users, academics, and practitioners

TE

MA

TE

RI

AL

10

Part I: R You Ready?

Tracing the history of R

Ross Ihaka and Robert Gentleman developed R as a free software environment for their teaching classes when they were colleagues at the University of Auckland in New Zealand. Because they were both familiar with S, a commercial programming language for statistics, it seemed natural to use similar syntax in their own work. After Ihaka and Gentleman announced their software on the S-news mailing list, several people became interested and started to collaborate with them, notably Martin Mchler. Currently, a group of 18 people has rights to modify the central archive of source code. This group is referred to as the R Development Core Team. In addition, many other people have contributed new code and bug fixes to the project. Here are some milestone dates in the development of R: Early 1990s: The development of R began. August 1993: The software was announced on the S-news mailing list. Since then, a set of active R mailing lists has been created. The web page at www.r-project. org/mail.html provides descriptions of these lists and instructions for subscribing. (For more information, turn to It provides an engaged community, later in this chapter.) June 1995: After some persuasive arguments by Martin Mchler (among others) to make the code available as free software, the code was made available under the Free Software Foundations GNU General Public License (GPL), Version 2. Mid-1997: The initial R Development Core Team was formed (although, at the time, it was simply known as the core group). February 2000: The first version of R, version 1.0.0, was released. Ross Ihaka wrote a comprehensive overview of the development of R. The web page http:// cran.r-project.org/doc/html/ interface98-paper/paper.html provides a fascinating history.

Recognizing the Benefits of Using R

Of the many attractive benefits of R, a few stand out: Its actively maintained, it has good connectivity to various types of data and other systems, and its versatile enough to solve problems in many domains. Possibly best of all, its available for free, in more than one sense of the word.

It comes as free, open-source code

R is available under an open-source license, which means that anyone can download and modify the code. This freedom is often referred to as free as in speech. R is also available free of charge a second kind of freedom, sometimes referred to as free as in beer. In practical terms, this means that you can download and use R free of charge.

Chapter 1: Introducing R: The Big Picture

Another benefit, albeit slightly more indirect, is that anybody can access the source code, modify it, and improve it. As a result, many excellent programmers have contributed improvements and fixes to the R code. For this reason, R is very stable and reliable. Any freedom also has associated obligations. In the case of R, these obligations are described in the conditions of the license under which it is released: GNU General Public License (GPL), Version 2. The full text of the license is available at www.r-project.org/COPYING. Its important to stress that the GPL does not pertain to your usage of R. There are no obligations for using the software the obligations just apply to redistribution. In short, if you change or redistribute the R source code, you have to make those changes available for anybody else to use.

11

It runs anywhere
The R Development Core Team has put a lot of effort into making R available for different types of hardware and software. This means that R is available for Windows, Unix systems (such as Linux), and the Mac.

It supports extensions
R itself is a powerful language that performs a wide variety of functions, such as data manipulation, statistical modeling, and graphics. One really big advantage of R, however, is its extensibility. Developers can easily write their own software and distribute it in the form of add-on packages. Because of the relative ease of creating these packages, literally thousands of them exist. In fact, many new (and not-so-new) statistical methods are published with an R package attached.

It provides an engaged community

The R user base keeps growing. Many people who use R eventually start helping new users and advocating the use of R in their workplaces and professional circles. Sometimes they also become active on the R mailing lists (www.r-project.org/mail.html) or question-and-answer (Q&A) websites such as Stack Overflow, a programming Q&A website (www. stackoverflow.com/questions/tagged/r) and CrossValidated, a statistics Q&A website (http://stats.stackexchange.com/questions/ tagged/r). In addition to these mailing lists and Q&A websites, R users participate in social networks such as Twitter (www.twitter.com/search/ rstats) and regional R conferences. (See Chapter 11 for more information on R communities.)

12

Part I: R You Ready?

It connects with other languages

As more and more people moved to R for their analyses, they started trying to combine R with their previous workflows, which led to a whole set of packages for linking R to file systems, databases, and other applications. Many of these packages have since been incorporated into the base installation of R. For example, the R package foreign (http://cran.r-project.org/ web/packages/foreign/index.html) is part of the standard R distribution and enables you to read data from the statistical packages SPSS, SAS, Stata, and others (see Chapter 12). Several add-on packages exist to connect R to database systems, such as the RODBC package, to read from databases using the Open Database Connectivity protocol (ODBC) (http://cran.r-project.org/web/packages/RODBC/ index.html), and the ROracle package, to read Oracle data bases (http:// cran.r-project.org/web/packages/ROracle/index.html). Initially, most of R was based on Fortran and C. Code from these two languages easily could be called from within R. As the community grew, C++, Java, Python, and other popular programming languages got more and more connected with R. Because many statisticians also worked with commercial programs, the R Development Core Team (and others) wrote tools to read data from those programs, including SAS Institutes SAS and IBMs SPSS. By now, many of the big commercial packages have add-ons to connect with R. Notably, SPSS has incorporated a link to R for its users, and SAS has numerous protocols that show you how to move data and graphics between the two packages.

Looking At Some of the Unique Features of R

R is more than just a domain-specific programming language aimed at statisticians. It has some unique features that make it very powerful, including the notion of vectors, which means that you can make calculations on many values at the same time.

Performing multiple calculations with vectors

R is a vector-based language. You can think of a vector as a row or column of numbers or text. The list of numbers {1,2,3,4,5}, for example, could be

Chapter 1: Introducing R: The Big Picture

a vector. Unlike most other programming languages, R allows you to apply functions to the whole vector in a single operation without the need for an explicit loop. Well illustrate with some real R code. First, well assign the values 1:5 to a vector that well call x:
> x <- 1:5 > x [1] 1 2 3 4 5

13

Next, well add the value 2 to each element in the vector x and print the result:
> x + 2 [1] 3 4 5 6 7

You can also add one vector to another. To add the values 6:10 elementwise to x, you do the following:
> x + 6:10 [1] 7 9 11 13 15

To do this in most other programming language would require an explicit loop to run through each value of x. This feature of R is extremely powerful because it lets you perform many operations in a single step. In programming languages that arent vectorized, youd have to program a loop to achieve the same result. We introduce the concept of vectors in Chapter 2 and expand on vectors and vectorization in much more depth in Chapter 4.

Processing more than just statistics

R was developed by statisticians to make statistical processing easier. This heritage continues, making R a very powerful tool for performing virtually any statistical computation. As R started to expand away from its origins in statistics, many people who would describe themselves as programmers rather than statisticians have become involved with R. The result is that R is now eminently suitable for a wide variety of nonstatistical tasks, including data processing, graphic visualization, and analysis of all sorts. R is being used in the fields of finance, natural language processing, genetics, biology, and market research, to name just a few.

14

Part I: R You Ready?

R is Turing complete, which means that you can use R alone to program anything you want. (Not every task is easy to program in R, though.) In this book, we assume that you want to find out about R programming, not statistics, although we provide an introduction to statistics in R in Part IV.

Running code without a compiler

R is an interpreted language, which means that contrary to compiled languages like C and Java you dont need a compiler to first create a program from your code before you can use it. R interprets the code you provide directly and converts it into lower-level calls to pre-compiled code/functions. In practice, it means that you simply write your code and send it to R, and the code runs, which makes the development cycle easy. This ease of development comes at the cost of speed of code execution, however. The downside of an interpreted language is that the code usually runs slower than compiled code runs. If you have experience in other languages, be aware that R is not C or Java. Although you can use R as a procedural language such as C or an object-oriented language such as Java, R is mostly based on the paradigm of functional programming. As we discuss later in this book, especially in Part III, this characteristic requires a bit of a different mindset. Forget what you know about other languages, and prepare for something completely different.