R Graphics (Chapman & Hall/CRC The R Series) (英語) ハードカバー – 2005/7/29
Kindle 端末は必要ありません。無料 Kindle アプリのいずれかをダウンロードすると、スマートフォン、タブレットPCで Kindle 本をお読みいただけます。
R is revolutionizing the world of statistical computing. Powerful, flexible, and best of all free, R is now the program of choice for tens of thousands of statisticians.
Destined to become an instant classic, R Graphics presents the first complete, authoritative exposition on the R graphical system. Paul Murrell, widely known as the leading expert on R graphics, has developed an in-depth resource that takes nothing for granted and helps both neophyte and seasoned users master the intricacies of R graphics. After an introductory overview of R graphics facilities, the presentation first focuses on the traditional graphics system, showing how to work the traditional functions, describing functions that are available to produce complete plots, and how to customize the details of plots.
The second part of the book describes the grid graphics system - a system unique to R and much more powerful than the traditional system. The author, who was integral in the development of the grid system, shows, starting from a blank page, how it can be used to produce graphical scenes. He also describes how to develop new graphical functions that are easy for others to use and build on. Appendices contain a brief introduction to the R system in general and discuss how the traditional and grid graphics systems can be combined.
Much of the information presented in this book cannot be found anywhere else. Well ahead of the curve, particularly regarding the grid system, R Graphics will have a major impact on the future direction of statistical graphics development.
The author maintains a website with more information.
"With R having become the lingua franca of statistics, 'R Graphics' is a must for many useRs and programmmeRs: Flexible programmable graphics having been a major strength of S from its beginning; this is even more true for R which has both improved the traditional graphics from S and introduced the new much more flexible 'grid' graphics system. Paul Murrell, a member of the R Core Development Team, has not only been the main author of 'grid' but has also been responsible for several recent enhancements to the underlying R graphics engine. Together with its online companion web site, this book will be an indispensable resource for almost everyone interested in how to produce R graphics efficiently and intelligently." -- Dr Martin Maechler, Seminar for Statistics, Swiss Federal Institute of Technology Zurich, Switzerland "This book starts where the graphics sections of other books on using S for data analysis typically end: high-level plots and their default settings. If everything you want to do to visualize your data can be done using the standard settings, then this book is not for you. But if you ever wanted to go beyond that line, from changing bits and pieces of a graph to writing your own visualization functions, then R Graphics'' has all you need to know (and much more). Starting with the basic plotting commands most users are familiar with from introductory texts, the book gives a comprehensive overview of the current state and design principles of visualizing data with R. Paul Murrell is one of the main authors of R's graphical facilities, and inventor of completely new features like the grid system or expressions for annotation of plots with mathematical formulae. His book iswritten in the spirit of S itself: It takes the reader on a journey, where beginners gradually are turned into programmers while learning the language, having ample material for both novices and experts. It will certainly claim its place on the bookshelf of reference guides next to my desktop." -- Friedrich Leisch, Institute for Statistics, Technical University of Vienna, Austria "R Graphics is exactly the sort of documentation that R needs. It is written clearly, with many examples, and will be useful for any level of R expertise from novice upwards. It contains more than a hundred figures containing model code and its output. There are extensive cross-references that make finding detailed information easy. My copy of the book is from the first printing, but it is exceptionally free of typographical and other errors. I've been using traditional S graphics in S-PLUS and R for 17 years, so I am very familiar with the system. However, there are some details that I've never memorized, so I've often needed to consult the manual page for the par() function. R Graphics will now be the first place to look for those sorts of details, specifically Chapter 3, which contains a series of diagrams and tables illustrating the choices. Being such an old-time user, I was not so familiar with some of the newer functions, such as layout() and xyz.coords(), and I have already modified some of my own code to make use of them. I was also very impressed with the book's descriptions of the grid and lattice packages. I have not studied the grid system before, though I have heard Murrell speak about it at conferences. The description in this book is perfect. It takes the reader from the basics through todevelopment of new types of graphics. After reading it, I feel that I understand the philosophy behind the design of grid, and am eager to make use of it in my own work. I especially appreciated the design advice in Chapter 7; I will be making use of it, and referring my students to it. I think every R user who uses graphics (which is essentially every R user) should have a copy of this book. The grid graphics package was a wonderful development, and this book is another one. Murrell is to be congratulated. -- Duncan Murdoch, University of Western Ontario, Canada
The book is divided into four major sections and there are ample examples of graphs and code. Further, each chapter starts with a brief preview and ends with a short summary paragraph. Overall, it was one of the clearest technical books I have read, and I thoroughly enjoyed it (I could not put it down all week :).
Part I deals with "traditional" graphics. It starts with an introduction to the high level functions that create typical plots like scatter plots, but also covers more esoteric types such as mosaic plots. It quickly moves on to low level functions that draw points, text, etc.; how the plotting regions and margins work, and how all of this can be edited and customized. By the end of the section, readers are introduced to all the tools to build their own custom plots complete with axes, annotations, legends, titles, even special shapes.
Part II introduces the grid graphics system. This is a massive section and nothing short of phenomenal. Two chapters are dedicated to introducing the lattice and ggplot2 packages. For readers interested in thoroughly learning either of those packages, I would suggest reading the book on Lattice: Multivariate Data Visualization with R (Use R) or ggplot2: Elegant Graphics for Data Analysis (Use R). The rest of the section (three chapters) is dedicated to the grid graphics model and how this can be used on its own to create plots and to edit and customize output from lattice and ggplot2. This allows the convenience of the lattice/ggplot2 systems while giving all the power of the low level grid functions. It also provides suggestions and guidelines for developing new grid graphical objects and functions (one of the main reasons I read the book).
Part III briefly introduces the underlying graphical engine and the various devices available (e.g., pdf, jpeg).
Part IV covers a host of addon graphics packages. For example, drawing maps, 3D plots, interactive and dynamic plots.
Other reviews discuss the contents and coverage so I won't repeat that. I'm a somewhat experienced R user - I'm still struggling to switch from SAS, but have used R exclusively for the past couple years, including a few fairly involved projects. When working with R, I'm usually (and it seems constantly) pulling an R reference off the shelf to help produce a specific product or analysis. I'd hoped this book would be like the SAS graphics documentation, with detailed descriptions of commands, good examples of code and output, and a bit of logic or 'how to'. This is a solid description of the graphics systems and it's certainly helped me understand the foundations of R graphics, but I've found it a hard-to-use reference book, and instead I'm usually pulling Venables and Ripley or something else (e.g. Crawley) off the shelf to find the specific graphics commands I need or for an example plot with the code that produced it. There are very few examples of using maps or other geographical backgrounds despite the plethora of analyses that are geographical, and the power of the maps package. If the description of the new version is accurate, it should address a number of these issues so I'd wait for the 2nd edition. I think much of the information I want is in this book, but the 'style' and organization of the book just doesn't seem to match the way I work.
By page count, this book devotes ~40% to traditional R graphics, 10% to lattice, and 40% to grid, in that order. The traditional graphics coverage is good, with interesting material on margins and other layout arcana. The grid coverage is quite technical, targeted largely at developers and power-users. The 10% lattice isn't especially illuminating. The grid material does inform the mechanisms behind lattice, and show how to mix grid and lattice. The traditional graphics material, does not apply to either grid or lattice.
If you're like me, this book isn't a good "first book". It covers the basics (which i'm comfortable with) and the complex (which i don't usually need), without a lot of middle ground, especially with respect to lattice. A good alternative might be "Lattice: Multivariate Data Visualization with R by Deepayan Sarkar" or ggplot2: Elegant Graphics for Data Analysis (Use R) by Hadley Wickham.
This book is well-written, and the grid coverage is thorough, but it could use a more descriptive title.
If you do know R, what this book *will* show you is how to do more complex things with R graphics. Half the book covers the traditional graphics model, while the other half covers the Grid and Trellis graphics models. This will be the interesting part of the book because Grid and Trellis look like they let users create really neat graphs and data representations with R.
I would have liked to see some more complete examples in the book, but at least there's an accompanying website that contains all the code used to generate the graphs and errata for the book. This would be a good addition to an R user's bookshelf.