Customer Reviews

Statistical Models in S

5 star:		(3)
4 star:		(0)
3 star:		(0)
2 star:		(0)
1 star:		(0)

 

 

S programmers refer to this as "the white book", and it is a key reference for understanding the methods implemented in several of S-PLUS' high-end statistical functions, including 'lm()', predict()', 'design()', 'aov()', 'glm()', 'gam()', 'loess()', 'tree()', 'burl.tree()', 'nls()' and 'ms()'.

It's apparently out of print, but it shouldn't be.

Even with the recent arrival of S-PLUS releases that incorporate S version 4 and many of the ideas discussed in "the green book" (<<Programming with Data>>, also by John Chambers), this classic S reference is an indispensable tool for the serious statistician. It needs to be reissued--with a white cover, of course.

Here are the titles of the chapters, for reference:

1. An Appetizer

2. Statistical Models

3. Data for Models

4. Linear Models

5. Analysis of Variance: Designed Experiments

6. Generalized Linear Models

7. Generalized Additive Models

8. Local Regression Models

9. Tree-Based Models

10. Nonlinear Models

A. Classes and Methods: Object-oriented Programming in S

B. S Functions and Classes

References

Index

If you really want to know what you're doing when you use S, buy this book. Don't waste your money on a book like Venables and Ripley -- you will be sorely dissappointed, unless you just want a large collections of example calls to canned S routines. The authors of the present book, on the other hand, are Chambers and Hastie of AT&T (where S was invented), and they clearly understand the importance of detailed explanations of the theory underlying the S functions they describe. Just as important, in my opinion, they also describe the algorithms used by these functions. These two components are missing from other books (like the popular Venables and Ripley) but they are critical in order to know -- and be able to explain and justify to others -- how and why your statistical analyses were performed and what the results really mean. The other way of doing statistics (i.e. throwing canned procedures at your data and seeing what pretty graphs and figures you can produce) is meaningless.

I am in the process of learning R, the open source implementation of the S language. This book is one of the classics describing the original S language. While some small parts, of this book, are now out of date, it remains a great source, of information about the design and use of S and by extension, of R.