Customer Reviews

Applied Functional Data Analysis

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

 

 

In 1997 Ramsay and Silverman wrote a standard text that stands out as a classic book teaching the functional approach. With this new book they take the approach of teaching by example.

Each of chapters 2-12 present a case study first. The one on childhood growth is standard fair but they also present unusual cases such as modeling a juggler and analyzing handwriting. In eachcase they show how special characteristics of the data are best portrayed by treating curves or functions as individual data points. The application is presented first. Then the authors show how as consultants they formuated a functional data approach to the problem. Sometimes transformations are applied first. In another case principal component curves are constructed. After each problem is presented with the approach taken, the authors go back and present a "what we have seen" section that reviews the key points and lessons learned from the examples.

This is a wonderful way to make the material very accessible to practitioners as well as statisticians and it illustrates the variety of problems and that can be handle by this approach to data analysis.

A nice, if overpriced, book, but not one that you need to keep, unlike Ramsay and Silverman's "Functional Data Analysis" (FDA). AFDA elaborates several examples introduced in FDA. The examples cover about 60% of FDA's material: smoothing is given much less space than in FDA, and regression (cf. FDA's Chapters 12-15) is left out. If you are choosing between the two books, go for FDA.