The second edition retains it commitment to the statistical programming language R. If anything the commitment is stronger. R provides access to state-of-the-art statistics, including those needed for statistical learning. It is also now a standard training component in top departments of statistics so for many readers, applications of the statistical procedures discussed will come quite naturally. Where it could be useful, I now include the R-code needed when the usual R documentation may be insuf?cient. That code is written to be accessible. Often there will be more elegant, or at least more ef?cient, ways to proceed. When practical, I develop examples using data that can be downloaded from one of the R libraries. But, R is a moving target. Code that runs now may not run in the future. In the year it took to complete this edition, many key procedures were updated several times, and there were three updates of R itself. Caveat emptor. Readers will also notice that the graphical output from the many procedures used do not have common format or color scheme. In some cases, it would have been very dif?cult to force a common set of graphing conventions, and it is probably important to show a good approximation of the default output in any case. Aesthetics and common formats can be a casualty.