Kansas City Perl Mongers: Book Review

That said, the book has something for everyone. Beginner and Intermediate level Perl programmers will find the book's materials information dense but approachable. Advanced programmers will find a well written refresher illustrating familiar concepts in Perl. Given the breadth of the book and the language specific insights, even experience Perl programmers should expect to learn a thing or two.

Topics covered include: linked lists, circular linked lists, garbage collection, doubly-linked lists, infinite lists, binary trees, heaps, binary heaps, janus heaps, sorting, searching, sets, matrices, graphs, strings, geometric algorithms, number systems, number theory, cryptography, probability, statistics, and numerical analysis.

I've found that while the book is somewhat daunting, it goes down best a little bit at a time: read a chapter, then work through the examples. The first time I attempted to read the book I got about a third of the way through before stopping, scanning the rest, and putting it back on the shelf. A couple months later as I was working on a problem involving graphs and trees, the book came back down. And as interesting and complicated problems present themselves, I often find myself reaching for it.

It covers an awful lot in 701 tightly packed pages. Have you ever wondered how regular expressions work? Hashes? Insights into their internal workings are in there. It is an excellent reference and fills a much needed gap. No other Perl book comes close to the breadth and depth coverage of these materials. That said, early printings of the first edition had a tremendous number of errors and typos. If you buy it used, I would highly recommend visiting the O'Reilly site to read the errata. ...

Execellent Book

     

When I heard that O'Reilly was publishing a book on Algorithms in Perl I couldn't wait to get my hands on it. Well last month I did and it was great!

The clearly written text contains the usual light, easy-reading tone and occasional humorous elements found in most O'Reilly books. The authors include plenty of pictures and diagrams for those who learn visually (rather then by reciting words out loud). The Perl code within is concise, with comments when necessary, and makes use of the objects when possible. If you plan to read this book you should know Perl because the more advanced level of the code could cause problems for the non-Perl or beginning Perl Programmer. However, to a Perl programmer who is comfortable with the language the code reads clean and understandably - sometimes it's even more clear then pseudocode.

The text covers a broad range of topics (with varying levels of complexity). When I was reading I recalled things I learned in college classes such as: Data Structures, Algorithm Analysis, Discrete Math, Calculous, Linear Algebra, Statistics, Compiler Design, Signal Processing, and even some good old fashion high school geometry. I found this extremely helpful because the broad nature of the book doesn't allow the authors to cover a topic in great detail. They do review each topic area giving the proper terminology used along with background of how the field developed.

Within the different chapters the authors present various code segments. For some segments the authors have written there own code to implement the algorithms. In other cases, as is Perl custom, the authors have searched CPAN for the modules that implement the algorithm. Then the example code demonstrates the proper use of that module.

One of the features I really enjoyed is that each chapter can stand on its own as a nice review of the algorithms in that section. (In cases where they build on other sections you are reminded where to go and read.) Another great feature the authors include is the references - all the web sites and books you'll need are listed in Appendix A, by topic area.

The only thing I really felt was missing was a discussion on some AI topics such as Neural Networks and Genetic Programming. (If you're interesting Neural Networks in general check out The Linux Journal July 1999 p 44. For Genetic Programming in Perl check out The Perl Journal Fall 1999 p 34)

Overall, this book explains methods of implementing Algorithms with Perl blending custom techniques with resources available (CPAN) in a "learn by example" approach. It contains 16 great chapters of background, theory, sample code, diagrams, and discussion. It has a good Appendix A (for additional info on algorithms) and a useful ASCII table in Appendix B. If you want to learn good ways to implement specific Algorithms with Perl - Read this book!

Excellent ! A "cookbook"...but for algorithms in Perl

     This book is certainly not meant for learning Perl nor being read from the first page to the last...it's a compilation of some of the advanced features of Perl aimed at resolving very specific algorithmic problems.

Two examples :

1) i had an efficiency problem with my code that required boosting some subroutines that were heavily used but could not be easily implemented with a data structure...so i discovered the "Memoize" module that performs caching on subroutines, allowing me to speed up my code by a factor of 10 !

2) i had to find the maximum likelihood estimation of several probability distributions on some important amount of data, and for multi-parameter distributions. I just used the complex module and LU decomposition to find the roots of my non-linear systems, which can be "easily" done by specifying the equations to be resolved (expressed as subroutines which values are instanciated by the resolver) and applying a heuristic root finding algorithm based on Newton's method combined with LU decomposition. It would have taken several days or even weeks to write it from scratch in C or C++, but most of the code already existed so it was written within one day !

Of course, it's not a general-purpose book on Perl programming but rather a cookbook for some very specific utilization of Perl like "non-linear equations systems root finding" or graphs theory problems...so don't expect this to be a well-organized book since it's not intended to be that way !