Customer Reviews

Algorithms in C, Part 5: Graph Algorithms (3rd Edition) (Pt.5)

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

 

 

As students in Robert Sedgewick's Algorithms and Data Structures Course at Princeton, we had the privilege of reading Part 5 of Algorithms in C in a preprint edition this spring. Its treatment of Graph Algorithms is as thorough and comprehensive as the treatment of sorting and searching in parts 1-4. The algorithms discussed range from the fundamental (Depth-first search, Dijkstra's algorithm), to the relatively obscure (Gabow's strong component algorithm), to the impossibly difficult (Network Simplex), all in great detail. The book also discusses real-world applications of these algorithms, such as arbitrage. It contains a good number of useful diagrams allowing step-by-step traces of the algorithms, which helps decipher the sometimes cryptic code.
A warning: the book is DENSE. It is packed with detailed information and can be a difficult read, especially the mathematical analysis of the algorithms.

All in all, a great book, though.

This is a great book for the reader interested in learning a collection of graph algorithms. Though the C code isn't very clear, the text makes up for it with concise well-written explanations. And the figures do an excellent job of illustrating the execution of the algorithms.

But the problem is that it's not so great a book to study the properties of those algorithms. Most theorems are way too dense and the author relies too much on English explanations, when algebraic expressions would be a lot more explanatory. Sometimes the reader gets a sense that the author is hand-waving instead of giving proper proofs.

I bought these two book for one reason: to get help implementing algorithms. Since there is lot of code in these books, I thought this would be a good pick. I was wrong. The author is consistently leaving out details vital to understanding the code. He also makes a big deal about abstract data types. This is gloriously of the case. If the readers are supposed to understand the code, transperency would be a more intelligent goal for the author. There are no comments in the code what so ever (yes, this is actually true, not even in the on-line code). The on-line code to this book is a total mess. You would be able to implement all the algorithms from scratch in less time than it would take you to try and piece together the code the author has left for you.

The pity is that there are very few other algorithm books that have real code. Demand a new edition from the author and don't by this one unless you can avoid it.

Robert Sedgewick is certainly well known to all people involved in the development of algorithms form different fields as an author of a number of books about algorithmic methods and data structures. And I am sure that we all appreciate this work because one simply needs all the time a reference to look up not only how one can solve a probem algorithmic but also how one can do this in an efficient way.

This book provides a good overview of algorithms dealing with graphs but the problem is that the connection between the given source code in C and its general exlanation failed. It is clear, that a solution to a problem does not depend on the underlying programing language, hence, Sedgewick's book is anyway not intened to do this otherwise it would wear another title. On the other hand, an algorithm without general explanation brings not much.

I think this book has a conceptual problem. None is interested in preimplemented code, because one normaly has a certain problem the algorithm has to be adapted. In general, this book is better then nothing but far from being a complete source of information concerning the functioning of the algorithms.