First of all, let me tell you on advance that I have not finished the book, but I am not nearly close to the "first pages" either: I think I have read enough of it to give it a good evaluation. I will post updates later, as need arises. UPDATE: I have already finished the book.
I don't know. Maybe it's because its presentation, the way the author narrates, or just because the renown of the book makes you biased, but as soon as you start reading the very first pages you know you are in for something very special. You know that, after reading this book, you will be a much better computer scientist. I will not lie: I am still in attending college, but I have always been a great programmer, and in all my programming classes I am top of the class. I was really full of myself before buying this book. How wrong was I!
The first thing you get after the preface is some advices on the exercises. The author tells you how some of them are really hard, or some are just really easy, etc. He also states that "There are far too many exercises for you to do read and assimilate them all..." When I read this, I gleaned each section and the amount of exercises (there are literally, thousands), and I said to myself: "I will work on, at the least, 80% of the exercises on each section. I will never avoid the hard or very hard ones" (Sedgewick explicitly tells you how to differentiate them). Alas, even some exercises that are not considered to be hard have been hell on me. Until know, I have been true to my word, but I will get into that later.
Of course, this book is not about the exercises, is about Algorithms, and I'd be dammed if Sedgewick is not considered among the best authors of all times on the topic. I've read other books on Algorithms, and I always find that most books fit into one of these two spectrums: either they are recipes, or they are strictly academic, with little practical use. Algorithms in C++ is neither. Sedgewick makes a deep and interesting analysis on the properties of each algorithm (and data structure), concentrating at first on its abstract properties, not even daring talk about implementations. He is extremely clear in his language. Look, I am an Ecuadorian, living in Ecuador, and my English is pretty much what I have learned in my school or by myself. I can understand Sedgewick perfectly. I don't think I have had to reread a thing twice because the writing was obscure.
After being all abstract and wonderfull and stuff, Sedgewick then implements, and he does so magnificently. Yes, he may have not the best of C++ practices (for example, he explains what C++ operator delete does, but seems to dislike the use of it), or even good programming practices at all (you have, in one program, variables named like: p, a, x, y, z, f), but his implementations are nevertheless clear, straight to the point and efficient. He also constantly reminds the reader how important is to separate implementations from abstract concepts, but also explains the advantages certain implementations have against others.
So, you have learned the concept of the algorithm and it's properties, implemented it, and just when you think it as all over, Sedgwick goes ahead and refines the abstract model of the algorithm, implements it again, and thus a refinement iteration process emerges! All this is aided by graphics, more programs, and clear and concise definitions (which are separated from the text in italics). This whole process is done in one section or many. For example, he explains Pushdown-Stacks in several sections, but using this same process. Don't get me wrong now, Sedgwick does not use a "template" for explaining each topic: he will change his ways of explaining things if necessary, and as a whole the book reads very naturally, it never gets old or repetitive.
Once you finish each section, you get a bunch of exercises. Trust me: WORK ON THEM. Especially the ones marked with unfilled circles. They will bring new insights on the topic (whereas after such grand explanations, you would think more insight could not be obtained). As I said before, I have worked, until now, on more than the 80% of all the exercises. Some require background programming experience, some a high level of math (but not never too high), and they will slow your reading down (there are loads and loads of them!), but if you do them, you will be cool. Your projects will be most efficient and elegant. Teachers will admire your work (though secretly you know they should be admiring Robert Sedgewick).
One final note: this is not an academic book on the analysis of algorithms. This book is strictly for people who really want to put the best algorithms and data structures into practice. Sedgewick will tell you about the O notation, and what it means for an algorithm to run proportional to some function, but it is not the goal of the book to make you sit down and mathematically extract the efficiency of an algorithm. For this Sedgwick has another book, which people hail and glorify too. But with Algorithms in C++ (and C, and Java) you will be able to sit in front of a computer, analyze some problem (or a previously implemented solution to that problem) and determine the most effective way of solving it (or if you can replace the code with a more efficient one).
After walls of text, I hide again. I really hope this helped.
- Amazon Business : For business-only pricing, quantity discounts and FREE Shipping. Register a free business account
