p*****2
发帖数: 21240 | 1 周三入手看了一半，怎么感觉什么都没看到呢。
就感觉几句话反复的讲，又经常引用别人说了什么。是我理解能力太差了吗？ |
k*********6
发帖数: 738 | 2 read and think, and you will get it... it is not a book teaching you how to
solve a few programming problems. It is the book teach me the way to think..
. |
w***y
发帖数: 6251 | |
h*******s
发帖数: 8454 | 4 不是有几个题挺常见的从这里来的么
【在 p*****2 的大作中提到】
: 周三入手看了一半，怎么感觉什么都没看到呢。
: 就感觉几句话反复的讲，又经常引用别人说了什么。是我理解能力太差了吗？
|
p*****2
发帖数: 21240 | 5 后半部看样子有一些内容，排序方面的。还有堆。等看完了再讨论一下。 |
p*****2
发帖数: 21240 | 6 I was really puzzled by some of the language in this book. Many books use
example from real life to make a problem easily understood. But this book
uses real example to complicate the problem. Many problems in this book can
be summarized or formulated by concise and easily understood sentences. But
this book use examples to make the problem complicate. Sometimes I feel that
I have walked a long way to reach what I can reach by one step after
reading the verbose examples; sometimes I just want to say "what at hell are
you talking about". |
p*****2
发帖数: 21240 | 7 "Programming Pearls" summarizes many important topics in Computer Science,
usually in a sub-par manner. About 10% of the book is dedicated to "thinking
outside the box," and I'd say these parts really shine. Unfortunately, the
other 90% isn't nearly as good. There are inconsistencies in notation,
improperly explained terminology, and incomplete analyses. For example, his
chapter on structured testing leaves out randomized testing (which I would
argue is just as important), despite the fact that it takes little space to
explain. I felt most chapters had deficiencies.
While I have a long list of complaints, my biggest one is his chapter
promoting a colleague's work on the n-body simulation problem, which at
times reads like an advertisement. He discusses each speed optimization
without discussing the impact on the error rate. Although he states his
focus is on performance, I can't imagine any good programmer ignoring
something as important as the error rate (a fast answer may be worthless if
it is incorrect)! He could have used this opportunity to have a great
discussion of the trade-off between speed and accuracy. Ignoring this aspect
struck me as odd because this kind of critical thinking (examining all
aspects of a system) is a key tenet to being a good programmer. It felt like
a very poor example, where the author was more concerned with promoting a
colleague than teaching the reader how to think critically.
I'm sure this book will definitely be helpful to neophytes. However, I think
there are better treatments of these subjects out there. No good book in
its 2nd edition should have some of the mistakes that are present in "
Programming Pearls". |
p*****2
发帖数: 21240 | 8 It's a collection of lessons appropriate for college students, nothing more.
If you've seen your fair share of sorts and trees, don't feel pressured by
my fellow reviewers into thinking this book is some sort of essential
programmers' bible.
The advice so highly praised here is borderline commonsensical: ask why you
need to change code before you do it, get as many specification details as
you can before you start a project, consider running times and minimizing
the amount of data you need to use.
Pearls is short on content for us non-newbies. |
p*****2
发帖数: 21240 | 9 This book is really addressed towards absolute beginner programmers. Maybe
if you've only taken an introductory course or two at college, or just have
the syntax of one language under your belt and want a broader perspective
this book will be useful.
If you have any experience, the examples are all very mundane. (E.g. "this
is what a minheap is"). If you have just finished reading a "Learn [Language
] in 24 hours!" type book, I guess this wouldn't be a bad follow up.
Otherwise, steer clear. |
w***y
发帖数: 6251 | 10 老大 6－9楼是你自己写的？ 这么大段的英文， admire |
|
|
l*****a
发帖数: 14598 | 11 这本书有用的算法就那么几个
最大连续和，array shuffle...
【在 p*****2 的大作中提到】
: 周三入手看了一半，怎么感觉什么都没看到呢。
: 就感觉几句话反复的讲，又经常引用别人说了什么。是我理解能力太差了吗？
|
l*********8
发帖数: 4642 | 12 这几段的观点是不一致的。所以我觉得是亚马逊的用户书评。
【在 w***y 的大作中提到】
: 老大 6－9楼是你自己写的？ 这么大段的英文， admire
|
p*****2
发帖数: 21240 | 13
smart
【在 l*********8 的大作中提到】
: 这几段的观点是不一致的。所以我觉得是亚马逊的用户书评。
|
p*****2
发帖数: 21240 | 14
这几个算法都知道了呀
【在 l*****a 的大作中提到】
: 这本书有用的算法就那么几个
: 最大连续和，array shuffle...
|
h****e
发帖数: 928 | 15 这本书第一版几年前看过，当时觉得最大连续数和的O(N)
解法，counting sort很巧，其它的就没有什么印象了。
后来看了面试题和多看了几章CLRS以后才知道这些只是
入门级的。或许自己当时太菜鸟了。 |
l*****a
发帖数: 14598 | 16 so u can ignore it
【在 p*****2 的大作中提到】
:
: 这几个算法都知道了呀
|
w****x
发帖数: 2483 | 17 有一个证明程序的正确性那章还不错, 我其实比较推荐algorithm in C, 真的很有用 |
p*****2
发帖数: 21240 | 18
那不白买了？哎。看来还得看CLRS呀。
【在 l*****a 的大作中提到】
: so u can ignore it
|
x*******5
发帖数: 152 | 19 这个书是要慢点看的，它的前言也说了，这本书前言说要求你尽全力去搞懂每一道习题
，不要急，相信它的话。是，就算法而言，这个肯定没有CLRS讲的多，但是本质上它会
给你思考的方法，因为面试题太多了（>=1k and keep growing，如果你去做它的每一
道题，我是说不是想想大概意思，而是如果能编程就编到没有BUG，如果能思考就使劲
思考，比如它的back of the envelop这章没有编程就是思维，我第一遍整整用了60天。
【在 p*****2 的大作中提到】
: 周三入手看了一半，怎么感觉什么都没看到呢。
: 就感觉几句话反复的讲，又经常引用别人说了什么。是我理解能力太差了吗？
|
x*******5
发帖数: 152 | 20 整本书读完，印象最深的却是它开篇的那个例子，如果进行排序。它启迪我的研究在于
对于任何一个问题需要思考的是问题本身的shape，然后根据自己的经验选择或者设计
适合这个问题的solution，这个对我有极大的启发，启发我对每一个问题尽全力去思考
它特殊的shape，而方法就像积木，如果有一样形状的最好，如果没有，还得自己上锯
子磨啊
后来在machine learning中我看到Vipnik Principle: never design a solution more
general than the problem. 大呼叫好啊 |
