Which one is faster - 未名空间MITBBS历史存档

国际科技财经博客移民网络热点娱乐民生时事公众号

Redian新闻

>未名空间

>DotNet - 窗口里的风景

Which one is faster

Which one is faster# DotNet - 窗口里的风景

c*t2011-02-18 08:02

1 楼

Given List targets (around 20 elements), how check if string s is in
the list, we have the following ways:
case 1: targets.Contains(s);
case 2: if-else...
case 3: switch ...
case 4: s == e1 || s == e2 || ...
which one is faster?

a92011-02-18 08:02

2 楼

你这是找几个啊？呵呵。

in

【在 c**t 的大作中提到】

: Given List targets (around 20 elements), how check if string s is in
: the list, we have the following ways:
: case 1: targets.Contains(s);
: case 2: if-else...
: case 3: switch ...
: case 4: s == e1 || s == e2 || ...
: which one is faster?

s*o2011-02-18 08:02

3 楼

my guess:
switch-case is the fastest since most compilers optimize it.
Contains is the slowest due to some overhead.
Conclusion: who cares...

is in

【在 c**t 的大作中提到】

a92011-02-18 08:02

4 楼

most compilers....
难道.net不就一个编译器吗？呵呵

【在 s***o 的大作中提到】

: my guess:
: switch-case is the fastest since most compilers optimize it.
: Contains is the slowest due to some overhead.
: Conclusion: who cares...
:
: is in

c*d2011-02-18 08:02

5 楼

switch怎么找法？

r*y2011-02-18 08:02

6 楼

20个元素？这点区别应该很小吧

in

【在 c**t 的大作中提到】

a92011-02-18 08:02

7 楼

貌似有个阀值，如果大于那个阀值，就会创建一个dictonary，在dictonary里查找。

【在 c**d 的大作中提到】

: switch怎么找法？

c*t2011-02-18 08:02

8 楼

randomly generated 20 elements and a new value to lookup, 10000 iteration:
case 2 < case 3 < case 4 < case 1. If-else runs fastest.

Given List targets (around 20 elements), how check if string s is in
the list, we have the following ways:
case 1: targets.Contains(s);
case 2: if-else...
case 3: switch ...
case 4: s == e1 || s == e2 || ...
which one is faster?

【在 c**t 的大作中提到】

o*e2011-02-18 08:02

9 楼

dont waste your time, just use list.Contains
generic List is smart, it builds a hashtable of the collection is big. I've
tested with 1k elements, List.Contains has no performance diff compared to
Dictionary.ContainKey

v*n2011-02-18 08:02

10 楼

I think it's using linear search. You can step into .net fx source code
to see the implementation
http://www.hanselman.com/blog/NETFrameworkLibrarySourceCodeAvai

ve

【在 o****e 的大作中提到】

: dont waste your time, just use list.Contains
: generic List is smart, it builds a hashtable of the collection is big. I've
: tested with 1k elements, List.Contains has no performance diff compared to
: Dictionary.ContainKey

c*t2011-02-18 08:02

11 楼

I like list.contains more than other ways, even it's performing nano seconds
longer. The code is much cleaner, and easier to maintain and to scale.

s*o2011-02-18 08:02

12 楼

scalability vs efficiency, there is always something to weigh and make
one's head spin

seconds

【在 c**t 的大作中提到】

: I like list.contains more than other ways, even it's performing nano seconds
: longer. The code is much cleaner, and easier to maintain and to scale.

a92011-02-18 08:02

13 楼

List也会创建hashtable?

ve

【在 o****e 的大作中提到】

s*o2011-02-18 08:02

14 楼

from MSDN
"This method performs a linear search; therefore, this method is an O(n)
operation, where n is Count."

【在 a9 的大作中提到】

: List也会创建hashtable?
:
: ve

a92011-02-18 08:02

15 楼

也就是说不会创建喽？呵呵。

【在 s***o 的大作中提到】

: from MSDN
: "This method performs a linear search; therefore, this method is an O(n)
: operation, where n is Count."

o*e2011-02-18 08:02

16 楼

take my word back, list with several k is not enough to tell the difference.
the List.Contains call is apparently O(n), just tested with 1M and 10M list
. well, even for such big list, the difference is only 20ms vs 200ms.