简体   繁体   English

如何转换清单 <string> 到字典 <string, int> ?

[英]How can I transform a List <string> to a Dictionary<string, int>?

How can I transform a List <string> to a Dictionary<string, int> ? 如何将List <string>转换为Dictionary<string, int>

This should be number from zero to n. 这应该是从零到n的数字。

You can use ToDictionary() to create a Dictionary from any IEnumerable . 您可以使用ToDictionary()从任何IEnumerable创建Dictionary

var list = new List<string>() { "Foo", "Bar", "Spam" };

// TKey is int, TValue is string
int i = 0;
Dictionary<int,string> dict1 = list.ToDictionary( _ => i++ );          

// TKey is string, TValue is int
i = 0;
Dictionary<string,int> dict2 = list.ToDictionary( x => x, _ => i++ );

The lambda passed to ToDictionary is the keySelector . 传递给ToDictionary的lambda是keySelector Usually you use this to select a property from the items in the IEnumerable to be the key, but here, we use it to provide a count. 通常,您可以使用它从IEnumerable的项中选择一个属性作为键,但是在这里,我们使用它来提供计数。

Edit: The second version uses two selectors. 编辑:第二个版本使用两个选择器。 The first is the key. 首先是关键。 By providing a x => x lambda, we are just using the string that came from the list. 通过提供x => x lambda,我们仅使用了来自列表的字符串。 The second selector is the value. 第二个选择器是值。 Here, we are providing the counter, i . 在这里,我们提供柜台i


Performance 性能

I decided to test out the performance of my method versus pst's method. 我决定测试我的方法与pst方法的性能。

Test Code: 测试代码:

static void Main(string[] args) {
    const int N = 1000000;
    const int M = 10;
    Stopwatch s;

    // Generate test list of strings.
    var list = Enumerable.Range(0, N).Select(n => n.ToString());

    // Just so it's enumerated once before.
    var ar = list.ToArray(); 

    // Try Jonathon's method M times.
    s = Stopwatch.StartNew();
    for (int x = 0; x < M; x++) {
        int i = 0;
        //var dict1 = list.ToDictionary(_ => i++);    // Before question edit
        var dict1 = list.ToDictionary(x => x, _ => i++);
    }
    s.Stop();
    Console.WriteLine("Jonathon's i++ method took {0} ms", s.ElapsedMilliseconds);

    // Try pst's method M times.
    s = Stopwatch.StartNew();
    for (int x = 0; x < M; x++) {
        var dict2 = list.Select((v, j) => new {v, j}).ToDictionary(p => p.v, p => p.j);
    }
    s.Stop();
    Console.WriteLine("psts' Select() method took {0} ms", s.ElapsedMilliseconds);

    Console.ReadLine();
}

Output: 输出:

Jonathon's i++ method took 3005 ms
pst's Select() method took 5019 ms

Output (After question edit): 输出(问题编辑后):

Jonathon's i++ method took 3521 ms
pst's Select() method took 5103 ms

In conclusion, it appears that there is a roughly 40% improvement by using a counter instead of creating the intermediate, anonymous-typed objects. 总之,通过使用计数器而不是创建中间的匿名类型的对象,似乎可以将大约40%的性能提高。

Assuming that each item maps to the key and and n represents the index in the source, then: 假设每个项目都映射到键,并且n代表源中的索引,则:

var dict = list
    .Select((v, i) => new {v, i})
    .ToDictionary(p => p.v, p => p.i);

I like this approach because, while the temporary anonymous type is a little wonky, it's a self-contained side-effect free expression. 我喜欢这种方法,因为,尽管临时匿名类型有些古怪,但它是一个自包含的副作用免费表达式。

(This will throw an exception if there are any duplicate items in the source list.) (如果源列表中有任何重复项,则将引发异常。)

声明:本站的技术帖子网页,遵循CC BY-SA 4.0协议,如果您需要转载,请注明本站网址或者原文地址。任何问题请咨询:yoyou2525@163.com.

 
粤ICP备18138465号  © 2020-2024 STACKOOM.COM