Parallel Linq的用法
并行集合
并行计算使用的多个线程同时进行计算，所以要控制每个线程对资源的访问，我们先来看一下平时常用的List<T>集合，在并行计算下的表现：
using System;
using System.Collections.Generic;
using System.Linq;
using System.Text;
using System.Threading.Tasks;
using System.Collections.Concurrent;
namespace ThreadPool
{
public class PEnumerable
{
public static void ListWithParallel()
{
List<int> list = new List<int>();
Parallel.For(0, 10000, item =>
{
list.Add(item);
});
Console.WriteLine("List‘s count is {0}",list.Count());
}
}
}
看到结果中显示的5861，而不是我们循环的是10000次啊！怎么结果不对呢？这是因为List<T>是非线程安全集合，意思就是说所有的线程都可以修改他的值。
下面我们来看下并行集合 —— 线程安全集合，在System.Collections.Concurrent命名空间中，首先来看一下ConcurrentBag<T>泛型集合，其用法和List<T>类似，先来写个方法测试一下：
public static void ConcurrentBagWithPallel()
{
ConcurrentBag<int> list = new ConcurrentBag<int>();
Parallel.For(0, 10000, item =>
{
list.Add(item);
});
Console.WriteLine("ConcurrentBag‘s count is {0}", list.Count());
}
可以看到，ConcurrentBag集合的结果是正确的，但ConcurrentBag中的数据并不是按照顺序排列的，顺序是乱的，随机的。我们平时使用的Max、First、Last等linq方法都还有。
关于线程安全的集合还有很多，和我们平时用的集合都差不多，比如类似Dictionary的ConcurrentDictionary，还有ConcurrentStack，ConcurrentQueue等。
AsParallel 方法
前面了解了并行的For和foreach,今天就来看一下Linq的并行版本是怎么样吧？为了测试，我们添加一个Custom类，代码如下：
public class Custom
{
public string Name { get; set; }
public int Age { get; set; }
public string Address { get; set; }
}
写如下测试代码：
public static void TestPLinq()
{
Stopwatch sw = new Stopwatch();
List<Custom> customs = new List<Custom>();
for (int i = 0; i < 2000000; i++)
{
customs.Add(new Custom() { Name = "Jack", Age = 21, Address = "NewYork" });
customs.Add(new Custom() { Name = "Jime", Age = 26, Address = "China" });
customs.Add(new Custom() { Name = "Tina", Age = 29, Address = "ShangHai" });
customs.Add(new Custom() { Name = "Luo", Age = 30, Address = "Beijing" });
customs.Add(new Custom() { Name = "Wang", Age = 60, Address = "Guangdong" });
customs.Add(new Custom() { Name = "Feng", Age = 25, Address = "YunNan" });
}
sw.Start();
var result = customs.Where<Custom>(c => c.Age > 26).ToList();
sw.Stop();
Console.WriteLine("Linq time is {0}.",sw.ElapsedMilliseconds);
sw.Restart();
sw.Start();
var result2 = customs.AsParallel().Where<Custom>(c => c.Age > 26).ToList();
sw.Stop();
Console.WriteLine("Parallel Linq time is {0}.", sw.ElapsedMilliseconds);
}
ToLookup方法
在项目中，我们经常要对数据做处理，比如分组统计，我们知道在linq中也可以实现，今天来学习一下新的ToLookup方法，写一个测试方法，代码如下：
{
Stopwatch stopWatch = new Stopwatch();
List<Custom> customs = new List<Custom>();
for (int i = 0; i < 2000000; i++)
{
customs.Add(new Custom() { Name = "Jack", Age = 21, Address = "NewYork" });
customs.Add(new Custom() { Name = "Jime", Age = 26, Address = "China" });
customs.Add(new Custom() { Name = "Tina", Age = 29, Address = "ShangHai" });
customs.Add(new Custom() { Name = "Luo", Age = 30, Address = "Beijing" });
customs.Add(new Custom() { Name = "Wang", Age = 60, Address = "Guangdong" });
customs.Add(new Custom() { Name = "Feng", Age = 25, Address = "YunNan" });
}
stopWatch.Restart();
var groupByAge = customs.GroupBy(item => item.Age).ToList();
foreach (var item in groupByAge)
{
Console.WriteLine("Age={0},count = {1}", item.Key, item.Count());
}
stopWatch.Stop();
Console.WriteLine("Linq group by time is: " + stopWatch.ElapsedMilliseconds);
stopWatch.Restart();
var lookupList = customs.ToLookup(i => i.Age);
foreach (var item in lookupList)
{
Console.WriteLine("LookUP:Age={0},count = {1}", item.Key, item.Count());
}
stopWatch.Stop();
Console.WriteLine("LookUp group by time is: " + stopWatch.ElapsedMilliseconds);
}
ToLookup方法是将集合转换成一个只读集合，所以在大数据量分组时性能优于List.