C# 并行编程 PLINQ

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：

       public static void OrderByTest()

      {
         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.

 

 

 

 

*****************************************************

*** No matter how far you go, looking back is also necessary. ***

*****************************************************

转载于:https://www.cnblogs.com/gangle/p/11578681.html