• c#扩展出MapReduce方法


    MapReduce方法主体:

     1 public static IDictionary<TKey, TResult> MapReduce<TInput, TKey, TValue, TResult>(this IList<TInput> inputList,
     2             Func<MapReduceData<TInput>, KeyValueClass<TKey, TValue>> map, Func<TKey, IList<TValue>, TResult> reduce)
     3         {
     4             object locker = new object();
     5             ConcurrentDictionary<TKey, TResult> result = new ConcurrentDictionary<TKey, TResult>();
     6             //保存map出来的结果
     7             ConcurrentDictionary<TKey, IList<TValue>> mapDic = new ConcurrentDictionary<TKey, IList<TValue>>();
     8             var parallelOptions = new ParallelOptions();
     9             parallelOptions.MaxDegreeOfParallelism = Environment.ProcessorCount;
    10             //并行map
    11             Parallel.For(0, inputList.Count(), parallelOptions, t =>
    12             {
    13                 MapReduceData<TInput> data = new MapReduceData<TInput>
    14                 {
    15                     Data = inputList[t],
    16                     Index = t,
    17                     List = inputList,
    18                 };
    19                 var pair = map(data);
    20                 if (pair != null && pair.Valid)
    21                 {
    22                     //锁住防止并发操作list造成数据缺失
    23                     lock (locker)
    24                     {
    25                         //将匹配出来的结果加入结果集放入字典
    26                         IList<TValue> list = null;
    27                         if (mapDic.ContainsKey(pair.Key))
    28                         {
    29                             list = mapDic[pair.Key];
    30                         }
    31                         else
    32                         {
    33                             list = new List<TValue>();
    34                             mapDic[pair.Key] = list;
    35                         }
    36                         list.Add(pair.Value);
    37                     }
    38                 }
    39             });
    40 
    41             //并行reduce
    42             Parallel.For(0, mapDic.Keys.Count, parallelOptions, t =>
    43             {
    44                 KeyValuePair<TKey, IList<TValue>> pair = mapDic.ElementAt(t);
    45                 result[pair.Key] = reduce(pair.Key, pair.Value);
    46             });
    47             return result;
    48         }
    View Code

    KeyValueClass定义:

     1 public class KeyValueClass<K, V>
     2     {
     3         public KeyValueClass(K key, V value)
     4         {
     5             Key = key;
     6             Value = value;
     7         }
     8 
     9         public KeyValueClass()
    10         {
    11 
    12         }
    13 
    14         public K Key { get; set; }
    15 
    16         public V Value { get; set; }
    17     }
    View Code

    Console测试:

     1 List<TestClass> listTestClass = new List<TestClass>();
     2             listTestClass.Add(new TestClass { a = "a", g = 1 });
     3             listTestClass.Add(new TestClass { a = "b", g = 3 });
     4             listTestClass.Add(new TestClass { a = "c", g = 4 });
     5             listTestClass.Add(new TestClass { a = "d", g = 2 });
     6             listTestClass.Add(new TestClass { a = "e", g = 1 });
     7             listTestClass.Add(new TestClass { a = "f", g = 2 });
     8             listTestClass.Add(new TestClass { a = "g", g = 5 });
     9             listTestClass.Add(new TestClass { a = "h", g = 6 });
    10             IDictionary<int, string> dic = listTestClass.MapReduce(t =>
    11             {
    12                 if (t.g < 5)
    13                 {
    14                     return new KeyValueClass<int, string>(t.g, t.a);
    15                 }
    16                 return null;
    17             }, (key, values) =>
    18            {
    19                return string.Join(",", values);
    20            });
    View Code

    TestClass定义:

     1 public class TestClass
     2     {
     3         public string a { get; set; }
     4         public string b { get; set; }
     5 
     6         public string d { get; set; }
     7 
     8         //public DateTime f { get; set; }
     9 
    10         public int g { get; set; }
    11 
    12         public List<TestClass> test { get; set; }
    13 
    14         public Dictionary<string, string> dic { get; set; }
    15     }
    View Code

    结果:

    1:a,e

    2:d,f

    3:b

    4:c

    词频性能测试

  • 相关阅读:
    Cocos2d-x 使用物理引擎进行碰撞检测
    Cocos2d粒子系统二
    cocos2d(粒子效果编辑器)
    关于Cococs中的CCActionEase(下)
    关于Cococs中的CCActionEase(中)
    关于Cococs中的CCActionEase
    call()和apply() 的区别
    WebStorm 使用快捷键大全
    WebStorm 的使用(一)
    导入xlsx,文件到sqlite3数据库
  • 原文地址:https://www.cnblogs.com/WindBlog/p/4262617.html
Copyright © 2020-2023  润新知