MapReduce TOP n

1.pom.xml

<dependency>
			<groupId>org.apache.hadoop</groupId>
			<artifactId>hadoop-hdfs</artifactId>
			<version>2.9.0</version>
		</dependency>
		
		<dependency>
			<groupId>org.apache.hadoop</groupId>
			<artifactId>hadoop-yarn-api</artifactId>
			<version>2.9.0</version>
		</dependency>

		<dependency>
			<groupId>org.apache.hadoop</groupId>
			<artifactId>hadoop-client</artifactId>
			<version>2.9.0</version>
		</dependency>

		<dependency>
			<groupId>org.apache.hadoop</groupId>
			<artifactId>hadoop-common</artifactId>
			<version>2.9.0</version>
		</dependency>

2. java文件

  1 package ex.topn;
  2 
  3 import java.util.Comparator;
  4 
  5 public class MySalaryComparator implements Comparator<Salary>  {
  6     @Override
  7     public int compare(Salary e1, Salary e2) {
  8         if(e1.getSum() > e2.getSum()) {
  9             return 1;
 10         }else{
 11             return -1;
 12         }
 13     }
 14 }
 15 
 16 
 17 package ex.topn;
 18 
 19 public class Salary {
 20     private int sum;
 21     
 22     public Salary(int sum) {
 23         this.sum = sum;
 24     }
 25     
 26     public int getSum() {
 27         return sum;
 28     }
 29     
 30     public void setSum(int sum) {
 31         this.sum = sum;
 32     }
 33 }
 34 
 35 
 36 
 37 
 38 package ex.topn;
 39 
 40 import java.io.IOException;
 41 import java.util.Iterator;
 42 import java.util.Map.Entry;
 43 import java.util.TreeMap;
 44 
 45 import org.apache.hadoop.io.LongWritable;
 46 import org.apache.hadoop.io.NullWritable;
 47 import org.apache.hadoop.io.Text;
 48 import org.apache.hadoop.mapreduce.Mapper;
 49 
 50 public class Top10Mapper extends Mapper<LongWritable, Text, NullWritable, Text> {
 51     public static TreeMap<Salary, Text> TopRecordMap = new TreeMap<Salary, Text>(new MySalaryComparator());
 52     
 53     public void map(LongWritable key, Text value, Context context) throws IOException, InterruptedException {
 54         String line = value.toString();
 55         String[] tokens = line.split(",");
 56         
 57         // get salary
 58         int salary = Integer.parseInt(tokens[3]);
 59         System.out.println("=> salary=" + salary);
 60         TopRecordMap.put(new Salary(salary), new Text(value));
 61         
 62         Iterator<Entry<Salary, Text>> iter = TopRecordMap.entrySet().iterator();
 63         Entry<Salary, Text> entry = null;
 64         
 65         System.out.println("TopRecordMap.size()=" + TopRecordMap.size());
 66         while(TopRecordMap.size() > 10) {
 67             entry = iter.next();
 68             iter.remove();
 69         }
 70     }    
 71     
 72     protected void cleanup(Context context) throws IOException, InterruptedException {
 73         // output our ten records to the reducers with a null key
 74     
 75         for(Text t : TopRecordMap.values()) {
 76             context.write(NullWritable.get(), t);
 77             System.out.println(" => " + t + ", " + t.toString());
 78         }
 79     }
 80 }
 81 
 82 
 83 
 84 
 85 package ex.topn;
 86 
 87 import java.io.IOException;
 88 import java.util.Iterator;
 89 import java.util.Map.Entry;
 90 import java.util.TreeMap;
 91 
 92 import org.apache.hadoop.io.NullWritable;
 93 import org.apache.hadoop.io.Text;
 94 import org.apache.hadoop.mapreduce.Reducer;
 95 
 96 public class Top10Reducer extends Reducer<NullWritable, Text, NullWritable, Text> {
 97     public static TreeMap<Salary, Text> TopRecordMap = new TreeMap<Salary, Text>(new MySalaryComparator());
 98     
 99     public void reduce(NullWritable key, Iterable<Text> values, Context context) throws IOException, InterruptedException {
100         
101         for(Text value : values) {
102             String line = value.toString();
103             if(line.length() > 0) {
104                 String[] tokens = line.split(",");
105                 int salary = Integer.parseInt(tokens[3]);
106                 System.out.println("=> salary=" + salary);
107                 TopRecordMap.put(new Salary(salary), new Text(value));
108                 
109                 Iterator<Entry<Salary, Text>> iter = TopRecordMap.entrySet().iterator();
110                 Entry<Salary, Text> entry = null;
111                 
112                 System.out.println("TopRecordMap.size()=" + TopRecordMap.size());
113                 while(TopRecordMap.size() > 10) {
114                     entry = iter.next();
115                     iter.remove();
116                 }        
117             }
118         }
119         
120 //        Iterator<Entry<Salary, Text>> iter = TopRecordMap.entrySet().iterator();
121 //        while(iter.hasNext()) {
122 //            System.out.println("    => " + iter.next().getValue());
123 //        }
124         
125         for(Text t : TopRecordMap.descendingMap().values()) {
126             context.write(NullWritable.get(), t);
127         }
128     }
129 }
130 
131 
132 
133 package ex.topn;
134 
135 import java.io.IOException;
136 import java.net.URI;
137 
138 import org.apache.hadoop.conf.Configuration;
139 import org.apache.hadoop.fs.FileSystem;
140 import org.apache.hadoop.fs.Path;
141 import org.apache.hadoop.io.NullWritable;
142 import org.apache.hadoop.io.Text;
143 import org.apache.hadoop.mapreduce.Job;
144 import org.apache.hadoop.mapreduce.lib.input.FileInputFormat;
145 import org.apache.hadoop.mapreduce.lib.output.FileOutputFormat;
146 
147 public class Topn {
148     
149     public static void main(String[] args) throws Exception {
150         Configuration conf = new Configuration();
151         
152         
153         Job job = Job.getInstance(conf,"top n");
154         job.setJarByClass(Topn.class);
155         job.setMapperClass(Top10Mapper.class);
156         job.setCombinerClass(Top10Reducer.class);
157         job.setReducerClass(Top10Reducer.class);
158         job.setNumReduceTasks(1);
159         
160         //job.setMapOutputKeyClass(NullWritable.class);
161         //job.setMapOutputValueClass(IntWritable.class);
162         
163         job.setOutputKeyClass(NullWritable.class);
164         job.setOutputValueClass(Text.class);
165         
166         FileInputFormat.addInputPath(job, new Path(args[0]));
167         FileOutputFormat.setOutputPath(job, new Path(args[1]));
168         System.exit(job.waitForCompletion(true) ? 0 : 1);
169     }170 }

3. 从eclipse export一个jar: topn.jar, 上传到 hadoop: /opt/hadoop-2.9.0/files

4. 准备三个文件，放入hdfs: /user/root/topn/input, 每行四个数字，原来的例子第四个字段是代表工资，在这些数据中取前10个工资最高的行，可以有重复。

这里用比较简单的数字代字代替，前三个字段根本用不到，所以大部分是1,2,3这种。

file03:
1985,1,2,11
1985,1,2,12
1985,1,2,13
1985,1,2,8
1985,1,2,7
1985,1,2,16
1985,1,2,7
1985,1,2,20
195,1,2,25
1,2,3,22
1,2,3,27
12,2,3,18
1,2,3,17
1,2,3,15


file04:
1,2,3,2
12,3,3,3
1,2,3,4
12,3,4,5
1,2,3,36
1,2,3,34
12,2,3,27
1,2,3,48
1,2,3,35
12,3,3,28
1,2,3,19
12,2,2,31
1,2,3,29


file05:
1,2,3,55
1,2,3,39
1,2,3,2
a,a,a,35

5.执行：

例子以root用户运行    
# cd /opt/hadoop-2.9.0
# bin/yarn jar files/topn.jar ex.topn.Topn topn/input topn/output

6.运行结果：

1,2,3,55
1,2,3,48
1,2,3,39
1,2,3,36
a,a,a,35
1,2,3,35
1,2,3,34
12,2,2,31
1,2,3,29
12,3,3,28

相关阅读:
concate string when group by
Regular Expression
Login failed for user
SQL Performance Tools
Web References
ARTetris-AR版俄罗斯方块的源码解析
 ARKit_1
链表
 线性表
 ORB-SLAM2的特征提取算法
原文地址：https://www.cnblogs.com/bear129/p/9415567.html