NO.5   算法测试（词条统计）

 一、安装Eclipse

下载Eclipse，解压安装，例如安装到/usr/local，即/usr/local/eclipse

4.3.1版本下载地址：http://pan.baidu.com/s/1eQkpRgu

二、在eclipse上安装hadoop插件

 

   1、下载hadoop插件(就是一个jar包文件：hadoop-eclipse-plugin-1.2.1.jar)

 

   2、把插件放到eclipse/plugins目录下

 

   3、重启eclipse，按如下步骤配置Hadoop installation directory。 

如果插件安装成功，打开Windows—Preferences后，在窗口左侧会有Hadoop Map/Reduce选项，点击此选项，在窗口右侧设置Hadoop安装路径。

4、配置Map/Reduce Locations

打开Windows—Open Perspective—Other

搜索“Map”

    选择Map/Reduce，点击OK

在右下方看到如下图所示

点击Map/Reduce Location选项卡，点击右边小象图标，打开Hadoop Location配置窗口：

输入Location Name，任意名称即可.配置Map/Reduce Master和DFS Mastrer，Host和Port配置成与core-site.xml的设置一致即可。

点击"Finish"按钮，关闭窗口。

点击左侧的DFSLocations—>myhadoop（上一步配置的location name)，如能看到user，表示安装成功

如果如下图所示表示安装失败，请检查Hadoop是否启动，以及eclipse配置是否正确。使用eclipse版本与jdk的版本对应，可以多安装几个jdk,灵活切换调用。

三、新建WordCount项目

File—>Project，选择Map/Reduce Project，输入项目名称WordCount等。

在WordCount项目里新建class，名称为WordCount，代码如下：

 

import java.io.IOException;

 

import java.util.StringTokenizer;

 

import org.apache.hadoop.conf.Configuration;

 

import org.apache.hadoop.fs.Path;

 

import org.apache.hadoop.io.IntWritable;

 

import org.apache.hadoop.io.Text;

 

import org.apache.hadoop.mapreduce.Job;

 

import org.apache.hadoop.mapreduce.Mapper;

 

import org.apache.hadoop.mapreduce.Reducer;

 

import org.apache.hadoop.mapreduce.lib.input.FileInputFormat;

 

import org.apache.hadoop.mapreduce.lib.output.FileOutputFormat;

 

import org.apache.hadoop.util.GenericOptionsParser;

 

public class WordCount {

                 public static class TokenizerMapper extends Mapper<Object, Text, Text, IntWritable>{                 

                                 private final static IntWritable one = new IntWritable(1);                               

                                 private Text word = new Text();                   

                                 public void map(Object key, Text value, Context context) throws IOException, InterruptedException {                               

                                                                StringTokenizer itr = new StringTokenizer(value.toString());                                                               

                                                                 while (itr.hasMoreTokens()) {

                                                                

                                                                                 word.set(itr.nextToken());

                                                                                

                                                                                context.write( word,one );

                                                                }

                                }

                }

 

                 public static class IntSumReducer extends Reducer<Text,IntWritable,Text,IntWritable> {

                

                                 private IntWritable result = new IntWritable();

                                

                                 public void reduce(Text key, Iterable<IntWritable> values,Context context) throws IOException, InterruptedException {                        

                                                 int sum = 0;

                                                

                                                 for (IntWritable val : values) {

                                                

                                                                sum += val.get();                                              

                                                }

                                                

                                                 result.set(sum);

                                                

                                                context.write(key, result);     

                                }

                }

 

                 public static void main(String[] args) throws Exception {

                                Configuration conf = new Configuration();

                                

                                String[] otherArgs = new GenericOptionsParser(conf, args).getRemainingArgs();

                                

                                 if (otherArgs.length != 2) {                      

                                                System. err.println("Usage: wordcount <in> <out>" );                                               

                                                System. exit(2);

                                }

                                

                                Job job = new Job(conf, "word count" );

                                

                                job.setJarByClass(WordCount. class);

                                

                                job.setMapperClass(TokenizerMapper. class);

                                

                                job.setCombinerClass(IntSumReducer. class);

                                

                                job.setReducerClass(IntSumReducer. class);

                                

                                job.setOutputKeyClass(Text. class);

                                

                                job.setOutputValueClass(IntWritable. class);

                                

                                FileInputFormat. addInputPath(job, new Path(otherArgs[0]));

                                

                                FileOutputFormat. setOutputPath(job, new Path(otherArgs[1]));

                                

                                System. exit(job.waitForCompletion(true) ? 0 : 1);

                }

}

 

 

四、运行

    1、在HDFS上创建目录input

        hadoop fs -mkdir input

或者在Eclipse中的使用快捷功能

2、拷贝本地Test1.txt 到HDFS的input里

         hadoop fs -copyFromLocal /usr/root/Test1.txt input

3、点击WordCount.java，右键，点击Run As—>Run Configurations，配置运行参数，即输入和输出文件夹

hdfs://localhost:9000/user/root/input         hdfs://localhost:9000/user/root/output

点击Run按钮，运行程序。

4、运行完成后，查看运行结果

        方法1：

        hadoop fs -ls output

        可以看到有两个输出结果，_SUCCESS和part-r-00000

        执行hadoop fs -cat output/*

 

方法2:

展开DFS Locations，如下图所示，双击打开part-r00000查看结果

////////////////////////////////////////////////////////////////////////////

小结：Hadoop程序处理流程

（1）将文件拆分为splits，并由MapReduce框架自动完成分割，将每一个split分割为<key,value>对

（2）每一对<key,value>调用一次map函数，处理后生产新的<key,value>对，由Context传递给reduce处理

（3）Mapper对<key，value>对进行按key值进行排序，并执行Combine过程，将key值相同的value进行合并。最后得到Mapper的最终输出结果

     （4）reduce处理，处理后将新的<key,value>对输出。