Hadoop-MapReduce2的几个基本示例

去重处理(Distinct)

类似于db中的select distinct(x) from table , 去重处理甚至比WordCount还要简单,假如我们要对以下文件的内容做去重处理(注:该文件也是后面几个示例的输入参数)

28832353027

基本上啥也不用做,在map阶段,把每一行的值当成key分发下去,然后在reduce阶段回收上来就可以了.

注:里面用到了一个自己写的类HDFSUtil,可以在 hadoop: hdfs API示例 一文中找到.

原理:map阶段完成后,在reduce开始之前,会有一个combine的过程,相同的key值会自动合并,所以自然而然的就去掉了重复.

 1 package yjmyzz.mr; 2  3 import org.apache.hadoop.conf.Configuration; 4 import org.apache.hadoop.fs.Path; 5 import org.apache.hadoop.io.NullWritable; 6 import org.apache.hadoop.io.Text; 7 import org.apache.hadoop.mapreduce.Job; 8 import org.apache.hadoop.mapreduce.Mapper; 9 import org.apache.hadoop.mapreduce.Reducer;10 import org.apache.hadoop.mapreduce.lib.input.FileInputFormat;11 import org.apache.hadoop.mapreduce.lib.output.FileOutputFormat;12 import org.apache.hadoop.util.GenericOptionsParser;13 14 import yjmyzz.util.HDFSUtil;15 16 import java.io.IOException;17 18 19 public class RemoveDup {20 21     public static class RemoveDupMapper22             extends Mapper<Object, Text, Text, NullWritable> {23 24         public void map(Object key, Text value, Context context)25                 throws IOException, InterruptedException {26             context.write(value, NullWritable.get());27             //System.out.println("map: key=" + key + ",value=" + value);28         }29 30     }31 32     public static class RemoveDupReducer extends Reducer<Text, NullWritable, Text, NullWritable> {33         public void reduce(Text key, Iterable<NullWritable> values, Context context)34                 throws IOException, InterruptedException {35             context.write(key, NullWritable.get());36             //System.out.println("reduce: key=" + key);37         }38     }39 40     public static void main(String[] args) throws Exception {41         Configuration conf = new Configuration();42         String[] otherArgs = new GenericOptionsParser(conf, args).getRemainingArgs();43         if (otherArgs.length < 2) {44             System.err.println("Usage: RemoveDup <in> [<in>...] <out>");45             System.exit(2);46         }47 48         //删除输出目录(可选,省得多次运行时,总是报OUTPUT目录已存在)49         HDFSUtil.deleteFile(conf, otherArgs[otherArgs.length - 1]);50 51         Job job = Job.getInstance(conf, "RemoveDup");52         job.setJarByClass(RemoveDup.class);53         job.setMapperClass(RemoveDupMapper.class);54         job.setCombinerClass(RemoveDupReducer.class);55         job.setReducerClass(RemoveDupReducer.class);56         job.setOutputKeyClass(Text.class);57         job.setOutputValueClass(NullWritable.class);58 59 60         for (int i = 0; i < otherArgs.length - 1; ++i) {61             FileInputFormat.addInputPath(job, new Path(otherArgs[i]));62         }63         FileOutputFormat.setOutputPath(job,64                 new Path(otherArgs[otherArgs.length - 1]));65         System.exit(job.waitForCompletion(true) ? 0 : 1);66     }67 68 69 }

输出:

023578

 

3) 记录计数(Count)

这个跟WordCount略有不同,类似于Select Count(*) from tables的效果,代码也超级简单,直接拿WordCount改一改就行了

 1 package yjmyzz.mr; 2  3 import org.apache.hadoop.conf.Configuration; 4 import org.apache.hadoop.fs.Path; 5 import org.apache.hadoop.io.IntWritable; 6 import org.apache.hadoop.io.Text; 7 import org.apache.hadoop.mapreduce.Job; 8 import org.apache.hadoop.mapreduce.Mapper; 9 import org.apache.hadoop.mapreduce.Reducer;10 import org.apache.hadoop.mapreduce.lib.input.FileInputFormat;11 import org.apache.hadoop.mapreduce.lib.output.FileOutputFormat;12 import org.apache.hadoop.util.GenericOptionsParser;13 import yjmyzz.util.HDFSUtil;14 15 import java.io.IOException;16 import java.util.StringTokenizer;17 18 19 public class RowCount {20 21     public static class RowCountMapper22             extends Mapper<Object, Text, Text, IntWritable> {23 24         private final static IntWritable one = new IntWritable(1);25         private final  static Text countKey = new Text("count");26 27         public void map(Object key, Text value, Context context) throws IOException, InterruptedException {28                 context.write(countKey, one);29         }30     }31 32     public static class RowCountReducer extends Reducer<Text, IntWritable, Text, IntWritable> {33         private IntWritable result = new IntWritable();34 35         public void reduce(Text key, Iterable<IntWritable> values, Context context) throws IOException, InterruptedException {36             int sum = 0;37             for (IntWritable val : values) {38                 sum += val.get();39             }40             result.set(sum);41             context.write(key, result);42         }43     }44 45     public static void main(String[] args) throws Exception {46         Configuration conf = new Configuration();47         String[] otherArgs = new GenericOptionsParser(conf, args).getRemainingArgs();48         if (otherArgs.length < 2) {49             System.err.println("Usage: RowCount <in> [<in>...] <out>");50             System.exit(2);51         }52         //删除输出目录(可选)53         HDFSUtil.deleteFile(conf, otherArgs[otherArgs.length - 1]);54 55         Job job = Job.getInstance(conf, "word count");56         job.setJarByClass(RowCount.class);57         job.setMapperClass(RowCountMapper.class);58         job.setCombinerClass(RowCountReducer.class);59         job.setReducerClass(RowCountReducer.class);60         job.setOutputKeyClass(Text.class);61         job.setOutputValueClass(IntWritable.class);62         for (int i = 0; i < otherArgs.length - 1; ++i) {63             FileInputFormat.addInputPath(job, new Path(otherArgs[i]));64         }65         FileOutputFormat.setOutputPath(job,66                 new Path(otherArgs[otherArgs.length - 1]));67         System.exit(job.waitForCompletion(true) ? 0 : 1);68     }69 70 71 }

输出: count 11

注:如果只想输出一个数字,不需要"count"这个key,可以改进一下:

 1 package yjmyzz.mr; 2  3 import org.apache.hadoop.conf.Configuration; 4 import org.apache.hadoop.fs.Path; 5 import org.apache.hadoop.io.LongWritable; 6 import org.apache.hadoop.io.NullWritable; 7 import org.apache.hadoop.io.Text; 8 import org.apache.hadoop.mapreduce.Job; 9 import org.apache.hadoop.mapreduce.Mapper;10 import org.apache.hadoop.mapreduce.Reducer;11 import org.apache.hadoop.mapreduce.lib.input.FileInputFormat;12 import org.apache.hadoop.mapreduce.lib.output.FileOutputFormat;13 import org.apache.hadoop.util.GenericOptionsParser;14 import yjmyzz.util.HDFSUtil;15 16 import java.io.IOException;17 18 19 public class RowCount2 {20 21     public static class RowCount2Mapper22             extends Mapper<LongWritable, Text, LongWritable, NullWritable> {23 24         public long count = 0;25 26         public void map(LongWritable key, Text value, Context context)27                 throws IOException, InterruptedException {28             count += 1;29         }30 31         protected void cleanup(Context context) throws IOException, InterruptedException {32             context.write(new LongWritable(count), NullWritable.get());33         }34 35     }36 37     public static class RowCount2Reducer extends Reducer<LongWritable, NullWritable, LongWritable, NullWritable> {38 39         public long count = 0;40 41         public void reduce(LongWritable key, Iterable<NullWritable> values, Context context)42                 throws IOException, InterruptedException {43             count += key.get();44         }45 46 47         protected void cleanup(Context context) throws IOException, InterruptedException {48             context.write(new LongWritable(count), NullWritable.get());49         }50 51     }52 53     public static void main(String[] args) throws Exception {54         Configuration conf = new Configuration();55         String[] otherArgs = new GenericOptionsParser(conf, args).getRemainingArgs();56         if (otherArgs.length < 2) {57             System.err.println("Usage: FindMax <in> [<in>...] <out>");58             System.exit(2);59         }60 61         //删除输出目录(可选,省得多次运行时,总是报OUTPUT目录已存在)62         HDFSUtil.deleteFile(conf, otherArgs[otherArgs.length - 1]);63 64         Job job = Job.getInstance(conf, "RowCount2");65         job.setJarByClass(RowCount2.class);66         job.setMapperClass(RowCount2Mapper.class);67         job.setCombinerClass(RowCount2Reducer.class);68         job.setReducerClass(RowCount2Reducer.class);69         job.setOutputKeyClass(LongWritable.class);70         job.setOutputValueClass(NullWritable.class);71 72         for (int i = 0; i < otherArgs.length - 1; ++i) {73             FileInputFormat.addInputPath(job, new Path(otherArgs[i]));74         }75         FileOutputFormat.setOutputPath(job,76                 new Path(otherArgs[otherArgs.length - 1]));77         System.exit(job.waitForCompletion(true) ? 0 : 1);78     }79 80 81 }

这样输出结果就只有一个数字11了.

注意: 这里context.write(xxx)只能写在cleanup方法中, 该方法在Mapper和Reducer接口中都有, 在map方法及reduce方法执行完后,会触发cleanup方法. 大家可以尝试下,把context.write(xxx)写在map和reduce方法中试试看,结果会出现多行记录,而不是预期的仅1个数字.

 

4)求最大值(Max)

 1 package yjmyzz.mr; 2  3 import org.apache.hadoop.conf.Configuration; 4 import org.apache.hadoop.fs.Path; 5 import org.apache.hadoop.io.LongWritable; 6 import org.apache.hadoop.io.NullWritable; 7 import org.apache.hadoop.io.Text; 8 import org.apache.hadoop.mapreduce.Job; 9 import org.apache.hadoop.mapreduce.Mapper;10 import org.apache.hadoop.mapreduce.Reducer;11 import org.apache.hadoop.mapreduce.lib.input.FileInputFormat;12 import org.apache.hadoop.mapreduce.lib.output.FileOutputFormat;13 import org.apache.hadoop.util.GenericOptionsParser;14 import yjmyzz.util.HDFSUtil;15 16 import java.io.IOException;17 18 19 public class Max {20 21     public static class MaxMapper22             extends Mapper<LongWritable, Text, LongWritable, NullWritable> {23 24         public long max = Long.MIN_VALUE;25 26         public void map(LongWritable key, Text value, Context context) throws IOException, InterruptedException {27             max = Math.max(Long.parseLong(value.toString()), max);28         }29 30         protected void cleanup(Mapper.Context context) throws IOException, InterruptedException {31             context.write(new LongWritable(max), NullWritable.get());32         }33 34     }35 36     public static class MaxReducer extends Reducer<LongWritable, NullWritable, LongWritable, NullWritable> {37 38         public long max = Long.MIN_VALUE;39 40         public void reduce(LongWritable key, Iterable<NullWritable> values, Context context) throws IOException, InterruptedException {41 42             max = Math.max(max, key.get());43 44         }45 46 47         protected void cleanup(Reducer.Context context) throws IOException, InterruptedException {48             context.write(new LongWritable(max), NullWritable.get());49         }50 51     }52 53     public static void main(String[] args) throws Exception {54         Configuration conf = new Configuration();55         String[] otherArgs = new GenericOptionsParser(conf, args).getRemainingArgs();56         if (otherArgs.length < 2) {57             System.err.println("Usage: Max <in> [<in>...] <out>");58             System.exit(2);59         }60 61         //删除输出目录(可选,省得多次运行时,总是报OUTPUT目录已存在)62         HDFSUtil.deleteFile(conf, otherArgs[otherArgs.length - 1]);63 64         Job job = Job.getInstance(conf, "Max");65         job.setJarByClass(Max.class);66         job.setMapperClass(MaxMapper.class);67         job.setCombinerClass(MaxReducer.class);68         job.setReducerClass(MaxReducer.class);69         job.setOutputKeyClass(LongWritable.class);70         job.setOutputValueClass(NullWritable.class);71 72         for (int i = 0; i < otherArgs.length - 1; ++i) {73             FileInputFormat.addInputPath(job, new Path(otherArgs[i]));74         }75         FileOutputFormat.setOutputPath(job,76                 new Path(otherArgs[otherArgs.length - 1]));77         System.exit(job.waitForCompletion(true) ? 0 : 1);78     }79 80 81 }

输出结果:8

如果看懂了刚才的Count2版本的代码,这个自然不用多解释.

 

5)求和(Sum)

 View Code

输出结果:43

Sum与刚才的Max原理如出一辙,不多解释了,依旧利用了cleanup方法

 

6)求平均值(Avg)

 View Code

输出:3.909090909090909

这个稍微要复杂一点,平均值大家都知道=Sum/Count,所以这其实前面Count与Max的综合运用而已,思路是在输出的key-value中,用max做key,用count做value,最终形成{sum,count}的输出,然后在最后的cleanup中,sum/count即得avg,但是有一个特点要注意的地方,由于Mapper与Reducer的output {key,value}类型并不一致,所以96-101行这里,分别设置了Map及Reduce的key,value输出类型,如果没有96-97这二行,100-101这二行会默认把Mapper,Combiner,Reducer这三者的输出类型设置成相同的类型.

 

7) 改进型的WordCount(按词频倒排)

官网示例WordCount只统计出单词出现的次数,并未按词频做倒排,下面的代码示例实现了该功能

 1 package yjmyzz.mr; 2  3 import org.apache.hadoop.conf.Configuration; 4 import org.apache.hadoop.fs.Path; 5 import org.apache.hadoop.io.IntWritable; 6 import org.apache.hadoop.io.LongWritable; 7 import org.apache.hadoop.io.NullWritable; 8 import org.apache.hadoop.io.Text; 9 import org.apache.hadoop.mapreduce.Job;10 import org.apache.hadoop.mapreduce.Mapper;11 import org.apache.hadoop.mapreduce.Reducer;12 import org.apache.hadoop.mapreduce.lib.input.FileInputFormat;13 import org.apache.hadoop.mapreduce.lib.output.FileOutputFormat;14 import org.apache.hadoop.util.GenericOptionsParser;15 import yjmyzz.util.HDFSUtil;16 17 import java.io.IOException;18 import java.util.Comparator;19 import java.util.StringTokenizer;20 import java.util.TreeMap;21 22 23 public class WordCount2 {24 25     public static class TokenizerMapper26             extends Mapper<Object, Text, Text, IntWritable> {27 28         private final static IntWritable one = new IntWritable(1);29         private Text word = new Text();30 31         public void map(Object key, Text value, Context context) throws IOException, InterruptedException {32             StringTokenizer itr = new StringTokenizer(value.toString());33             while (itr.hasMoreTokens()) {34                 word.set(itr.nextToken());35                 context.write(word, one);36             }37         }38     }39 40     public static class IntSumReducer extends Reducer<Text, IntWritable, Text, IntWritable> {41 42         //定义treeMap来保持统计结果,由于treeMap是按key升序排列的,这里要人为指定Comparator以实现倒排43         private TreeMap<Integer, String> treeMap = new TreeMap<Integer, String>(new Comparator<Integer>() {44             @Override45             public int compare(Integer x, Integer y) {46                 return y.compareTo(x);47             }48         });49 50         public void reduce(Text key, Iterable<IntWritable> values, Context context) throws IOException, InterruptedException {51             //reduce后的结果放入treeMap,而不是向context中记入结果52             int sum = 0;53             for (IntWritable val : values) {54                 sum += val.get();55             }56             if (treeMap.containsKey(sum)){57                 String value = treeMap.get(sum) + "," + key.toString();58                 treeMap.put(sum,value);59             }60             else {61                 treeMap.put(sum, key.toString());62             }63         }64 65         protected void cleanup(Context context) throws IOException, InterruptedException {66             //将treeMap中的结果,按value-key顺序写入contex中67             for (Integer key : treeMap.keySet()) {68                 context.write(new Text(treeMap.get(key)), new IntWritable(key));69             }70         }71     }72 73     public static void main(String[] args) throws Exception {74         Configuration conf = new Configuration();75         String[] otherArgs = new GenericOptionsParser(conf, args).getRemainingArgs();76         if (otherArgs.length < 2) {77             System.err.println("Usage: wordcount2 <in> [<in>...] <out>");78             System.exit(2);79         }80         //删除输出目录81         HDFSUtil.deleteFile(conf, otherArgs[otherArgs.length - 1]);82         Job job = Job.getInstance(conf, "word count2");83         job.setJarByClass(WordCount2.class);84         job.setMapperClass(TokenizerMapper.class);85         job.setCombinerClass(IntSumReducer.class);86         job.setReducerClass(IntSumReducer.class);87         job.setOutputKeyClass(Text.class);88         job.setOutputValueClass(IntWritable.class);89         for (int i = 0; i < otherArgs.length - 1; ++i) {90             FileInputFormat.addInputPath(job, new Path(otherArgs[i]));91         }92         FileOutputFormat.setOutputPath(job,93                 new Path(otherArgs[otherArgs.length - 1]));94         System.exit(job.waitForCompletion(true) ? 0 : 1);95     }96 97 98 }

原理: 依然用到了cleanup,此外为了实现排序,采用了TreeMap这种内置了key排序的数据结构.

这里为了展示更直观,选用了电影<超能陆战队>主题曲的第一段歌词做为输入:

They say we are what we areBut we do not have to beI am  bad behavior but I do it in the best wayI will be the watcherOf the eternal flameI will be the guard dogof all your fever dreams

原版的WordCount处理完后,结果如下:

But1I4Of1They1all1am1are2bad1be3behavior1best1but1do2dog1dreams1eternal1fever1flame1guard1have1in1it1not1of1say1the4to1watcher1way1we3what1will2your1

改进后的WordCount2处理结果如下:

I,the4be,we3are,do,will2But,Of,They,all,am,bad,behavior,best,but,dog,dreams,eternal,fever,flame,guard,have,in,it,not,of,say,to,watcher,way,what,your1