java—mapreduce中第一个reduce任务的输出值为< key value>对

我希望我的第一个reduce任务生成smth like（当然，<sum，count>）；在第二个reduce任务中，我将计算每个课程的sum/count。第一个减速机任务充当合并器、求和器和计数器；第二个reduce任务找到每个课程的平均值并输出平均值。我只是找不到什么是最好的类型存储输出值作为密钥对，然后能够检索和计算他们。hashmap不起作用。

import java.io.IOException;
import java.util.HashMap;
import java.util.Map;
import java.util.Map.Entry;
import java.util.StringTokenizer;

import org.apache.hadoop.conf.Configuration;
import org.apache.hadoop.fs.Path;
import org.apache.hadoop.io.FloatWritable;
import org.apache.hadoop.io.IntWritable;
import org.apache.hadoop.io.LongWritable;
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.Mapper.Context;
import org.apache.hadoop.mapreduce.lib.input.FileInputFormat;
import org.apache.hadoop.mapreduce.lib.output.FileOutputFormat;
import org.apache.hadoop.mapreduce.lib.output.MultipleOutputs;

public class AvgGrading {

    public static void main(String[] args) throws IllegalArgumentException, IOException, ClassNotFoundException, InterruptedException {
        Configuration conf = new Configuration();
        Job job = Job.getInstance(conf, "avg grading");
        job.setJarByClass(AvgGrading.class);
        job.setMapperClass(MapForAverage.class);
        job.setCombinerClass(ReduceForAverage.class);
        job.setNumReduceTasks(2);
        job.setReducerClass(ReduceForFinal.class);
        job.setMapOutputKeyClass(LongWritable.class);
        job.setMapOutputValueClass(Object.class);
        job.setOutputKeyClass(LongWritable.class);
        job.setOutputValueClass(FloatWritable.class);
        FileInputFormat.addInputPath(job, new Path(args[0]));
        FileOutputFormat.setOutputPath(job, new Path(args[1]));
        System.exit(job.waitForCompletion(true) ? 0 : 1);
    }

    public static class MapForAverage extends Mapper<LongWritable, Text, LongWritable, Object> {

        public void map(LongWritable key, Text value, Context con) throws IOException, InterruptedException {
            String [] word = value.toString().split(", ");
            float grade = Integer.parseInt(word[1]);
            int course = Integer.parseInt(word[0]);
            Map <Float,Long> m = new HashMap<Float,Long>();
            m.put(grade, (long) 1);
            con.write(new LongWritable(course), m);
        }
    }

    public static class ReduceForAverage extends Reducer<LongWritable, Object, LongWritable, Object> {
        private FloatWritable result = new FloatWritable();
        public void reduce(LongWritable course, Map<Float,Long> values, Context con)
                throws IOException, InterruptedException {

            Map <Float,Long> m = new HashMap<Float,Long>();

            float sum = 0;
            long count =0;
            for (Map.Entry<Float, Long> entry : values.entrySet()) {
                sum += entry.getKey();
                count++;
            }
            m.put(sum, count);

            con.write(course, m);
        }
    }

    public static class ReduceForFinal extends Reducer<LongWritable, Object, LongWritable, FloatWritable> {
        private FloatWritable result = new FloatWritable();
        public void reduce(LongWritable course, Map<Long,Float>values, Context con)
                throws IOException, InterruptedException {

            long key = 0;
            float value=0;

            for ( Map.Entry<Long, Float> entry : values.entrySet()) {
                 key = entry.getKey();
                 value = entry.getValue();
            }
            float res= key/value;

            con.write(course, new FloatWritable(res));
        }
    }
}

请注意，我无法遍历 Iterable < Map<Float,Int>> 在reduce任务中，我传递的是简单的map，这可能是不正确的。
错误代码为：

Unable to initialize MapOutputCollector org.apache.hadoop.mapred.MapTask$MapOutputBuffer

java.lang.nullpointerexception异常
二级减速器故障

map不实现可写的，你说你的combiner和reducer输入值的类是object，而你发射map，你只需要为此创建一个自定义类。记住，如果您想在hadoop中发出一些东西，那么定制类必须实现可写的。你可以这样做：

public class Counter implements Writable {

       private float sum;
       private long count;

       public Counter(float sum, long count){
              this.sum = sum;
              this.count = count;
       }

       /* Methods to get and set private variables of the class */

       public float getSum() {
              return sum;
       }

       public void setSum(float sumValue) {
              sum=sumValue;
       }

       public long getCount() {
              return count;
       }

       public void setCount(long countValue) {
              count=countValue;
       }

       /* Methods to serialize and deserialize the contents of the
          instances of this class */

       @Override /* Serialize the fields of this object to out */ 
       public void write(DataOutput out) throws IOException{
              out.writeFloat(sum);
              out.writeLong(count);
       }

      @Override /* Deserialize the fields of this object from in */
      public void readFields(DataInputin) throws IOException{
                  sum=in.readFloat();
                  count=in.readLong();
       }
       }

因此，在您的第一个Map器中，您可以通过以下方式创建并发出计数器：

Counter counter = new Counter(grade, 1);
       con.write(course, counter);

此时，在第一个减速机中，您将拥有一个表示进程的键和一个iterable值，该值是所有计数器的iterable值，使用该iterable可以计算平均值。记住更新mapper和reducers类参数，使其与新参数保持一致。

java—mapreduce中第一个reduce任务的输出值为< key value>对

1条答案

相关问题

热门标签

最新问答