Map/reduce java-创建未知输出的代码

代码运行良好，但输出不符合预期。我的代码是：

public class Test {

 public static class MapReduceMapper extends Mapper<LongWritable, Text, IntWritable, IntWritable> {

    public void map(LongWritable key, Text value, OutputCollector<IntWritable, IntWritable> output) throws IOException, InterruptedException {
        Scanner scanner = new Scanner(value.toString());
        String row;
        String[] pre;
        int[] tokens;
        while (scanner.hasNext()) {
            row = scanner.nextLine();
            pre = row.split("\\t");
            tokens = new int[pre.length];

            for(int i=0; i<pre.length;i++) {
                tokens[i] = Integer.parseInt(pre[i]);
            }

                output.collect(new IntWritable(tokens[0]), new IntWritable(tokens[1]));

        }
    }
 } 

 public static class MapReduceReducer extends Reducer<IntWritable, IntWritable, Text, NullWritable> {

 NullWritable NULL = NullWritable.get();

    public void reduce(IntWritable key, Iterable<IntWritable> values, OutputCollector<Text, NullWritable> output) 
      throws IOException, InterruptedException {

        for (IntWritable val : values) {
            int a = val.get();
                count++;
        }

        String keyValue = key.get() + ": ";
        output.collect(new Text(keyValue + "Mean = " + (sum / count)), NULL);
        output.collect(new Text(keyValue + "Count = " + count), NULL);


    }
 }

 public static void main(String[] args) throws Exception {
    Configuration conf = new Configuration();
    FileSystem fs = FileSystem.get(conf);

    Job job = new Job(conf, "mapreduce");

    job.setJarByClass(test.class);

    job.setOutputKeyClass(LongWritable.class);
    job.setOutputValueClass(Text.class);

    job.setMapperClass(MapReduceMapper.class);
    job.setReducerClass(MapReduceReducer.class);

    job.setInputFormatClass(TextInputFormat.class);
    job.setOutputFormatClass(TextOutputFormat.class);

    FileInputFormat.addInputPath(job, new Path("")); 
    String outputFile = "/home/kevmccar/mapreduce/output/";  
    Path outPath = new Path(outputFile);
    fs.delete(outPath, true);
    FileOutputFormat.setOutputPath(job, new Path(outputFile)); 

    job.waitForCompletion(true);
 }

}

我希望输出的格式为

1: Average = 12312
1: Count = 6564 

但它看起来像

5244    8       121602
5253    10      663603
5263    2       32288
5271    6       221095
5280    10      350834
5290    2       245710
5299    1       318947
5308    9       440945
5317    4       638909
...

---

为什么输出是这样？

我将

OutputCollector

更改为

Context

，只是因为我更熟悉

Context

，我不太确定

OutputCollector

是如何工作的。然而，这种改变似乎起到了作用，所以我假设使用OutputCollector声明键和值数据类型的方式有问题。以下是我的代码供参考：

public class HelloWorld {

   public static class MapReduceMapper extends Mapper<LongWritable, Text, IntWritable, IntWritable> {


     public void map(LongWritable key, Text value, Context context) throws IOException, InterruptedException {
        Scanner scanner = new Scanner(value.toString());
        String row;
        String[] pre;
        int[] tokens;
        while (scanner.hasNext()) {
          row = scanner.nextLine();
          pre = row.split("\\t");
          tokens = new int[pre.length];

          for(int i=0; i<pre.length;i++) {
            tokens[i] = Integer.parseInt(pre[i]);
          }
          System.err.println("MapKey: " + tokens[0] + "MapValue: " + tokens[1]);
          context.write(new IntWritable(tokens[0]), new IntWritable(tokens[1]));

        }
        scanner.close();
     }
  } 

 public static class MapReduceReducer extends Reducer<IntWritable, IntWritable, Text, NullWritable> {

   NullWritable NULL = NullWritable.get();

   public void reduce(IntWritable key, Iterable<IntWritable> values, Context context) throws IOException, InterruptedException {
    int sum = 0;
    int count = 0;
    int max = Integer.MIN_VALUE;
    int min = Integer.MAX_VALUE;
    for (IntWritable val : values) {
        int a = val.get();
            if (a > max) {
                max = a; }
            else if (a < min) {
                a = min;
                        }
            sum += a;
            count++;
    }

    String keyValue = "Product "  + key.get() + ": ";
    context.write(new Text(keyValue + "Mean = " + (sum / count)), NULL);
    context.write(new Text(keyValue + "Count = " + count), NULL);
    context.write(new Text(keyValue + "Min = " + min), NULL);
    context.write(new Text(keyValue + "Max = " + max), NULL);

   }
 }

 public static void main(String[] args) throws Exception {
   Configuration conf = new Configuration();
   FileSystem fs = FileSystem.get(conf);

   Job job = new Job(conf, "mapreduce");

   job.setJarByClass(HelloWorld.class);

   job.setMapOutputKeyClass(IntWritable.class);
   job.setMapOutputValueClass(IntWritable.class);
   job.setOutputKeyClass(Text.class);
   job.setOutputValueClass(NullWritable.class);

   job.setMapperClass(MapReduceMapper.class);
   job.setReducerClass(MapReduceReducer.class);

   job.setInputFormatClass(TextInputFormat.class);
   job.setOutputFormatClass(TextOutputFormat.class);

   FileInputFormat.addInputPath(job, new Path(args[0])); 
   String outputFile = args[1];  
   Path outPath = new Path(outputFile);
   fs.delete(outPath, true);
   FileOutputFormat.setOutputPath(job, new Path(outputFile)); 

   job.waitForCompletion(true);
 }

}

样本输出：

Product 5: Mean = 289182
Product 5: Count = 1
Product 5: Min = 2147483647
Product 5: Max = 289182

---

我认为Map方法中的

LongWritable键

并不是您所期望的那样。传递给map方法的键是文件中行（值）的位置，行是

1 1029109

，等等。您可能希望使用传递给映射器的

值，它包含文件中的键和值对。我将映射器中的.toString更改为我以前使用的line.toString。我用这两个函数运行，结果是相同的输出。你的
行定义在哪里？哇，对不起，我是说value.toString而不是line.toString，我更详细地检查了你的代码，但我无法确定错误在哪里。不过，我建议您更详细地了解这一点：
job.setOutputKeyClass（LongWritable.class）；job.setOutputValueClass（Text.class）
据我所知，当map和reduce发出相同类型的键和值时，应该使用
setOutputKeyClass（）和setOutputValueClass
。因为在你的例子中，你在Map&Reduce中发出不同类型的键和值，你最好使用像
setMapOutputKeyClass
和
setMapOutputValueClass
这样的方法，对于reducer也是如此。是的，这似乎起到了作用。谢谢你花了这么多时间。
1   1029109
5   289182
6   547849

Product 5: Mean = 289182
Product 5: Count = 1
Product 5: Min = 2147483647
Product 5: Max = 289182