0

I'm referring to this question Calling MapReduce Twice

The solution would be call MapReduce twice. The first MapReduce calculates the count of each word, and then in the second MapReduce, one computes the frequency of each count. My question is how to link these two MapReduce, i.e, how to pass the output of the 1st Reducer to the input of the 2nd Mapper (is this an automatic parser)?

Configuration conf = new Configuration();
Job job = Job.getInstance(conf, "Example");
job.setMapperClass(Mapper1.class);
job.setCombinerClass(Reducer1.class);
job.setReducerClass(Reducer1.class);
job.setMapperClass(Mapper2.class);
job.setCombinerClass(Reducer2.class);
job.setReducerClass(Reducer2.class);
Yunhan Zou
  • 369
  • 3
  • 14
  • Possible duplicate of [Chaining multiple MapReduce jobs in Hadoop](https://stackoverflow.com/questions/2499585/chaining-multiple-mapreduce-jobs-in-hadoop) – Ben Watson Mar 13 '19 at 09:03
  • I hope this helps - "chaining mapreduce jobs" returns a lot of results. – Ben Watson Mar 13 '19 at 09:04

0 Answers0