how to write int values in hadoop reducer

Question

I am writing a mapreduce with Hadoop. In reduce method, I want to use context.write() . But output is int type. How can I do this? When I use context.write() it shows an error:

The second argument cannot be int.

This is my code:

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

            int count = 0;

            for (NullWritable nullWritable : values) {
                count++;
            }

             //context.write(key, count); 
}

This reduce counts something. Then it should write key and count variable.

How can I do that?

Answer

I found my answer. I should new a IntWritable class and use its method (set(intValue)).

Like following code:

IntWritable c = new IntWritable();
    public void reduce(Text key, Iterable<NullWritable> values, Context context) throws IOException, InterruptedException {

                int count = 0;


                for (NullWritable nullWritable : values) {
                    count++;
                }

                c.set(count);
                context.write(key, c);

            }

Jijo Jijo · Accepted Answer · 2014-01-16T18:02:45

You can also use it like this:

context.write(key,new IntWritable(count));

how to write int values in hadoop reducer

1 Answers