public static void main(String[] args) throws Exception {
Conf conf = new Conf(args, "-i input -o output");
conf.setMapMemoryMB(4096);
conf.setTaskTimeout(180000);
conf.setMapSpeculativeExecution(false);
conf.setMaxMapAttempts(1);
String input = conf.get("input");
Path out = new Path(conf.get("output"));
Job job = new Job(conf, input, out);
InputFormat.setNonSplitable(job);
setupInput(job, input);
job.setNumReduceTasks(1);
job.setMapperClass(VocabularyMap.class);
job.setMapOutputKeyClass(Text.class);
job.setMapOutputValueClass(LongWritable.class);
job.setReducerClass(VocabularyReduce.class);
job.setGroupingComparatorClass(Text.Comparator.class);
job.setOutputFormatClass(NullOutputFormat.class);
job.waitForCompletion(true);
}
/**
* @param args the command line arguments
* @throws java.io.IOException
*/
public static void main(String[] args) throws IOException, Exception {
Conf conf = new Conf();
Job job = new Job(conf, "TweetsLanguage");
Scan scan = new Scan();
TableMapReduceUtil.initTableMapperJob(
"hhscyber:tweets", scan, LanguageMapper.class, null, null, job);
job.setNumReduceTasks(0);
TableMapReduceUtil.initTableReducerJob("hhscyber:tweets_lang", null, job);
job.waitForCompletion(true);
}