Start line:  
End line:  

Snippet Preview

Snippet HTML Code

Stack Overflow Questions
  * Copyright [2013-2014] eBay Software Foundation
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
  * You may obtain a copy of the License at
 * Unless required by applicable law or agreed to in writing, software
 * distributed under the License is distributed on an "AS IS" BASIS,
 * See the License for the specific language governing permissions and
 * limitations under the License.
package ml.shifu.guagua.mapreduce.example.sum;
import  org.apache.hadoop.conf.Configuration;
import  org.apache.hadoop.fs.FSDataOutputStream;
import  org.apache.hadoop.fs.FileSystem;
import  org.apache.hadoop.fs.Path;
SumOutput is used to write the final model output to file system.
public class SumOutput extends
        BasicMasterInterceptor<GuaguaWritableAdapter<LongWritable>, GuaguaWritableAdapter<LongWritable>> {
    private static final Logger LOG = LoggerFactory.getLogger(SumOutput.class);

Get output file setting and write final sum value to HDFS file.
    public void postApplication(
            final MasterContext<GuaguaWritableAdapter<LongWritable>, GuaguaWritableAdapter<LongWritable>> context) {
        .info("SumOutput starts to write final sum value to file.");
        Path out = new Path(context.getProps().getProperty("guagua.sum.output"));
        .info("Writing results to {}"out.toString());
        PrintWriter pw = null;
        try {
            FSDataOutputStream fos = FileSystem.get(new Configuration()).create(out);
            pw = new PrintWriter(fos);
        } catch (IOException e) {
            .error("Error in writing output."e);
        } finally {
New to GrepCode? Check out our FAQ X