public static void main(String[] args) throws Exception { if (args.length != 2) { System.err.println("Usage: OutLinks <input path> <output path>"); System.exit(-1); } Job job = new Job(); job.setJarByClass(OutLinks.class); FileInputFormat.addInputPath(job, new Path(args[0])); FileOutputFormat.setOutputPath(job, new Path(args[1])); job.setOutputFormatClass(MapFileOutputFormat.class); job.setMapperClass(OutLinkNumMapper.class); job.setReducerClass(OutLinkNumReducer.class); job.setMapOutputKeyClass(Text.class); job.setMapOutputValueClass(IntWritable.class); job.setOutputKeyClass(Text.class); job.setOutputValueClass(DoubleWritable.class); job.waitForCompletion(true); }
public static void main(String[] args) throws Exception { if(args.length != 2) { System.err.println("Usage: UrlModulus <input path> <output path>"); System.exit(-1); } Configuration conf = new Configuration(); Job job = new Job(conf, "UrlModulus"); job.setJarByClass(UrlModulus.class); FileInputFormat.addInputPath(job, new Path(args[0])); FileOutputFormat.setOutputPath(job, new Path(args[1])); job.setOutputFormatClass(MapFileOutputFormat.class); job.setMapperClass(UrlModulusMapper.class); job.setMapOutputKeyClass(Text.class); job.setMapOutputValueClass(DoubleWritable.class); job.setReducerClass(UrlModulusReducer.class); job.setOutputKeyClass(Text.class); job.setOutputValueClass(DoubleWritable.class); job.setNumReduceTasks(1); job.waitForCompletion(true); }
public static void main(String[] args) throws Exception { if (args.length != 2) { System.err.println("Usage: OutLinks <input path> <output path>"); System.exit(-1); } Job job = new Job(); job.setJarByClass(PageRank.class); FileInputFormat.addInputPath(job, new Path(args[0])); FileOutputFormat.setOutputPath(job, new Path(args[1])); job.setOutputFormatClass(MapFileOutputFormat.class); job.setMapperClass(PageRankMapper.class); job.setReducerClass(PageRankReducer.class); job.setOutputKeyClass(Text.class); job.setOutputValueClass(DoubleWritable.class); PageRankRead = new MapFileRead("/input/PageRankMap"); //OutLinksRead = new MapFileRead("/input/OutLinksMap"); job.waitForCompletion(true); }
private int getNumPages(Configuration conf, Path titlesDir) throws Exception { int numPages = 0; IntWritable pageNumber = new IntWritable(); MapFile.Reader[] readers = MapFileOutputFormat.getReaders(titlesDir, conf); for (int i = 0; i < readers.length; i++) { readers[i].finalKey(pageNumber); if (pageNumber.get() > numPages) { numPages = pageNumber.get(); } } for (MapFile.Reader reader : readers) { reader.close(); } return numPages; }
@Override public int run(String[] args) throws Exception { if (args.length != 2) { JobBuilder.printUsage(this, "<path> <key>"); return -1; } Path path = new Path(args[0]); IntWritable key = new IntWritable(Integer.parseInt(args[1])); Reader[] readers = MapFileOutputFormat.getReaders(path, getConf()); Partitioner<IntWritable, Text> partitioner = new HashPartitioner<IntWritable, Text>(); Text val = new Text(); Writable entry = MapFileOutputFormat.getEntry(readers, partitioner, key, val); if (entry == null) { System.err.println("Key not found: " + key); return -1; } NcdcRecordParser parser = new NcdcRecordParser(); parser.parse(val.toString()); System.out.printf("%s\t%s\n", parser.getStationId(), parser.getYear()); return 0; }
@Override public int run(String[] args) throws Exception { Job job = JobBuilder.parseInputAndOutput(this, getConf(), args); if (job == null) { return -1; } job.setInputFormatClass(SequenceFileInputFormat.class); job.setOutputKeyClass(IntWritable.class); job.setOutputFormatClass(MapFileOutputFormat.class); SequenceFileOutputFormat.setCompressOutput(job, true); SequenceFileOutputFormat.setOutputCompressorClass(job, BZip2Codec.class); SequenceFileOutputFormat.setOutputCompressionType(job, CompressionType.BLOCK); return job.waitForCompletion(true) ? 0 : 1; }
public Job createSubmittableJob(String[] args) throws IOException { Path partitionsPath = new Path(destPath, PARTITIONS_FILE_NAME); generatePartitions(partitionsPath); Job job = Job.getInstance(getConf(), getConf().get("mapreduce.job.name", "hashTable_" + tableHash.tableName)); Configuration jobConf = job.getConfiguration(); jobConf.setLong(HASH_BATCH_SIZE_CONF_KEY, tableHash.batchSize); job.setJarByClass(HashTable.class); TableMapReduceUtil.initTableMapperJob(tableHash.tableName, tableHash.initScan(), HashMapper.class, ImmutableBytesWritable.class, ImmutableBytesWritable.class, job); // use a TotalOrderPartitioner and reducers to group region output into hash files job.setPartitionerClass(TotalOrderPartitioner.class); TotalOrderPartitioner.setPartitionFile(jobConf, partitionsPath); job.setReducerClass(Reducer.class); // identity reducer job.setNumReduceTasks(tableHash.numHashFiles); job.setOutputKeyClass(ImmutableBytesWritable.class); job.setOutputValueClass(ImmutableBytesWritable.class); job.setOutputFormatClass(MapFileOutputFormat.class); FileOutputFormat.setOutputPath(job, new Path(destPath, HASH_DATA_DIR)); return job; }
public int run(String[] args) throws Exception { if (args.length != 2) { System.out.println("Usage: TitleIndex <titles-sorted.txt> <output-dir>"); ToolRunner.printGenericCommandUsage(System.out); return 2; } Path titlesFile = new Path(args[0]); Path outputDir = new Path(args[1]); Configuration conf = getConf(); // Do not create _SUCCESS files. MapFileOutputFormat.getReaders calls // try to read the _SUCCESS as another MapFile dir. conf.set("mapreduce.fileoutputcommitter.marksuccessfuljobs", "false"); // This job creates a MapFile of the titles indexed by the page id. // UnsplittableTextInputFormat is used to ensure that the same map task // gets all the lines in the titlesFile and it can count the line // numbers. The number of reduce tasks is set to 0. Job job = Job.getInstance(conf, "TitleIndex"); job.setJarByClass(InLinks.class); job.setInputFormatClass(UnsplittableTextInputFormat.class); job.setMapperClass(TitleIndexMapper.class); job.setOutputFormatClass(MapFileOutputFormat.class); job.setOutputKeyClass(IntWritable.class); job.setOutputValueClass(Text.class); FileInputFormat.addInputPath(job, titlesFile); FileOutputFormat.setOutputPath(job, outputDir); job.setNumReduceTasks(0); job.waitForCompletion(true); return 0; }
@Override protected void cleanup(Context context) throws IOException, InterruptedException { Configuration conf = context.getConfiguration(); Path titlesDir = new Path(conf.get("pagerank.titles_dir")); MapFile.Reader[] readers = MapFileOutputFormat.getReaders(titlesDir, conf); Partitioner<IntWritable, Text> partitioner = new HashPartitioner<IntWritable, Text>(); IntWritable page = new IntWritable(); Text title = new Text(); float[] pageRanks = new float[topN.size()]; String[] titles = new String[topN.size()]; // The order of the entries is reversed. The priority queue is in // non-decreasing order and we want the highest PageRank first. for (int i = pageRanks.length - 1; i >= 0; i--) { Map.Entry<Float, Integer> entry = topN.poll(); // Get the title of the page from the title index. page.set(entry.getValue()); MapFileOutputFormat.getEntry(readers, partitioner, page, title); pageRanks[i] = entry.getKey(); titles[i] = title.toString(); } for (MapFile.Reader reader : readers) { reader.close(); } for (int i = 0; i < pageRanks.length; i++) { context.write(new FloatWritable(pageRanks[i]), new Text(titles[i])); } }
private void pageRankIteration(int iter, Configuration conf, Path outputDir) throws Exception { // This job performs an iteration of the power iteration method to // compute PageRank. The map task processes each block M_{i,j}, loads // the corresponding stripe j of the vector v_{k-1} and produces the // partial result of the stripe i of the vector v_k. The reduce task // sums all the partial results of v_k and adds the teleportation factor // (the combiner only sums all the partial results). See Section 5.2 // (and 5.2.3 in particular) of Mining of Massive Datasets // (http://infolab.stanford.edu/~ullman/mmds.html) for details. The // output is written in a "vk" subdir of the output dir, where k is the // iteration number. MapFileOutputFormat is used to keep an array of the // stripes of v. Job job = Job.getInstance(conf, "PageRank:Iteration"); job.setJarByClass(PageRank.class); job.setInputFormatClass(SequenceFileInputFormat.class); job.setMapperClass(PageRankIterationMapper.class); job.setMapOutputKeyClass(ShortWritable.class); job.setMapOutputValueClass(FloatArrayWritable.class); job.setCombinerClass(PageRankIterationCombiner.class); job.setReducerClass(PageRankIterationReducer.class); job.setOutputFormatClass(MapFileOutputFormat.class); job.setOutputKeyClass(ShortWritable.class); job.setOutputValueClass(FloatArrayWritable.class); FileInputFormat.addInputPath(job, new Path(outputDir, "M")); FileOutputFormat.setOutputPath(job, new Path(outputDir, "v" + iter)); job.waitForCompletion(true); }
@Override protected void cleanup(Context context) throws IOException, InterruptedException { Configuration conf = context.getConfiguration(); Path titlesDir = new Path(conf.get("inlinks.titles_dir")); MapFile.Reader[] readers = MapFileOutputFormat.getReaders(titlesDir, conf); Partitioner<IntWritable, Text> partitioner = new HashPartitioner<IntWritable, Text>(); IntWritable page = new IntWritable(); Text title = new Text(); int[] inLinks = new int[topN.size()]; String[] titles = new String[topN.size()]; for (int i = inLinks.length - 1; i >= 0; i--) { Map.Entry<Integer, Integer> entry = topN.poll(); page.set(entry.getValue()); MapFileOutputFormat.getEntry(readers, partitioner, page, title); inLinks[i] = entry.getKey(); titles[i] = title.toString(); } for (MapFile.Reader reader : readers) { reader.close(); } for (int i = 0; i < inLinks.length; i++) { context.write(new IntWritable(inLinks[i]), new Text(titles[i])); } }
@Override public int run(String[] args) throws Exception { if (args.length != 2) { JobBuilder.printUsage(this, "<path> <key>"); return -1; } Path path = new Path(args[0]); IntWritable key = new IntWritable(Integer.parseInt(args[1])); Reader[] readers = MapFileOutputFormat.getReaders(path, getConf()); Partitioner<IntWritable, Text> partitioner = new HashPartitioner<IntWritable, Text>(); Text val = new Text(); Reader reader = readers[partitioner.getPartition(key, val, readers.length)]; Writable entry = reader.get(key, val); if (entry == null) { System.err.println("Key not found: " + key); return -1; } NcdcRecordParser parser = new NcdcRecordParser(); IntWritable nextKey = new IntWritable(); do { parser.parse(val.toString()); System.out.printf("%s\t%s\n", parser.getStationId(), parser.getYear()); } while (reader.next(nextKey, val) && key.equals(nextKey)); return 0; }
@Override public void map(ShortArrayWritable inKey, MatrixBlockWritable inValue, Context context) throws IOException, InterruptedException { // This task gets each block M_{i,j}, loads the corresponding stripe j // of the vector v_{k-1} and produces the partial result of the stripe i // of the vector v_k. Configuration conf = context.getConfiguration(); int iter = Integer.parseInt(conf.get("pagerank.iteration")); int numPages = Integer.parseInt(conf.get("pagerank.num_pages")); short blockSize = Short.parseShort(conf.get("pagerank.block_size")); Writable[] blockIndexes = inKey.get(); short i = ((ShortWritable) blockIndexes[0]).get(); short j = ((ShortWritable) blockIndexes[1]).get(); int vjSize = (j > numPages / blockSize) ? (numPages % blockSize) : blockSize; FloatWritable[] vj = new FloatWritable[vjSize]; if (iter == 1) { // Initial PageRank vector with 1/n for all pages. for (int k = 0; k < vj.length; k++) { vj[k] = new FloatWritable(1.0f / numPages); } } else { // Load the stripe j of the vector v_{k-1} from the MapFiles. Path outputDir = MapFileOutputFormat.getOutputPath(context).getParent(); Path vjDir = new Path(outputDir, "v" + (iter - 1)); MapFile.Reader[] readers = MapFileOutputFormat.getReaders(vjDir, conf); Partitioner<ShortWritable, FloatArrayWritable> partitioner = new HashPartitioner<ShortWritable, FloatArrayWritable>(); ShortWritable key = new ShortWritable(j); FloatArrayWritable value = new FloatArrayWritable(); MapFileOutputFormat.getEntry(readers, partitioner, key, value); Writable[] writables = value.get(); for (int k = 0; k < vj.length; k++) { vj[k] = (FloatWritable) writables[k]; } for (MapFile.Reader reader : readers) { reader.close(); } } // Initialize the partial result i of the vector v_k. int viSize = (i > numPages / blockSize) ? (numPages % blockSize) : blockSize; FloatWritable[] vi = new FloatWritable[viSize]; for (int k = 0; k < vi.length; k++) { vi[k] = new FloatWritable(0); } // Multiply M_{i,j} by the stripe j of the vector v_{k-1} to obtain the // partial result i of the vector v_k. Writable[][] blockColumns = inValue.get(); for (int k = 0; k < blockColumns.length; k++) { Writable[] blockColumn = blockColumns[k]; if (blockColumn.length > 0) { int vDegree = ((ShortWritable) blockColumn[0]).get(); for (int columnIndex = 1; columnIndex < blockColumn.length; columnIndex++) { int l = ((ShortWritable) blockColumn[columnIndex]).get(); vi[l].set(vi[l].get() + (1.0f / vDegree) * vj[k].get()); } } } context.write(new ShortWritable(i), new FloatArrayWritable(vi)); }
/** * Save a {@code JavaRDD<List<Writable>>} to a Hadoop {@link org.apache.hadoop.io.MapFile}. Each record is * given a <i>unique and contiguous</i> {@link LongWritable} key, and values are stored as * {@link RecordWritable} instances.<br> * <b>Note</b>: If contiguous keys are not required, using a sequence file instead is preferable from a performance * point of view. Contiguous keys are often only required for non-Spark use cases, such as with * {@link org.datavec.hadoop.records.reader.mapfile.MapFileRecordReader} * <p> * Use {@link #restoreMapFileSequences(String, JavaSparkContext)} to restore values saved with this method. * * @param path Path to save the MapFile * @param rdd RDD to save * @param c Configuration object, used to customise options for the map file * @param maxOutputFiles Nullable. If non-null: first coalesce the RDD to the specified size (number of partitions) * to limit the maximum number of output map files * @see #saveMapFileSequences(String, JavaRDD) * @see #saveSequenceFile(String, JavaRDD) */ public static void saveMapFile(String path, JavaRDD<List<Writable>> rdd, Configuration c, @Nullable Integer maxOutputFiles) { path = FilenameUtils.normalize(path, true); if (maxOutputFiles != null) { rdd = rdd.coalesce(maxOutputFiles); } JavaPairRDD<List<Writable>, Long> dataIndexPairs = rdd.zipWithIndex(); //Note: Long values are unique + contiguous, but requires a count JavaPairRDD<LongWritable, RecordWritable> keyedByIndex = dataIndexPairs.mapToPair(new RecordSavePrepPairFunction()); keyedByIndex.saveAsNewAPIHadoopFile(path, LongWritable.class, RecordWritable.class, MapFileOutputFormat.class, c); }
/** * Save a {@code JavaRDD<List<List<Writable>>>} to a Hadoop {@link org.apache.hadoop.io.MapFile}. Each record is * given a <i>unique and contiguous</i> {@link LongWritable} key, and values are stored as * {@link SequenceRecordWritable} instances.<br> * <b>Note</b>: If contiguous keys are not required, using a sequence file instead is preferable from a performance * point of view. Contiguous keys are often only required for non-Spark use cases, such as with * {@link org.datavec.hadoop.records.reader.mapfile.MapFileSequenceRecordReader}<br> * <p> * Use {@link #restoreMapFileSequences(String, JavaSparkContext)} to restore values saved with this method. * * @param path Path to save the MapFile * @param rdd RDD to save * @param c Configuration object, used to customise options for the map file * @see #saveMapFileSequences(String, JavaRDD) * @see #saveSequenceFile(String, JavaRDD) */ public static void saveMapFileSequences(String path, JavaRDD<List<List<Writable>>> rdd, Configuration c, @Nullable Integer maxOutputFiles) { path = FilenameUtils.normalize(path, true); if (maxOutputFiles != null) { rdd = rdd.coalesce(maxOutputFiles); } JavaPairRDD<List<List<Writable>>, Long> dataIndexPairs = rdd.zipWithIndex(); JavaPairRDD<LongWritable, SequenceRecordWritable> keyedByIndex = dataIndexPairs.mapToPair(new SequenceRecordSavePrepPairFunction()); keyedByIndex.saveAsNewAPIHadoopFile(path, LongWritable.class, SequenceRecordWritable.class, MapFileOutputFormat.class, c); }