mahout贝叶斯算法开发思路（拓展篇）2

如果想直接下面算法调用包，可以直接在mahout贝叶斯算法拓展下载，该算法调用的方式如下：

$HADOOP_HOME/bin hadoop jar mahout.jar mahout.fansy.bayes.BayerRunner -i hdfs_input_path -o hdfs_output_path -scl : -scv ,

调用参数如下：

usage: <command> [Generic Options] [Job-Specific Options]
Generic Options:
 -archives <paths>              comma separated archives to be unarchived
                                on the compute machines.
 -conf <configuration file>     specify an application configuration file
 -D <property=value>            use value for given property
 -files <paths>                 comma separated files to be copied to the
                                map reduce cluster
 -fs <local|namenode:port>      specify a namenode
 -jt <local|jobtracker:port>    specify a job tracker
 -libjars <paths>               comma separated jar files to include in
                                the classpath.
 -tokenCacheFile <tokensFile>   name of the file with the tokens
Job-Specific Options:                                                           
  --input (-i) input                                    Path to job input       
                                                        directory.              
  --output (-o) output                                  The directory pathname  
                                                        for output.             
  --splitCharacterVector (-scv) splitCharacterVector    Vector split            
                                                        character,default is    
                                                        ','                     
  --splitCharacterLabel (-scl) splitCharacterLabel      Vector and Label split  
                                                        character,default is    
                                                        ':'                     
  --help (-h)                                           Print out help          
  --tempDir tempDir                                     Intermediate output     
                                                        directory               
  --startPhase startPhase                               First phase to run      
  --endPhase endPhase                                   Last phase to run

接上篇分析下面的步骤：

4. 获取贝叶斯模型的属性值2：

这一步骤相当于 TrainNaiveBayesJob的第二个prepareJob，其中mapper和reducer都是参考这个job的，基本没有修改代码；代码如下：

package mahout.fansy.bayes;

import org.apache.hadoop.conf.Configuration;
import org.apache.hadoop.fs.Path;
import org.apache.hadoop.io.Text;
import org.apache.hadoop.mapreduce.Job;
import org.apache.hadoop.mapreduce.lib.input.SequenceFileInputFormat;
import org.apache.hadoop.mapreduce.lib.output.SequenceFileOutputFormat;
import org.apache.hadoop.util.ToolRunner;
import org.apache.mahout.classifier.naivebayes.training.WeightsMapper;
import org.apache.mahout.common.AbstractJob;
import org.apache.mahout.common.HadoopUtil;
import org.apache.mahout.common.mapreduce.VectorSumReducer;
import org.apache.mahout.math.VectorWritable;
/**
 * 贝叶斯算法第二个job任务相当于 TrainNaiveBayesJob的第二个prepareJob
 * Mapper，Reducer还用原来的
 * @author Administrator
 *
 */
public class BayesJob2 extends AbstractJob {
	/**
	 * @param args
	 * @throws Exception 
	 */
	public static void main(String[] args) throws Exception {
		ToolRunner.run(new Configuration(), new BayesJob2(),args);
	}
	
	@Override
	public int run(String[] args) throws Exception {
		addInputOption();
	    addOutputOption();
	    addOption("labelNumber","ln", "The number of the labele ");
	    if (parseArguments(args) == null) {
		      return -1;
		}
	    Path input = getInputPath();
	    Path output = getOutputPath();
	    String labelNumber=getOption("labelNumber");
	    Configuration conf=getConf();
	    conf.set(WeightsMapper.class.getName() + ".numLabels",labelNumber);
	    HadoopUtil.delete(conf, output);
	    Job job=new Job(conf);
	    job.setJobName("job2 get weightsFeture and weightsLabel by job1's output:"+input.toString());
	    job.setJarByClass(BayesJob2.class); 
	    
	    job.setInputFormatClass(SequenceFileInputFormat.class);
	    job.setOutputFormatClass(SequenceFileOutputFormat.class);
	    
	    job.setMapperClass(WeightsMapper.class);
	    job.setMapOutputKeyClass(Text.class);
	    job.setMapOutputValueClass(VectorWritable.class);
	    job.setCombinerClass(VectorSumReducer.class);
	    job.setReducerClass(VectorSumReducer.class);
	    job.setOutputKeyClass(Text.class);
	    job.setOutputValueClass(VectorWritable.class);
	    SequenceFileInputFormat.setInputPaths(job, input);
	    SequenceFileOutputFormat.setOutputPath(job, output);
	    
	    if(job.waitForCompletion(true)){
	    	return 0;
	    }
		return -1;
	}

}

其单独调用方式如下：

usage: <command> [Generic Options] [Job-Specific Options]
Generic Options:
 -archives <paths>              comma separated archives to be unarchived
                                on the compute machines.
 -conf <configuration file>     specify an application configuration file
 -D <property=value>            use value for given property
 -files <paths>                 comma separated files to be copied to the
                                map reduce cluster
 -fs <local|namenode:port>      specify a namenode
 -jt <local|jobtracker:port>    specify a job tracker
 -libjars <paths>               comma separated jar files to include in
                                the classpath.
 -tokenCacheFile <tokensFile>   name of the file with the tokens
Job-Specific Options:                                                           
  --input (-i) input                 Path to job input directory.               
  --output (-o) output               The directory pathname for output.         
  --labelNumber (-ln) labelNumber    The number of the labele                   
  --help (-h)                        Print out help                             
  --tempDir tempDir                  Intermediate output directory              
  --startPhase startPhase            First phase to run                         
  --endPhase endPhase                Last phase to run   

其实也就是设置一个标识的个数而已，其他参考AbstractJob的默认参数；

5.贝叶斯模型写入文件：

这一步把3、4步骤的输出进行转换然后作为贝叶斯模型的一部分，然后把贝叶斯模型写入文件，其中的转换以及写入文件都参考BayesUtils中的相关方法，具体代码如下：

package mahout.fansy.bayes;

import java.io.IOException;

import mahout.fansy.bayes.util.OperateArgs;

import org.apache.commons.cli.ParseException;
import org.apache.hadoop.conf.Configuration;
import org.apache.hadoop.fs.Path;
import org.apache.hadoop.io.IntWritable;
import org.apache.hadoop.io.Text;
import org.apache.mahout.classifier.naivebayes.NaiveBayesModel;
import org.apache.mahout.classifier.naivebayes.training.ThetaMapper;
import org.apache.mahout.classifier.naivebayes.training.TrainNaiveBayesJob;
import org.apache.mahout.common.Pair;
import org.apache.mahout.common.iterator.sequencefile.PathFilters;
import org.apache.mahout.common.iterator.sequencefile.PathType;
import org.apache.mahout.common.iterator.sequencefile.SequenceFileDirIterable;
import org.apache.mahout.math.Matrix;
import org.apache.mahout.math.SparseMatrix;
import org.apache.mahout.math.Vector;
import org.apache.mahout.math.VectorWritable;

import com.google.common.base.Preconditions;

public class WriteBayesModel extends OperateArgs{

	/**
	 * @param args,输入和输出都是没有用的，输入是job1和job 2 的输出，输出是model的路径
	 * model存储的路径是 输出路径下面的naiveBayesModel.bin文件
	 * @throws ParseException 
	 * @throws IOException 
	 */
	public static void main(String[] args) throws IOException, ParseException {
		String[] arg={"-jt","ubuntu:9001",
				"-i","",
				"-o","",
				"-mp","hdfs://ubuntu:9000/user/mahout/output_bayes/bayesModel",
				"-bj1","hdfs://ubuntu:9000/user/mahout/output_bayes/job1",
				"-bj2","hdfs://ubuntu:9000/user/mahout/output_bayes/job2"};
		new WriteBayesModel().run(arg);
	}
	/**
	 * 把model写入文件中
	 * @param args
	 * @throws IOException
	 * @throws ParseException
	 */
	public  int run(String[] args) throws IOException, ParseException{
	
		// modelPath
        setOption("mp","modelPath",true,"the path for bayesian model to store",true);  
        // bayes job 1 path
        setOption("bj1","bayesJob1",true,"the path for bayes job 1",true);  
        // bayes job 2 path
        setOption("bj2","bayesJob2",true,"the path for bayes job 2",true);  
		if(!parseArgs(args)){
			return -1;
		}
		String job1Path=getNameValue("bj1");
		String job2Path=getNameValue("bj2");
		Configuration conf=getConf();
		String modelPath=getNameValue("mp");
		NaiveBayesModel naiveBayesModel=readFromPaths(job1Path,job2Path,conf);
		naiveBayesModel.validate();
	    naiveBayesModel.serialize(new Path(modelPath), getConf());
	    System.out.println("Write bayesian model to '"+modelPath+"/naiveBayesModel.bin'");
	    return 0;
	}
	/**
	 * 摘自BayesUtils的readModelFromDir方法，只修改了相关路径
	 * @param job1Path
	 * @param job2Path
	 * @param conf
	 * @return
	 */
	public  NaiveBayesModel readFromPaths(String job1Path,String job2Path,Configuration conf){
		float alphaI = conf.getFloat(ThetaMapper.ALPHA_I, 1.0f);
	    // read feature sums and label sums
	    Vector scoresPerLabel = null;
	    Vector scoresPerFeature = null;
	    for (Pair<Text,VectorWritable> record : new SequenceFileDirIterable<Text, VectorWritable>(
	        new Path(job2Path), PathType.LIST, PathFilters.partFilter(), conf)) {
	      String key = record.getFirst().toString();
	      VectorWritable value = record.getSecond();
	      if (key.equals(TrainNaiveBayesJob.WEIGHTS_PER_FEATURE)) {
	        scoresPerFeature = value.get();
	      } else if (key.equals(TrainNaiveBayesJob.WEIGHTS_PER_LABEL)) {
	        scoresPerLabel = value.get();
	      }
	    }

	    Preconditions.checkNotNull(scoresPerFeature);
	    Preconditions.checkNotNull(scoresPerLabel);

	    Matrix scoresPerLabelAndFeature = new SparseMatrix(scoresPerLabel.size(), scoresPerFeature.size());
	    for (Pair<IntWritable,VectorWritable> entry : new SequenceFileDirIterable<IntWritable,VectorWritable>(
	        new Path(job1Path), PathType.LIST, PathFilters.partFilter(), conf)) {
	      scoresPerLabelAndFeature.assignRow(entry.getFirst().get(), entry.getSecond().get());
	    }

	    Vector perlabelThetaNormalizer = scoresPerLabel.like();
	    return new NaiveBayesModel(scoresPerLabelAndFeature, scoresPerFeature, scoresPerLabel, perlabelThetaNormalizer,
	        alphaI);
	}
	
}

6. 应用贝叶斯模型分类原始数据：

这个部分的代码也基本是参考mahout中贝叶斯算法的源码，只是修改了其中的解析部分的代码而已，具体如下：

package mahout.fansy.bayes;

import java.io.IOException;

import org.apache.hadoop.conf.Configuration;
import org.apache.hadoop.fs.Path;
import org.apache.hadoop.io.Text;
import org.apache.hadoop.mapreduce.Job;
import org.apache.hadoop.mapreduce.Mapper;
import org.apache.hadoop.mapreduce.lib.input.SequenceFileInputFormat;
import org.apache.hadoop.mapreduce.lib.output.SequenceFileOutputFormat;
import org.apache.hadoop.util.ToolRunner;
import org.apache.mahout.classifier.naivebayes.AbstractNaiveBayesClassifier;
import org.apache.mahout.classifier.naivebayes.NaiveBayesModel;
import org.apache.mahout.classifier.naivebayes.StandardNaiveBayesClassifier;
import org.apache.mahout.classifier.naivebayes.training.WeightsMapper;
import org.apache.mahout.common.AbstractJob;
import org.apache.mahout.common.HadoopUtil;
import org.apache.mahout.math.Vector;
import org.apache.mahout.math.VectorWritable;
/**
 * 用于分类的Job
 * @author Administrator
 *
 */
public class BayesClassifyJob extends AbstractJob {
	/**
	 * @param args
	 * @throws Exception 
	 */
	public static void main(String[] args) throws Exception {
		ToolRunner.run(new Configuration(), new BayesClassifyJob(),args);
	}
	
	@Override
	public int run(String[] args) throws Exception {
		addInputOption();
	    addOutputOption();
	    addOption("model","m", "The file where bayesian model store ");
	    addOption("labelNumber","ln", "The labels number ");
	    if (parseArguments(args) == null) {
		      return -1;
		}
	    Path input = getInputPath();
	    Path output = getOutputPath();
	    String labelNumber=getOption("labelNumber");
	    String modelPath=getOption("model");
	    Configuration conf=getConf();
	    conf.set(WeightsMapper.class.getName() + ".numLabels",labelNumber);
	    HadoopUtil.cacheFiles(new Path(modelPath), conf);
	    HadoopUtil.delete(conf, output);
	    Job job=new Job(conf);
	    job.setJobName("Use bayesian model to classify the  input:"+input.getName());
	    job.setJarByClass(BayesClassifyJob.class); 
	    
	    job.setInputFormatClass(SequenceFileInputFormat.class);
	    job.setOutputFormatClass(SequenceFileOutputFormat.class);
	    
	    job.setMapperClass(BayesClasifyMapper.class);
	    job.setMapOutputKeyClass(Text.class);
	    job.setMapOutputValueClass(VectorWritable.class);
	    job.setNumReduceTasks(0);
	    job.setOutputKeyClass(Text.class);
	    job.setOutputValueClass(VectorWritable.class);
	    SequenceFileInputFormat.setInputPaths(job, input);
	    SequenceFileOutputFormat.setOutputPath(job, output);
	    
	    if(job.waitForCompletion(true)){
	    	return 0;
	    }
		return -1;
	}
	/**
	 *  自定义Mapper，只修改了解析部分代码
	 * @author Administrator
	 *
	 */
	public static class BayesClasifyMapper extends Mapper<Text, VectorWritable, Text, VectorWritable>{
		private AbstractNaiveBayesClassifier classifier;
			@Override
		  public void setup(Context context) throws IOException, InterruptedException {
		    System.out.println("Setup");
		    Configuration conf = context.getConfiguration();
		    Path modelPath = HadoopUtil.cachedFile(conf);
		    NaiveBayesModel model = NaiveBayesModel.materialize(modelPath, conf);
		    classifier = new StandardNaiveBayesClassifier(model);
		  }

		  @Override
		  public void map(Text key, VectorWritable value, Context context) throws IOException, InterruptedException {
		    Vector result = classifier.classifyFull(value.get());
		    //the key is the expected value
		    context.write(new Text(key.toString()), new VectorWritable(result));
		  }
	}
}

如果要单独运行这一步，可以参考：

usage: <command> [Generic Options] [Job-Specific Options]
Generic Options:
 -archives <paths>              comma separated archives to be unarchived
                                on the compute machines.
 -conf <configuration file>     specify an application configuration file
 -D <property=value>            use value for given property
 -files <paths>                 comma separated files to be copied to the
                                map reduce cluster
 -fs <local|namenode:port>      specify a namenode
 -jt <local|jobtracker:port>    specify a job tracker
 -libjars <paths>               comma separated jar files to include in
                                the classpath.
 -tokenCacheFile <tokensFile>   name of the file with the tokens
Job-Specific Options:                                                           
  --input (-i) input                 Path to job input directory.               
  --output (-o) output               The directory pathname for output.         
  --model (-m) model                 The file where bayesian model store        
  --labelNumber (-ln) labelNumber    The labels number                          
  --help (-h)                        Print out help                             
  --tempDir tempDir                  Intermediate output directory              
  --startPhase startPhase            First phase to run                         
  --endPhase endPhase                Last phase to run 

只需提供model的路径和标识的个数这两个参数即可；

7. 对第6步分类的结果进行评价，这部分的代码如下：

package mahout.fansy.bayes;

import java.io.IOException;
import java.util.Map;

import mahout.fansy.bayes.util.OperateArgs;

import org.apache.commons.cli.ParseException;
import org.apache.hadoop.conf.Configuration;
import org.apache.hadoop.fs.Path;
import org.apache.hadoop.io.Text;
import org.apache.mahout.classifier.ClassifierResult;
import org.apache.mahout.classifier.ResultAnalyzer;
import org.apache.mahout.classifier.naivebayes.BayesUtils;
import org.apache.mahout.common.Pair;
import org.apache.mahout.common.iterator.sequencefile.PathFilters;
import org.apache.mahout.common.iterator.sequencefile.PathType;
import org.apache.mahout.common.iterator.sequencefile.SequenceFileDirIterable;
import org.apache.mahout.math.Vector;
import org.apache.mahout.math.VectorWritable;
import org.slf4j.Logger;
import org.slf4j.LoggerFactory;

public class AnalyzeBayesModel extends OperateArgs{

	/**
	 * 输入是BayesClassifyJob的输出
	 * -o 参数没作用
	 */
	private static final Logger log = LoggerFactory.getLogger(AnalyzeBayesModel.class);
	public static void main(String[] args) throws IOException, ParseException {
		String[] arg={"-jt","ubuntu:9001",
				"-i","hdfs://ubuntu:9000/user/mahout/output_bayes/classifyJob",
				"-o","",
				"-li","hdfs://ubuntu:9000/user/mahout/output_bayes/index.bin"
				};
		new AnalyzeBayesModel().run(arg);
	}
	/**
	 * 分析BayesClassifyJob输出文件和labelIndex做对比，分析正确率
	 * @param args
	 * @throws IOException
	 * @throws ParseException
	 */
	public  int run(String[] args) throws IOException, ParseException{
	
		 // labelIndex
        setOption("li","labelIndex",true,"the path where labelIndex store",true);  

		if(!parseArgs(args)){
			return -1;
		}
		Configuration conf=getConf();
		String labelIndex=getNameValue("labelIndex");
		String input=getInput();
		Path inputPath=new Path(input);
		//load the labels
	    Map<Integer, String> labelMap = BayesUtils.readLabelIndex(getConf(), new Path(labelIndex));

	    //loop over the results and create the confusion matrix
	    SequenceFileDirIterable<Text, VectorWritable> dirIterable =
	        new SequenceFileDirIterable<Text, VectorWritable>(inputPath,
	                                                          PathType.LIST,
	                                                          PathFilters.partFilter(),
	                                                          conf);
	    ResultAnalyzer analyzer = new ResultAnalyzer(labelMap.values(), "DEFAULT");
	    analyzeResults(labelMap, dirIterable, analyzer);

	    log.info("{} Results: {}",  "Standard NB", analyzer);
	    return 0;
	}
	/**
	 * 摘自TestNaiveBayesDriver中的analyzeResults方法
	 */
	private  void analyzeResults(Map<Integer, String> labelMap,
            SequenceFileDirIterable<Text, VectorWritable> dirIterable,
            ResultAnalyzer analyzer) {
		for (Pair<Text, VectorWritable> pair : dirIterable) {
			int bestIdx = Integer.MIN_VALUE;
			double bestScore = Long.MIN_VALUE;
			for (Vector.Element element : pair.getSecond().get()) {
				if (element.get() > bestScore) {
					bestScore = element.get();
					bestIdx = element.index();
				}
			}
			if (bestIdx != Integer.MIN_VALUE) {
				ClassifierResult classifierResult = new ClassifierResult(labelMap.get(bestIdx), bestScore);
				analyzer.addInstance(pair.getFirst().toString(), classifierResult);
			}
		}
	}
	
}

运行拓展篇1中的数据得到的模型的分类结果如下：

13/09/14 14:52:13 INFO bayes.AnalyzeBayesModel: Standard NB Results: =======================================================
Summary
-------------------------------------------------------
Correctly Classified Instances          :          7	        70%
Incorrectly Classified Instances        :          3	        30%
Total Classified Instances              :         10

=======================================================
Confusion Matrix
-------------------------------------------------------
a    	b    	c    	d    	<--Classified as
3    	0    	0    	0    	 |  3     	a     = 1
0    	1    	0    	1    	 |  2     	b     = 2
1    	1    	2    	0    	 |  4     	c     = 3
0    	0    	0    	1    	 |  1     	d     = 4

运行后可以在hdfs上面看到如下的文件夹：

任务列表如下：

分享，成长，快乐

转载请注明blog地址：http://blog.csdn.net/fansy1990