利用艺术家的整数ID映射将标签转换为向量

标签：

<strong><span style="font-size:18px;">/***
 * @author YangXin
 * @info Mapper选择艺术家的整数特征ID然后建立单个特征的向量。这些一维的部分
 * 向量会传给Reducer，后者会将这些向量简单地进行联结，生成一个完整的向量。
 */
package unitTwelve;

import java.io.IOException;
import java.util.HashMap;
import java.util.Map;
import java.util.regex.Pattern;
import org.apache.hadoop.conf.Configuration;
import org.apache.hadoop.io.DefaultStringifier;
import org.apache.hadoop.io.LongWritable;
import org.apache.hadoop.io.Text;
import org.apache.hadoop.mapreduce.Mapper;
import org.apache.hadoop.util.GenericsUtil;
import org.apache.mahout.math.NamedVector;
import org.apache.mahout.math.SequentialAccessSparseVector;
import org.apache.mahout.math.VectorWritable;

public class VectorMapper extends Mapper<LongWritable, Text, Text, VectorWritable>{
	private Pattern splitter;
	private VectorWritable writer;
	private Map<String, Integer> dictionary = new HashMap<String, Integer>();
	
	@Override
	protected void map(LongWritable key, Text value, Context context) throws IOException, InterruptedException{
		String[] fields = splitter.split(value.toString());
		if(fields.length < 4){
			context.getCounter("Map", "LinesWithErrors").increment(1);
			return;
		}
		String arrtist = fields[1];
		String tag = fields[2];
		double weight = Double.parseDouble(fields[3]);
		NamedVector vector = new NamedVector(new SequentialAccessSparseVector(dictionary.size()), tag);
		vector.set(dictionary.get(value), weight);
		writer.set(vector);
		context.write(new Text(tag), writer);
	}
	
	@Override
	protected void setup(Context context) throws IOException, InterruptedException{
		super.setup(context);
		Configuration conf = context.getConfiguration();
		DefaultStringifier<Map<String, Integer>> mapStringifier = new DefaultStringifier<Map<String, Integer>>(conf, GenericsUtil.getClass(dictionary));
		dictionary = mapStringifier.fromString(conf.get("dictionary"));
		splitter = Pattern.compile("<sep>");
		writer = new VectorWritable();
	}
}
</span></strong>

利用艺术家的整数ID映射将标签转换为向量

标签：

原文地址：http://blog.csdn.net/u012965373/article/details/50811633