MapReduce算法形式一:WordCount

MapReduce算法形式一:WordCount

这种形式可以做一些网站登陆次数,或者某个电商网站的商品销量啊诸如此类的,主要就是求和,但是求和之前还是要好好清洗数据的,以免数据缺省值太多,影响真实性。

废话不多说,上代码吧,我把注释一行行的都写了~~可可可可~

先封装了数据行的对象:

public class Log {
  private String time;
  private String UID;
  private String keyWord;
  private int rank;
  private int order;
  private String URL;

  public String getTime() {
    return time;
  }
  public void setTime(String time) {
    this.time = time;
  }
  public String getUID() {
    return UID;
  }
  public void setUID(String uID) {
    UID = uID;
  }
  public String getKeyWord() {
    return keyWord;
  }
  public void setKeyWord(String keyWord) {
    this.keyWord = keyWord;
  }
  public int getRank() {
    return rank;
  }
  public void setRank(int rank) {
    this.rank = rank;
  }
  public int getOrder() {
    return order;
  }
  public void setOrder(int order) {
    this.order = order;
  }
  public String getURL() {
    return URL;
  }
  public void setURL(String uRL) {
    URL = uRL;
  }

  public Log(String time, String uID, String keyWord, int rank, int order,String uRL) {
    super();
    this.time = time;
    this.UID = uID;
    this.keyWord = keyWord;
    this.rank = rank;
    this.order = order;
    this.URL = uRL;
  }

  public Log() {
    super();
  }

/*
* 对行记录日志信息进行封装成对象
* 并将对象返回
*/
  public static Log getInfo(String value){
    Log log = new Log();

    //将一条日志记录转换成一个数组
    String[] lines = value.toString().trim().split("\t");
    //判断行记录中间是否有缺省值
    if(lines.length == 6){
      //行记录封装
      log.setTime(lines[0].trim());
      log.setUID(lines[1].trim());
      log.setKeyWord(lines[2].trim());
      log.setRank(Integer.parseInt(lines[3].trim()));
      log.setOrder(Integer.parseInt(lines[4].trim()));
      log.setURL(lines[5].trim());
    }
      return log;
  }

}

 

 

mr中的代码:

public class PVSum {
/**案例一:WordCount
*
* 非空查询条数
* 不去重,直接统计总和即可
*
* 假设:
* 日志格式如下:(已经过清洗,以制表符分割)
* 20111230050630 时间time
* 2a12e06f50ad41063ed2b62bffac29ad 用户UID
* 361泰国电影 搜索的关键词keyword
* 5 rank搜索结果排序
* 8 order点击次数
* http://www.57ge.com/play/?play_2371_1_361.html 访问的URL
*
* @param args
* @throws Exception
*/
public static void main(String[] path) throws Exception {
  if(path.length != 2){
    System.out.println("please input full path!");
    System.exit(0);
  }

  Job job = Job.getInstance(new Configuration(), PVSum.class.getSimpleName());
  job.setJarByClass(PVSum.class);

  FileInputFormat.setInputPaths(job, new Path(path[0]));
  FileOutputFormat.setOutputPath(job, new Path(path[1]));

  job.setMapperClass(PVSumMap.class);
  job.setReducerClass(PVSumReduce.class);

  job.setOutputKeyClass(Text.class);
  job.setOutputValueClass(IntWritable.class);

  job.waitForCompletion(true);
}

public static class PVSumMap extends Mapper<LongWritable, Text, Text, IntWritable> {
  IntWritable one = new IntWritable(1);//记录数量,一条记录即为1
  Text text = new Text("非空关键词的PV访问量总计:");
  protected void map(LongWritable key, Text value,org.apache.hadoop.mapreduce.Mapper<LongWritable, Text, Text, IntWritable>.Context context)
            throws java.io.IOException, InterruptedException {
    //获取每条记录的对象
    Log log = Log.getInfo(value.toString().trim());
    //判断关键字是否为空
    if(log.getKeyWord().trim() != null && !log.getKeyWord().trim().equals("")){
      //写入数据
      context.write(text, one);
      //map : <非空关键词的PV访问量总计:, 1>
    }
  };
}

//shuffle : <非空关键词的PV访问量总计:, {1, 1, 1...}>

public static class PVSumReduce extends Reducer<Text, IntWritable, Text, IntWritable> {
  protected void reduce(Text key, java.lang.Iterable<IntWritable> values, 

              org.apache.hadoop.mapreduce.Reducer<Text, IntWritable, Text, IntWritable>.Context context)

              throws java.io.IOException, InterruptedException {
      int sum = 0;//记录总条数
      for (IntWritable count : values) {
        sum += count.get();
      }
      context.write(key, new IntWritable(sum));
  };
}

}

 

posted @ 2016-09-10 09:38  yoghurt2016  阅读(397)  评论(0编辑  收藏  举报