第一个MapReduce的例子

Hadoop Guide的第一个MapReduce的例子是处理气象数据的（数据来源ncdc），终于跑通了。总结一下步骤，安装hadoop不在本文中介绍

1 数据预处理

1.1 下载数据

测试数据需要在ncdc的官方ftp上进行下载，年份跨度范围1901到2016，不写个脚本下载，靠手工是行不通的，脚本如下：
download.sh

!bin/bash
for i in {1901..2015}
do
wget --execute robots=off -r -np -nH -P./ncdc/ --cut-dirs=4 -R index.html* ftp://ftp.ncdc.noaa.gov/pub/data/gsod/$i/gsod_$i.tar
done

执行命令[nohup sh download.sh &] 会把所有ftp上gsod_*.tar的压缩包下载下来，下载完之后需要预处理这些数据

1.2 数据预处理

每一个tar压缩包包含n个gz压缩包，每一个gz包含一个数据文本，关于ncdc气象数据的每个字段的描述在这里,格式举例如下

STN--- WBAN   YEARMODA    TEMP       DEWP      SLP        STP       VISIB      WDSP     MXSPD   GUST    MAX     MIN   PRCP   SNDP   FRSHTT
607450 99999  20100101    56.1 22    33.0 22  1012.4  8   975.5  8    5.6 22    9.1 22   19.0  999.9    63.9    48.2*  0.00G 999.9  000000
607450 99999  20100102    53.0 23    34.2 23  1019.5  8   982.2  8    5.8 23    6.7 23   12.0  999.9    66.7    39.0   0.00G 999.9  000000
607450 99999  20100103    50.5 23    34.3 23  1022.2  8   984.6  8    6.2 23    7.7 23   12.0  999.9    64.9    36.5   0.00G 999.9  000000
607450 99999  20100104    53.0 22    34.5 22  1016.5  8   979.3  8    6.4 22    6.5 22   15.9  999.9    64.9    42.8*  0.00G 999.9  000000

预处理的目标是把压缩包里面的数据按照一个年份一个txt的形式存在，为了Map阶段读书去数据方便，去除第一行的title。处理脚本如下，processh.sh

#!/bin/bash

for i in {1901..2017}
do
  tar xf ./ncdc/gsod_$i.tar -C ./ncdc
  gunzip ./ncdc/*.gz

  rm -rf ncdc/input_gsod_$i.txt
  touch ncdc/input_gsod_$i.txt

  for file in ./ncdc/*.op
  do
    sed -i '1d' $file
    cat $file >> ./ncdc/input_gsod_$i.txt
  done
  rm -rf ./ncdc/*.op
  echo "file gsod_$i has processed "
done

1.3 Load数据到HDFS上

创建input目录：

hdfs dfs -mkdir /ncdc

put数据到hdfs上：

hdfs dfs -put ./ncdc/*.txt /ncdc/*

检查hdfs上的数据，如果所有的年份的数据都load到了hdfs上就OK

hdfs dfs -ls /ncdc

2 MapReduce程序

2.1 MapReduce程序

package com.oldtrafford.hadoop;

import org.apache.hadoop.fs.Path;
import org.apache.hadoop.io.IntWritable;
import org.apache.hadoop.io.LongWritable;
import org.apache.hadoop.io.Text;
import org.apache.hadoop.mapreduce.Job;
import org.apache.hadoop.mapreduce.Mapper;
import org.apache.hadoop.mapreduce.Reducer;
import org.apache.hadoop.mapreduce.lib.input.FileInputFormat;
import org.apache.hadoop.mapreduce.lib.output.FileOutputFormat;
import java.io.IOException;

public class MaxTemperature {

    public static void main (String[] args) throws IOException, ClassNotFoundException, InterruptedException {
        if(args.length!=2){
            System.err.println("usage: maxtemperature <input>");
            System.exit(-1);
        }

        Job job = new Job();
        job.setJarByClass(MaxTemperature.class);
        job.setJobName("MaxTemperature");

        FileInputFormat.addInputPath(job, new Path(args[0]));
        FileOutputFormat.setOutputPath(job, new Path(args[1]));

        job.setMapperClass(MaxTemperatureMapper.class);
        job.setReducerClass(MaxTemperatureReducer.class);

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

        job.waitForCompletion(true);
        System.out.println("Finished");
    }

    static class MaxTemperatureMapper extends Mapper<LongWritable, Text, Text, IntWritable> {
        private static final String MISSING = "9999.9";
        @Override
        public void map(LongWritable key, Text value, Mapper.Context context) throws IOException, InterruptedException {
            String line = value.toString();
            String year = line.substring(14, 18);

            String tempretureStr = line.substring(24,30).trim();
            int temperature = -1;
            if(!MISSING.equals(tempretureStr)){
                temperature = (int)(Double.parseDouble(tempretureStr)*10);
                context.write(new Text(year), new IntWritable(temperature));
            }
        }
    }

    static class MaxTemperatureReducer extends Reducer<Text, IntWritable, Text, IntWritable> {
        @Override
        public void reduce(Text key, Iterable<IntWritable> values, Context context) throws IOException, InterruptedException {
            int maxValue = Integer.MIN_VALUE;
            for(IntWritable value : values) {
                maxValue = Math.max(maxValue, value.get());
            }
            context.write(key, new IntWritable(maxValue));
        }
    }
}

2.2 打包代码&&执行

打包代码
打包mapreduce的代码用最简单的maven命令，会产生一个jar包。然后把这个jar传输到hadoop机器的上一台机器上。

执行hadoop

hadoop jar temperature.jar com.oldtrafford.hadoop.MaxTemperature /ncdc/* /ncdc_output

3 查看执行结果

查看mapreduce结果

hdfs dfs -cat /ncdc_output/*