5大视频网站数据分析mapreduce


一、需求
 自定义输入格式 完成统计任务 输出多个文件

输入数据:5个网站的 每天电视剧的 播放量 收藏数 评论数 踩数 赞数

输出数据:按网站类别 统计每个电视剧的每个指标的总量

任务目标:自定义输入格式 完成统计任务 输出多个文件

二、数据

部分数据

这里写图片描述

三、思路

第一步:定义一个电视剧热度数据的bean。

第二步:定义一个读取热度数据的InputFormat类。

第三步:写MapReduce统计程序

第四步:上传tvplay.txt数据集到HDFS,并运行程序

四、代码

1.利用WritableComparable接口,自定义一个TVWritable类,实现WritableComparable类,将各个参数封装起来,便于计算。
package com.pc.hadoop.pc.tv;

import java.io.DataInput;
import java.io.DataOutput;
import java.io.IOException;

import org.apache.hadoop.io.WritableComparable;

public class TVWritable implements WritableComparable

{
    //定义5个成员变量
    private int view;
    private int collection;
    private int comment;
    private int diss;
    private int up;

    //构造函数
    public TVWritable(){}


    //定义一个set方法,用this关键字对封装好的数据进行引用
    public void set(int view,int collection,int comment, int diss,int up)
    {

        this.view = view;
        this.collection = collection;
        this.comment = comment;
        this.diss = diss;
        this.up = up;
    }

    //使用get和set对封装好的数据进行存取
    public int getView()
    {
        return view;
    }
    public void setView(int view)
    {
        this.view = view;
    }


    public int getCollection()
    {
        return collection;
    }
    public void setCollection(int collection)
    {
        this.collection = collection;
    }


    public int getComment()
    {
        return comment;
    }
    public void setComment(int comment)
    {
        this.comment = comment;
    }


    public int getDiss()
    {
        return diss;
    }
    public void setDiss(int diss)
    {
        this.diss = diss;
    }


    public int getUp()
    {
        return up;
    }
    public void setUp(int up)
    {
        this.up = up;
    }

    //实现WritableComparaqble的redafields()方法,以便该数据能被序列化后完成网络传输或文件输入。
    @Override
    public void readFields(DataInput in) throws IOException
    {
        // TODO Auto-generated method stub

        view = in.readInt();
        collection = in.readInt();
        comment = in.readInt();
        diss = in.readInt();
        up = in.readInt();

    }

    //实现WritableComparaqble的write()方法,以便该数据能被反序列化后完成网络传输或文件输入。
    @Override
    public void write(DataOutput out) throws IOException
    {
        // TODO Auto-generated method stub
        out.writeInt(view);
        out.writeInt(collection);
        out.writeInt(comment);
        out.writeInt(diss);
        out.writeInt(up);
    }

    //使用compareTo对其中的数据进行比较
    @Override
    public int compareTo(Object o)
    {
        // TODO Auto-generated method stub
        return 0;
    }

}
2.自定义一个TVInputFormat类取继承FileInputFormat文件输入格式这个父类,然后对createRecordReader()方法进行重写,其实质则是重写TVRecordReader()这个方法,
得到其返回值,利用TVRecordReader()这个方法去继承RecordReader()这个方法。
package com.pc.hadoop.pc.tv;

import java.io.IOException;

import org.apache.hadoop.conf.Configuration;
import org.apache.hadoop.fs.FSDataInputStream;
import org.apache.hadoop.fs.FileSystem;
import org.apache.hadoop.fs.Path;
import org.apache.hadoop.io.Text;
import org.apache.hadoop.mapreduce.InputSplit;
import org.apache.hadoop.mapreduce.RecordReader;
import org.apache.hadoop.mapreduce.TaskAttemptContext;
import org.apache.hadoop.mapreduce.lib.input.FileInputFormat;
import org.apache.hadoop.mapreduce.lib.input.FileSplit;
import org.apache.hadoop.util.LineReader;

public class TVInputFormat extends FileInputFormat<Text,TVWritable>
{
    protected boolean isSplitable()
    {
        return false;
    }

    @Override
    public RecordReader<Text, TVWritable> createRecordReader(InputSplit inputsplit, TaskAttemptContext context) throws IOException, InterruptedException
    {
        // TODO Auto-generated method stub
        return new TVRecordReader();
    }

    public static class TVRecordReader extends RecordReader<Text,TVWritable>
    {
        public LineReader in; //自定义行读取器
        public Text lineKey; //声明key类型
        public TVWritable lineValue; //自定义value
        public Text line; //每行数据类型

        //
        @Override
        public void close() throws IOException
        {
            // TODO Auto-generated method stub
            if(in != null)
            {
                in.close();
            }

        }

        //获取当前key
        @Override
        public Text getCurrentKey() throws IOException, InterruptedException
        {
            // TODO Auto-generated method stub
            return lineKey;
        }

        //获取当前value
        @Override
        public TVWritable getCurrentValue() throws IOException, InterruptedException

        {
            // TODO Auto-generated method stub
            return lineValue;
        }

        //获取当前进程
        @Override
        public float getProgress() throws IOException, InterruptedException

        {
            // TODO Auto-generated method stub
            return 0;
        }

        //初始化
        @Override
        public void initialize(InputSplit inputsplit, TaskAttemptContext context) throws IOException, InterruptedException

        {
            // TODO Auto-generated method stub

            FileSplit split = (FileSplit) inputsplit;//获取分片内容
            Configuration job = context.getConfiguration();//读取配置信息
            Path file = split.getPath();//获取路径
            FileSystem fs = file.getFileSystem(job);//获取文件系统

            FSDataInputStream filein = fs.open(file);//通过文件系统打开文件,对文件进行读取
            in = new LineReader(filein,job);
            lineKey = new Text();//新建一个Text实例作为自定义输入格式的key
            lineValue = new TVWritable();
            line = new Text();



        }

        @Override
        public boolean nextKeyValue() throws IOException, InterruptedException

        {
            // TODO Auto-generated method stub
            int lineSize = in.readLine(line);
            if(lineSize == 0)
            return false;
            //读取每行数据解数组i
            String[] i = line.toString().split(" ");
            if(i.length != 7)

            {
                throw new IOException("Invalid record received");
            }
            //自定义key和value的值
            lineKey.set(i[0]+" "+i[1]);//电视剧名称和所属视频网站
            lineValue.set(Integer.parseInt(i[2].trim()),
            Integer.parseInt(i[3].trim()),
            Integer.parseInt(i[4].trim()),
            Integer.parseInt(i[5].trim()),
            Integer.parseInt(i[6].trim() ));
            return true;

        }
    }
}

3.使用MapperReducer对输入的数据进行进行相应的处理输出想要得到的结果。
 在reduce在定义一个多输出的对象MultipleOutputs
/**
     * @input Params Text TvPlayData
     * @output Params Text TvPlayData
     * @author yangjun
     * @function 直接输出
     */
    public static class TVPlayMapper extends
            Mapper<Text, TVWritable, Text, TVWritable> {
        @Override
        protected void map(Text key, TVWritable value, Context context)
                throws IOException, InterruptedException {
            context.write(key, value);
        }
    }
    /**
     * @input Params Text TvPlayData
     * @output Params Text Text
     * @author yangjun
     * @fuction 统计每部电视剧的 点播数 收藏数等  按source输出到不同文件夹下
     */
    public static class TVPlayReducer extends
            Reducer<Text, TVWritable, Text, Text> {
        private Text m_key = new Text();
        private Text m_value = new Text();
        private MultipleOutputs<Text, Text> mos;

        protected void setup(Context context) throws IOException,
                InterruptedException {
            mos = new MultipleOutputs<Text, Text>(context);
        }//将 MultipleOutputs 的初始化放在 setup() 中,因为在 setup() 只会被调用一次
//定义reduce() 方法里的 multipleOutputs.write(…)。你需要把以前的 context.write(…) 替换成现在的这个
        protected void reduce(Text Key, Iterable<TVWritable> Values,
                Context context) throws IOException, InterruptedException {
             int view = 0;
             int collection = 0;
             int comment = 0;
             int diss = 0;
             int up = 0;
            for (TVWritable a:Values) {
                 view += a.getView();
                 collection += a.getCollection();
                 comment +=a.getComment();
                 diss += a.getDiss();
                 up += a.getUp();
            }
            //tvname  source
            String[] records = Key.toString().split(" ");
            // 1优酷2搜狐3土豆4爱奇艺5迅雷看看
            String source = records[1];// 媒体类别
            m_key.set(records[0]);
            m_value.set(view+" "+collection+" "+comment+" "+diss+" "+up);
            if (source.equals("1")) {
                mos.write("youku", m_key, m_value);
            } else if (source.equals("2")) {
                mos.write("souhu", m_key, m_value);
            } else if (source.equals("3")) {
                mos.write("tudou", m_key, m_value);
            } else if (source.equals("4")) {
                mos.write("aiqiyi", m_key, m_value);
            } else if (source.equals("5")) {
                mos.write("xunlei", m_key, m_value);
            }
        }

        protected void cleanup(Context context) throws IOException,
                InterruptedException {
            mos.close();   //关闭 MultipleOutputs,也就是关闭 RecordWriter,并且是一堆 RecordWriter,因为这里会有很多 reduce 被调用。
        }
    }

4 运行run函数对作业进行运行,并自定义输出MultipleOutputs函数调用addNameoutput方法对其进行设置多路径的输出。
@Override
    public int run(String[] args) throws Exception {
        Configuration conf = new Configuration();// 配置文件对象
        Path mypath = new Path(args[1]);
        FileSystem hdfs = mypath.getFileSystem(conf);// 创建输出路径
        if (hdfs.isDirectory(mypath)) {
            hdfs.delete(mypath, true);
        }

        Job job = new Job(conf, "tvplay");// 构造任务
        job.setJarByClass(TVplay.class);// 设置主类

        job.setMapperClass(TVPlayMapper.class);// 设置Mapper
        job.setMapOutputKeyClass(Text.class);// key输出类型
        job.setMapOutputValueClass(TVWritable.class);// value输出类型
        job.setInputFormatClass(TVInputFormat.class);//自定义输入格式

        job.setReducerClass(TVPlayReducer.class);// 设置Reducer
        job.setOutputKeyClass(Text.class);// reduce key类型
        job.setOutputValueClass(Text.class);// reduce value类型
        // 自定义文件输出格式,通过路径名(pathname)来指定输出路径
        MultipleOutputs.addNamedOutput(job, "youku", TextOutputFormat.class,
                Text.class, Text.class);
        MultipleOutputs.addNamedOutput(job, "souhu", TextOutputFormat.class,
                Text.class, Text.class);
        MultipleOutputs.addNamedOutput(job, "tudou", TextOutputFormat.class,
                Text.class, Text.class);
        MultipleOutputs.addNamedOutput(job, "aiqiyi", TextOutputFormat.class,
                Text.class, Text.class);
        MultipleOutputs.addNamedOutput(job, "xunlei", TextOutputFormat.class,
                Text.class, Text.class);

        FileInputFormat.addInputPath(job, new Path(args[0]));// 输入路径
        FileOutputFormat.setOutputPath(job, new Path(args[1]));// 输出路径
        job.waitForCompletion(true);
        return 0;
    }
    public static void main(String[] args) throws Exception {
        String[] args0 = { "hdfs://pc1:9000/home/hadoop/tvplay/tvplay.txt",
                "hdfs://pc1:9000/home/hadoop/tvplay/out/" };
        int ec = ToolRunner.run(new Configuration(), new TVplay(), args0);
        //public static int run(Configuration conf,Tool tool, String[] args),可以在job运行的时候指定配置文件或其他参数
        //这个方法调用tool的run(String[])方法,并使用conf中的参数,以及args中的参数,而args一般来源于命令行。
        System.exit(ec);
    }

原文地址:https://www.cnblogs.com/fengyouheng/p/10266812.html