Spark SQL窗口函数

窗口函数又叫着窗口分析函数，Spark 1.4版本SparkSQL支持窗口分析函数，主要用于分组函数；理解窗口函数，可以参考blog去了理解： http://www.cnblogs.com/CareySon/p/3411176.html

数据准备（用空格隔开）

Spark 100
Hadoop 65
Spark 99
Hadoop 61
Spark 195
Hadoop 60
Spark 98
Hadoop 69
Spark 91
Hadoop 98
Spark 88
Hadoop 99
Spark 68
Hadoop 60
Spark 79
Hadoop 97
Spark 69
Hadoop 96

代码编写

package com.imf.spark.sql

import org.apache.spark.SparkConf
import org.apache.spark.SparkContext
import org.apache.spark.sql.hive.HiveContext

object SparkSQLWindowFunctionOps {
  def main(args: Array[String]): Unit = {
    val conf = new SparkConf()
    conf.setAppName("SparkSQLWindowFunctionOps for scala")
    conf.setMaster("spark://master1:7077")
    val sc = new SparkContext(conf)

    val hiveContext = new HiveContext(sc);
    hiveContext.sql("use testdb")//使用hive中的testdb数据库
    hiveContext.sql("drop table if exists scores")
    hiveContext.sql("create table if not exists scores(name String,score int) "
      +"ROW FORMAT DELIMITED FIELDS TERMINATED BY ' ' LINES TERMINATED BY '\\n'")
    hiveContext.sql("load data local inpath '/usr/local/sparkApps/SparkSQLWindowFunctionOps/TopNGroup.txt' INTO TABLE scores")
    /**
     * 使用子查询的方式完成目标数据的提取，在目标函数内幕使用窗口函数row_number来进行分组排序：
     * partition by :指定窗口函数分组的key
     * order by :分组后进行排序
     */
    val result = hiveContext.sql("select name,score " 
                                  +" from ( "
                                  +" select name,score,row_number() over(partition by name order by score desc)rank from scores ) sub_scores"
                                  +" where rank <=4")

      result .show();//在Driver的控制台上打印出结果内容

    //保存到hive数据仓库中

    hiveContext.sql("drop table if exists sortedResultScores")
    result.saveAsTable("sortedResultScores")

  }
}

调度脚本
/usr/local/spark/spark-1.6.0-bin-hadoop2.6/bin/spark-submit \
--class com.imf.spark.sql.SparkSQLWindowFunctionOps \
--files /usr/local/hive/apache-hive-1.2.1-bin/conf/hive-site.xml \
--master spark://master1:7077 \
/usr/local/sparkApps/SparkSQLWindowFunctionOps/SparkSQLWindowFunctionOps.jar

查看结果

hive> show tables;
OK
people
peopleresult
peoplescores
scores
sortedresultscores
student
student2
student3
student4
tbsogou
tmp_pre_hour_seach_info
Time taken: 0.395 seconds, Fetched: 11 row(s)
hive> select * from scores;
OK
Spark    100
Hadoop    65
Spark    99
Hadoop    61
Spark    195
Hadoop    60
Spark    98
Hadoop    69
Spark    91
Hadoop    98
Spark    88
Hadoop    99
Spark    68
Hadoop    60
Spark    79
Hadoop    97
Spark    69
Hadoop    96
Time taken: 2.426 seconds, Fetched: 18 row(s)
hive> select * from sortedresultscores;
OK
SLF4J: Failed to load class "org.slf4j.impl.StaticLoggerBinder".
SLF4J: Defaulting to no-operation (NOP) logger implementation
SLF4J: See http://www.slf4j.org/codes.html#StaticLoggerBinder for further details.
Spark    195
Spark    100
Spark    99
Spark    98
Hadoop    99
Hadoop    98
Hadoop    97
Hadoop    96
Time taken: 0.229 seconds, Fetched: 8 row(s)

来源参考：
王家林：中国Spark第一人
DT大数据梦工厂
新浪微博：http://weibo.com.ilovepains/
手机：18610086859
QQ：1740415547
联系邮箱[email protected]