系列文章目录

spark第一章：环境安装
spark第二章：sparkcore实例

文章目录

• 系列文章目录
• 前言
• 一、idea创建项目
• 二、编写实例
• • 1.WordCount
  • 2.RDD实例
  • 3.Spark实例
• 总结

前言

上次我们搭建了环境，现在就要开始上实例，这次拖了比较长的时间，实在是sparkcore的知识点有点多，而且例子有些复杂，尽自己最大的能力说清楚，说不清楚也没办法了。

一、idea创建项目

这个可以参考我搭建scala环境
Scala第一章:环境搭建
pom添加需要的依赖

    org.apache.sparkspark-core_2.123.2.3

二、编写实例

1.WordCount

先用一下WordCount进行进行一下环境测试

package com.atguigu.bigdata.spark.core.rdd.wcimport org.apache.spark.rdd.RDD
import org.apache.spark.{SparkConf, SparkContext}object WordCount {def main(args: Array[String]): Unit = {//  创建 Spark 运行配置对象val sparkConf: SparkConf = new SparkConf().setMaster("local[*]").setAppName("WordCount")// 创建 Spark 上下文环境对象（连接对象）val sc : SparkContext = new SparkContext(sparkConf)// 读取文件 获取一行一行的数据val lines: RDD[String] = sc.textFile("datas/word.txt")// 将一行数据进行拆分val words: RDD[String] = lines.flatMap(_.split(" "))// 将数据根据单次进行分组，便于统计val wordToOne: RDD[(String, Int)] = words.map(word => (word, 1))// 对分组后的数据进行转换val wordToSum: RDD[(String, Int)] = wordToOne.reduceByKey(_ + _)// 打印输出val array: Array[(String, Int)] = wordToSum.collect()array.foreach(println)sc.stop()}

如果能出现如下结果，代表环境正常，可以继续进行实验。

2.RDD实例

1.数据准备
agent.log：时间戳，省份，城市，用户，广告，中间字段使用空格分隔。

实验数据可以到尚硅谷的平台寻找。
RDD_seq.scala

package com.atguigu.bigdata.spark.core.rdd.builderimport org.apache.spark.rdd.RDD
import org.apache.spark.{SparkConf, SparkContext}object RDD_seq {def main(args: Array[String]): Unit = {val sparkConf: SparkConf = new SparkConf().setMaster("local[*]").setAppName("Operator")val sc = new SparkContext(sparkConf)//1.获取原始数据val dataRDD: RDD[String] = sc.textFile("datas/agent.log")//2.将原始数据进行结构的转换。方便统计// 时间戳 省份 城市 用户 广告 =>// ((省份，广告)，1)val mapRDD: RDD[((String, String), Int)] = dataRDD.map((line: String) => {val datas: Array[String] = line.split(" ")((datas(1), datas(4)), 1)})//3.将转换结构厚的数据进行分组聚合// ((省份，广告)，1)=>((省份，广告)，sum)val reduceRDD: RDD[((String, String), Int)] = mapRDD.reduceByKey((_: Int) + (_: Int))//4.将聚合的结果进行结构的转换// ((省份，广告)，sum) =>(省份，(广告，sum))val newMapRDD: RDD[(String, (String, Int))] = reduceRDD.map {case ((prv, ad), sum) =>(prv, (ad, sum))}//5.将转换结构后的数据根据省份进行分组//(省份，[(广告A，sumA)(广告B，sumB)])val groupRDD: RDD[(String, Iterable[(String, Int)])] = newMapRDD.groupByKey()//6.将分组后的数据组内排序（降序） 取前三名val resultRDD: RDD[(String, List[(String, Int)])] = groupRDD.mapValues((iter: Iterable[(String, Int)]) => {iter.toList.sortBy((_: (String, Int))._2)(Ordering.Int.reverse).take(3)})//7.采集数据打印在控制台resultRDD.collect().foreach(println)sc.stop()}
}

3.Spark实例

1.数据准备
user_visit_action.txt
主要包含用户的 4 种行为：搜索，点击，下单，支付。数据规则如下：

• 数据文件中每行数据采用下划线分隔数据
• 每一行数据表示用户的一次行为，这个行为只能是 4 种行为的一种
• 如果搜索关键字为 null,表示数据不是搜索数据
• 如果点击的品类 ID 和产品 ID 为-1，表示数据不是点击数据
• 针对于下单行为，一次可以下单多个商品，所以品类 ID 和产品 ID 可以是多个，id 之间采用逗号分隔，如果本次不是下单行为，则数据采用null 表示
• 支付行为和下单行为类似
  

需求 1：Top10 热门品类
品类是指产品的分类，大型电商网站品类分多级，咱们的项目中品类只有一级，不同的
公司可能对热门的定义不一样。我们按照每个品类的点击、下单、支付的量来统计热门品类。

鞋 点击数 下单数 支付数
衣服 点击数 下单数 支付数
电脑 点击数 下单数 支付数
例如，综合排名 = 点击数20%+下单数30%+支付数*50%
本项目需求优化为：先按照点击数排名，靠前的就排名高；如果点击数相同，再比较下
单数；下单数再相同，就比较支付数
HotCategoryTop10Analysis.scala

package com.atguigu.bigdata.spark.core.rdd.seqimport org.apache.spark.rdd.RDD
import org.apache.spark.{SparkConf, SparkContext}object HotCategoryTop10Analysis {def main(args: Array[String]): Unit = {val sparkConf: SparkConf =new SparkConf().setMaster("local[*]").setAppName("HotCategoryTop10Analysis")val sc = new SparkContext(sparkConf)//1.读取原始日志数据val actionRDD: RDD[String] = sc.textFile("datas/user_visit_action.txt")//2.统计品类的点击数量：（品类ID，点击数量）val clickActionRDD: RDD[String] = actionRDD.filter((action: String) => {val datas: Array[String] = action.split("_")datas(6) != "-1"})val clickCountRDD: RDD[(String, Int)] = clickActionRDD.map((action: String) => {val datas: Array[String] = action.split("_")(datas(6), 1)}).reduceByKey((_: Int) + (_: Int))//3.统计品类的下单数量：（品类ID，下单数量）val orderActionRDD: RDD[String] = actionRDD.filter((action: String) => {val datas: Array[String] = action.split("_")datas(8) != "null"})val orderCountRDD: RDD[(String, Int)] = orderActionRDD.flatMap((action: String) => {val datas: Array[String] = action.split("_")val cid: String = datas(8)val cids: Array[String] = cid.split(",")cids.map((id: String) => (id, 1))}).reduceByKey((_: Int) + (_: Int))//4.统计品类的支付数量：（品类ID，支付数量）val payActionRDD: RDD[String] = actionRDD.filter((action: String) => {val datas: Array[String] = action.split("_")datas(10) != "null"})val payCountRDD: RDD[(String, Int)] = payActionRDD.flatMap((action: String) => {val datas: Array[String] = action.split("_")val cid: String = datas(10)val cids: Array[String] = cid.split(",")cids.map((id: String) => (id, 1))}).reduceByKey((_: Int) + (_: Int))//5.将品类进行排序，并且取钱10名val cogroupRDD: RDD[(String, (Iterable[Int], Iterable[Int], Iterable[Int]))] =clickCountRDD.cogroup(orderCountRDD, payCountRDD)val analysisRDD: RDD[(String, (Int, Int, Int))] = cogroupRDD.mapValues {case (clickIter, orderIter, payIter) =>var clickCnt = 0val iter1: Iterator[Int] = clickIter.iteratorif (iter1.hasNext) {clickCnt = iter1.next()}var orderCnt = 0val iter2: Iterator[Int] = orderIter.iteratorif (iter2.hasNext) {orderCnt = iter2.next()}var payCnt = 0val iter3: Iterator[Int] = payIter.iteratorif (iter3.hasNext) {payCnt = iter3.next()}(clickCnt, orderCnt, payCnt)}val resultRDD: Array[(String, (Int, Int, Int))] = analysisRDD.sortBy((_: (String, (Int, Int, Int)))._2, ascending = false).take(10)//6.将结果采集到控制台打印出来resultRDD.foreach(println)sc.stop()}
}

HotCategoryTop10Analysis1.scala

package com.atguigu.bigdata.spark.core.rdd.seqimport org.apache.spark.rdd.RDD
import org.apache.spark.{SparkConf, SparkContext}object HotCategoryTop10Analysis1 {def main(args: Array[String]): Unit = {val sparkConf: SparkConf = new SparkConf().setMaster("local[*]").setAppName("HotCategoryTop10Analysis")val sc = new SparkContext(sparkConf)//1.读取原始日志数据val actionRDD: RDD[String] = sc.textFile("datas/user_visit_action.txt")actionRDD.cache()//2.统计品类的点击数量：（品类ID，点击数量）val clickActionRDD: RDD[String] = actionRDD.filter((action: String) => {val datas: Array[String] = action.split("_")datas(6) != "-1"})val clickCountRDD: RDD[(String, Int)] = clickActionRDD.map((action: String) => {val datas: Array[String] = action.split("_")(datas(6), 1)}).reduceByKey((_: Int) + (_: Int))//3.统计品类的下单数量：（品类ID，下单数量）val orderActionRDD: RDD[String] = actionRDD.filter((action: String) => {val datas: Array[String] = action.split("_")datas(8) != "null"})val orderCountRDD: RDD[(String, Int)] = orderActionRDD.flatMap((action: String) => {val datas: Array[String] = action.split("_")val cid: String = datas(8)val cids: Array[String] = cid.split(",")cids.map((id: String) => (id, 1))}).reduceByKey((_: Int) + (_: Int))//4.统计品类的支付数量：（品类ID，支付数量）val payActionRDD: RDD[String] = actionRDD.filter((action: String) => {val datas: Array[String] = action.split("_")datas(10) != "null"})val payCountRDD: RDD[(String, Int)] = payActionRDD.flatMap((action: String) => {val datas: Array[String] = action.split("_")val cid: String = datas(10)val cids: Array[String] = cid.split(",")cids.map((id: String) => (id, 1))}).reduceByKey((_: Int) + (_: Int))//5.将品类进行排序，并且取钱10名val rdd1: RDD[(String, (Int, Int, Int))] = clickCountRDD.map {case (cid, cnt) =>(cid, (cnt, 0, 0))}val rdd2: RDD[(String, (Int, Int, Int))] = orderCountRDD.map {case (cid, cnt) =>(cid, (0, cnt, 0))}val rdd3: RDD[(String, (Int, Int, Int))] = payCountRDD.map {case (cid, cnt) =>(cid, (0, 0, cnt))}//将三个数据源合并在一起，统一进行聚合计算val sourceRDD: RDD[(String, (Int, Int, Int))] = rdd1.union(rdd2).union(rdd3)val analysisRDD: RDD[(String, (Int, Int, Int))] = sourceRDD.reduceByKey((t1: (Int, Int, Int), t2: (Int, Int, Int)) => {(t1._1 + t2._1, t1._2 + t2._2 ,t1._3 + t2._3)})val resultRDD: Array[(String, (Int, Int, Int))] = analysisRDD.sortBy((_: (String, (Int, Int, Int)))._2, ascending = false).take(10)//6.将结果采集到控制台打印出来resultRDD.foreach(println)sc.stop()}
}

HotCategoryTop10Analysis2.scala

package com.atguigu.bigdata.spark.core.rdd.seqimport org.apache.spark.rdd.RDD
import org.apache.spark.{SparkConf, SparkContext}object HotCategoryTop10Analysis2 {def main(args: Array[String]): Unit = {val sparkConf: SparkConf = new SparkConf().setMaster("local[*]").setAppName("HotCategoryTop10Analysis")val sc = new SparkContext(sparkConf)//1.读取原始日志数据val actionRDD: RDD[String] = sc.textFile("datas/user_visit_action.txt")//2.将数据转换结构val flatRDD: RDD[(String, (Int, Int, Int))] = actionRDD.flatMap(action => {val datas = action.split("_")if (datas(6) != "-1") {//点击List((datas(6), (1, 0, 0)))} else if (datas(8) != "null") {//下单val ids = datas(8).split(",")ids.map(id => (id, (0, 1, 0)))} else if (datas(10) != "null") {//支付val ids = datas(10).split(",")ids.map(id => (id, (0, 0, 1)))} else {Nil}})//3.将相同的品类ID的数据进行分组聚合val analysisRDD=flatRDD.reduceByKey((t1,t2)=>(t1._1 + t2._1, t1._2 + t2._2 ,t1._3 + t2._3))//4.将统计结果根据数据进行降序出来，去前10名val resultRDD: Array[(String, (Int, Int, Int))] = analysisRDD.sortBy((_: (String, (Int, Int, Int)))._2, ascending = false).take(10)//5.将结果采集到控制台打印出来resultRDD.foreach(println)sc.stop()}
}

HotCategoryTop10Analysis3.scala
package com.atguigu.bigdata.spark.core.rdd.seq

import org.apache.spark.rdd.RDD
import org.apache.spark.util.AccumulatorV2
import org.apache.spark.{SparkConf, SparkContext}

import scala.collection.mutable

object HotCategoryTop10Analysis3 {
def main(args: Array[String]): Unit = {
val sparkConf: SparkConf = new SparkConf().setMaster(“local[*]”).setAppName(“HotCategoryTop10Analysis”)
val sc = new SparkContext(sparkConf)

//1.读取原始日志数据
val actionRDD: RDD[String] = sc.textFile("datas/user_visit_action.txt")val acc =new HotCategoryAccumulator
sc.register(acc,"hotCategory")
//2.将数据转换结构
actionRDD.foreach(action => {val datas = action.split("_")if (datas(6) != "-1") {//点击acc.add((datas(6),"click"))} else if (datas(8) != "null") {//下单val ids = datas(8).split(",")ids.foreach(id=>{acc.add(id,"order")})} else if (datas(10) != "null") {//支付val ids = datas(10).split(",")ids.foreach(id=>{acc.add(id,"pay")})}}
)val accVal: mutable.Map[String, HotCategory] = acc.value
val categories: mutable.Iterable[HotCategory] = accVal.map(_._2)val sort=categories.toList.sortWith((left,right)=>{if (left.clickCnt>right.clickCnt) {true} else if (left.clickCnt==right.clickCnt){if (left.orderCnt>right.orderCnt) {true} else if(left.orderCnt==right.orderCnt) {left.payCnt>right.payCnt}else {false}}else{false}}
)//5.将结果采集到控制台打印出来
sort.take(10).foreach(println)sc.stop()

}
case class HotCategory(cid:String,var clickCnt:Int,var orderCnt :Int,var payCnt:Int)

class HotCategoryAccumulator extends AccumulatorV2[(String,String),mutable.Map[String,HotCategory]]{

private val hcMap=mutable.Map[String,HotCategory]()override def isZero: Boolean = {hcMap.isEmpty
}override def copy(): AccumulatorV2[(String, String), mutable.Map[String, HotCategory]] = {new HotCategoryAccumulator
}override def reset(): Unit = {hcMap.clear()
}override def add(v: (String, String)): Unit = {val cid= v._1val actionType= v._2val category: HotCategory = hcMap.getOrElse(cid, HotCategory(cid, 0, 0, 0))if (actionType =="click"){category.clickCnt+=1}else if (actionType == "order") {category.orderCnt+=1}else if (actionType =="pay"){category.payCnt+=1}hcMap.update(cid,category)
}override def merge(other: AccumulatorV2[(String, String), mutable.Map[String, HotCategory]]): Unit = {val map1=this.hcMapval map2=other.valuemap2.foreach{case (cid,hc)=>{val category: HotCategory = map1.getOrElse(cid, HotCategory(cid, 0, 0, 0))category.clickCnt+=hc.clickCntcategory.orderCnt+=hc.orderCntcategory.payCnt+=hc.payCntmap1.update(cid,category)}}
}override def value: mutable.Map[String, HotCategory] = hcMap

}
}

需求 2：Top10 热门品类中每个品类的 Top10 活跃 Session 统计
在需求一的基础上，增加每个品类用户 session 的点击统计
HotCategoryTop10SessionAnalysis.scala

package com.atguigu.bigdata.spark.core.rdd.seqimport org.apache.spark.rdd.RDD
import org.apache.spark.{SparkConf, SparkContext}object HotCategoryTop10SessionAnalysis {def main(args: Array[String]): Unit = {val sparkConf: SparkConf = new SparkConf().setMaster("local[*]").setAppName("HotCategoryTop10Analysis")val sc = new SparkContext(sparkConf)val actionRDD: RDD[String] = sc.textFile("datas/user_visit_action.txt")actionRDD.cache()val top10Ids: Array[String] = top10Category(actionRDD)//1.过滤原始数据,保留点击和前10品类IDval filterActionRDD: RDD[String] = actionRDD.filter(action => {val datas: Array[String] = action.split("_")if (datas(6) != "-1") {top10Ids.contains(datas(6))} else {false}})//2.根据评类ID和session进行点击量的统计val reduceRDD: RDD[((String, String), Int)] = filterActionRDD.map(action => {val datas: Array[String] = action.split("_")((datas(6), datas(2)), 1)}) reduceByKey (_ + _)//3.将统计的结果进行结构的转换val mapRDD: RDD[(String, (String, Int))] = reduceRDD.map {case ((cid, sid), sum) => {(cid, (sid, sum))}}//4.相同的品类进行分组val groupRDD: RDD[(String, Iterable[(String, Int)])] = mapRDD.groupByKey()//5.将分组后的数据进行点击量的排序，取前10名val resultRDD: RDD[(String, List[(String, Int)])] = groupRDD.mapValues(iter => {iter.toList.sortBy(_._2)(Ordering.Int.reverse).take(10)})resultRDD.foreach(println)sc.stop()}def top10Category(actionRDD: RDD[String]) ={val flatRDD: RDD[(String, (Int, Int, Int))] = actionRDD.flatMap(action => {val datas = action.split("_")if (datas(6) != "-1") {List((datas(6), (1, 0, 0)))} else if (datas(8) != "null") {val ids = datas(8).split(",")ids.map(id => (id, (0, 1, 0)))} else if (datas(10) != "null") {val ids = datas(10).split(",")ids.map(id => (id, (0, 0, 1)))} else {Nil}})val analysisRDD=flatRDD.reduceByKey((t1,t2)=>(t1._1 + t2._1, t1._2 + t2._2 ,t1._3 + t2._3))analysisRDD.sortBy((_: (String, (Int, Int, Int)))._2, ascending = false).take(10).map(_._1)}
}

需求 3：页面单跳转换率统计
页面单跳转化率
计算页面单跳转化率，什么是页面单跳转换率，比如一个用户在一次 Session 过程中
访问的页面路径 3,5,7,9,10,21，那么页面 3 跳到页面 5 叫一次单跳，7-9 也叫一次单跳，
那么单跳转化率就是要统计页面点击的概率。
比如：计算 3-5 的单跳转化率，先获取符合条件的 Session 对于页面 3 的访问次数（PV） 为 A，然后获取符合条件的 Session 中访问了页面 3 又紧接着访问了页面 5 的次数为 B，
那么 B/A 就是 3-5 的页面单跳转化率

PageflowAnalysis.scala

package com.atguigu.bigdata.spark.core.rdd.seqimport org.apache.spark.rdd.RDD
import org.apache.spark.{SparkConf, SparkContext}object PageflowAnalysis {def main(args: Array[String]): Unit = {val sparkConf: SparkConf = new SparkConf().setMaster("local[*]").setAppName("HotCategoryTop10Analysis")val sc = new SparkContext(sparkConf)val actionRDD: RDD[String] = sc.textFile("datas/user_visit_action.txt")val actionDataRDD=actionRDD.map(action=>{val datas: Array[String] = action.split("_")UserVisitAction(datas(0),datas(1).toLong,datas(2),datas(3).toLong,datas(4),datas(5),datas(6).toLong,datas(7).toLong,datas(8),datas(9),datas(10),datas(11),datas(12).toLong,)})actionDataRDD.cache()val pageidToCountMap: Map[Long, Long] = actionDataRDD.map(action => {(action.page_id, 1L)}).reduceByKey(_ + _).collect().toMapval sessionRDD: RDD[(String, Iterable[UserVisitAction])] = actionDataRDD.groupBy(_.session_id)val mvRDD: RDD[(String, List[((Long, Long), Int)])] = sessionRDD.mapValues(iter => {val sortList: List[UserVisitAction] = iter.toList.sortBy(_.action_time)val flowIds: List[Long] = sortList.map(_.page_id)val pageflowIds: List[(Long, Long)] = flowIds.zip(flowIds.tail)pageflowIds.map(t => {(t, 1)})})val flatRDD: RDD[((Long, Long), Int)] = mvRDD.map(_._2).flatMap(list => list)val dataRDD: RDD[((Long, Long), Int)] = flatRDD.reduceByKey(_ + _)dataRDD.foreach {case ((pageid1, pageid2), sum) => {val lon: Long = pageidToCountMap.getOrElse(pageid1, 0L)println(s"页面${pageid1}跳转到页面${pageid2}单跳转换率为:"+(sum.toDouble/lon))}}sc.stop()}case class UserVisitAction(date: String,//用户点击行为的日期user_id: Long,//用户的 IDsession_id: String,//Session 的 IDpage_id: Long,//某个页面的 IDaction_time: String,//动作的时间点search_keyword: String,//用户搜索的关键词click_category_id: Long,//某一个商品品类的 IDclick_product_id: Long,//某一个商品的 IDorder_category_ids: String,//一次订单中所有品类的 ID 集合order_product_ids: String,//一次订单中所有商品的 ID 集合pay_category_ids: String,//一次支付中所有品类的 ID 集合pay_product_ids: String,//一次支付中所有商品的 ID 集合city_id: Long)//城市 id
}

可以看到这个数据非常多，所以我们精简一下，只求出几个特定页面的跳出率。
PageflowAnalysis2.scala

package com.atguigu.bigdata.spark.core.rdd.seqimport org.apache.spark.rdd.RDD
import org.apache.spark.{SparkConf, SparkContext}object PageflowAnalysis {def main(args: Array[String]): Unit = {val sparkConf: SparkConf = new SparkConf().setMaster("local[*]").setAppName("HotCategoryTop10Analysis")val sc = new SparkContext(sparkConf)val actionRDD: RDD[String] = sc.textFile("datas/user_visit_action.txt")val actionDataRDD=actionRDD.map(action=>{val datas: Array[String] = action.split("_")UserVisitAction(datas(0),datas(1).toLong,datas(2),datas(3).toLong,datas(4),datas(5),datas(6).toLong,datas(7).toLong,datas(8),datas(9),datas(10),datas(11),datas(12).toLong,)})actionDataRDD.cache()val pageidToCountMap: Map[Long, Long] = actionDataRDD.map(action => {(action.page_id, 1L)}).reduceByKey(_ + _).collect().toMapval sessionRDD: RDD[(String, Iterable[UserVisitAction])] = actionDataRDD.groupBy(_.session_id)val mvRDD: RDD[(String, List[((Long, Long), Int)])] = sessionRDD.mapValues(iter => {val sortList: List[UserVisitAction] = iter.toList.sortBy(_.action_time)val flowIds: List[Long] = sortList.map(_.page_id)val pageflowIds: List[(Long, Long)] = flowIds.zip(flowIds.tail)pageflowIds.map(t => {(t, 1)})})val flatRDD: RDD[((Long, Long), Int)] = mvRDD.map(_._2).flatMap(list => list)val dataRDD: RDD[((Long, Long), Int)] = flatRDD.reduceByKey(_ + _)dataRDD.foreach {case ((pageid1, pageid2), sum) => {val lon: Long = pageidToCountMap.getOrElse(pageid1, 0L)println(s"页面${pageid1}跳转到页面${pageid2}单跳转换率为:"+(sum.toDouble/lon))}}sc.stop()}case class UserVisitAction(date: String,//用户点击行为的日期user_id: Long,//用户的 IDsession_id: String,//Session 的 IDpage_id: Long,//某个页面的 IDaction_time: String,//动作的时间点search_keyword: String,//用户搜索的关键词click_category_id: Long,//某一个商品品类的 IDclick_product_id: Long,//某一个商品的 IDorder_category_ids: String,//一次订单中所有品类的 ID 集合order_product_ids: String,//一次订单中所有商品的 ID 集合pay_category_ids: String,//一次支付中所有品类的 ID 集合pay_product_ids: String,//一次支付中所有商品的 ID 集合city_id: Long)//城市 id
}

总结

这次的spark实例就到这里