Spark sortByKey功能
按照Key进行排序
案例演示
需求:对val rdd: RDD[(Int, String)] =sc.makeRDD(List((3, “aa”), (6, “cc”), (2, “bb”), (1, “dd”)))按key进行升序和降序排序,然后输出到控制台,形式如下
(1,dd)(2,bb)(3,aa)(6,cc)
**************************************
(6,cc)(3,aa)(2,bb)(1,dd)
package com.xcu.bigdata.spark.core.pg02_rdd.pg022_rdd_transform
import org.apache.spark.rdd.RDD
import org.apache.spark.{SparkConf, SparkContext}
/**
* @Desc : 按照Key进行排序
*/
object Spark18_SortByKey {
def main(args: Array[String]): Unit = {
//创建配置文件
val conf: SparkConf = new SparkConf().setAppName("Spark18_SortByKey").setMaster("local[*]")
//创建SparkContext,该对象是提交的入口
val sc = new SparkContext(conf)
//创建RDD
val rdd: RDD[(Int, String)] = sc.makeRDD(List((3, "aa"), (6, "cc"), (2, "bb"), (1, "dd")))
//按照key对rdd中的元素进行排序 默认升序
val resRDD1: RDD[(Int, String)] = rdd.sortByKey()
//打印输出
resRDD1.collect().foreach(println)
println("**************************************")
//降序
val resRDD2: RDD[(Int, String)] = rdd.sortByKey(ascending = false)
//打印输出
resRDD2.collect().foreach(println)
//释放资源
sc.stop()
}
}
按照value进行排序
map进行结构转换,对调原来的k,v的位置
伪代码:
map(_.swap)
排完序后再使用map进行对调
自定义排序规则对对象进行排序
package com.xcu.bigdata.spark.core.pg02_rdd.pg022_rdd_transform
import org.apache.spark.rdd.RDD
import org.apache.spark.{SparkConf, SparkContext}
/**
* @Desc : 自定义排序规则
*/
object Spark18_CustomSortByKey {
def main(args: Array[String]): Unit = {
//创建配置文件
val conf: SparkConf = new SparkConf().setAppName("Spark18_CustomSortByKey").setMaster("local[*]")
//创建SparkContext,该对象是提交的入口
val sc = new SparkContext(conf)
//如果key为自定义类型,要求必须混入Ordered特质
//构造kv类型的数据
val stdList: List[(Student, Int)] = List(
(new Student("zs", 18), 1),
(new Student("li", 18), 1),
(new Student("zs", 19), 1),
(new Student("wangwu", 18), 1),
(new Student("zs", 20), 1)
)
//创建
val stdRDD: RDD[(Student, Int)] = sc.makeRDD(stdList)
val resRDD: RDD[(Student, Int)] = stdRDD.sortByKey()
//打印输出
resRDD.collect().foreach(println)
//释放资源
sc.stop()
}
}
class Student(var name: String, var age: Int) extends Ordered[Student] with Serializable {
//自定义比较规则
override def compare(that: Student): Int = {
//先按照名称升序排序,如果名称相同的话,在按照年龄降序排序
var res: Int = this.name.compareTo(that.name)
if (res == 0) {
res = that.age - this.age
}
res
}
override def toString: String = s"Student($name, $age)"
}
输出:
(Student(li, 18),1)
(Student(wangwu, 18),1)
(Student(zs, 20),1)
(Student(zs, 19),1)
(Student(zs, 18),1)