RDD的方法

方法	介绍	简单使用
flatmap	对RDD中的每一个元素进行先map再压扁,最后返回操作的结果	scala> sc.parallelize(Array("a b c", "d e f", "h i j")).collect res31: Array[String] = Array(a b c, d e f, h i j) scala> sc.parallelize(Array("a b c", "d e f", "h i j")).flatMap(_.split(" ")).collect res32: Array[String] = Array(a, b, c, d, e, f, h, i, j)
sortBy	sortBy(x=>x,true) //默认升序 sortBy(x=>x+"",true)//变成了字符串,结果为字典顺序	scala> sc.parallelize(List(5, 6, 4, 7, 3, 8, 2, 9, 1, 10)).sortBy(x=>x,true).collect res33: Array[Int] = Array(1, 2, 3, 4, 5, 6, 7, 8, 9, 10) scala> sc.parallelize(List(5, 6, 4, 7, 3, 8, 2, 9, 1, 10)).sortBy(x=>x+"",true).collect res34: Array[Int] = Array(1, 10, 2, 3, 4, 5, 6, 7, 8, 9)
union	并集
intersection	交集
subtract	差集
cartesian	笛卡尔积
join	join(内连接)聚合具有相同key组成的value元组	val rdd1 = sc.parallelize(List(("tom", 1), ("jerry", 2), ("kitty", 3))) val rdd2 = sc.parallelize(List(("jerry", 9), ("tom", 8), ("shuke", 7), ("tom", 2)))   scala> rdd1.join(rdd2).collect res39: Array[(String, (Int, Int))] = Array((tom,(1,8)), (tom,(1,2)), (jerry,(2,9)))
rightOuterJoin		scala>  rdd1.rightOuterJoin(rdd2).collect res43: Array[(String, (Option[Int], Int))] = Array((tom,(Some(1),8)), (tom,(Some(1),2)), (jerry,(Some(2),9)), (shuke,(None,7)))
groupbykey	groupByKey()的功能是，对具有相同键的值进行分组	scala> val rdd6 = sc.parallelize(Array(("tom",1), ("jerry",2), ("kitty",3),      | ("jerry",9), ("tom",8), ("shuke",7), ("tom",2))) rdd6: org.apache.spark.rdd.RDD[(String, Int)] = ParallelCollectionRDD[109] at parallelize at <console>:24 scala> rdd6.groupByKey.collect res44: Array[(String, Iterable[Int])] = Array((tom,CompactBuffer(8, 2, 1)), (jerry,CompactBuffer(9, 2)), (shuke,CompactBuffer(7)), (kitty,CompactBuffer(3)))
cogroup	先在RDD内部按照key分组,再在多个RDD间按照key分组	val rdd1 = sc.parallelize(List(("tom", 1), ("tom", 2), ("jerry", 3), ("kitty", 2))) val rdd2 = sc.parallelize(List(("jerry", 2), ("tom", 1), ("shuke", 2))) val rdd3 = rdd1.cogroup(rdd2)     scala> rdd1.cogroup(rdd2).collect res47: Array[(String, (Iterable[Int], Iterable[Int]))] = Array((tom,(CompactBuffer(1, 2),CompactBuffer(1))), (jerry,(CompactBuffer(3),CompactBuffer(2))), (shuke,(CompactBuffer(),CompactBuffer(2))), (kitty,(CompactBuffer(2),CompactBuffer())))
groupBy	根据指定的函数中的规则/key进行分组	scala> val intRdd = sc.parallelize(List(1,2,3,4,5,6)) intRdd: org.apache.spark.rdd.RDD[Int] = ParallelCollectionRDD[117] at parallelize at <console>:24 scala> intRdd.groupBy(x=>{if(x%2==0) "even" else "odd"}).collect res48: Array[(String, Iterable[Int])] = Array((even,CompactBuffer(4, 6, 2)), (odd,CompactBuffer(1, 3, 5)))
reduce	注意reduce是Action算子	val rdd1 = sc.parallelize(List(1, 2, 3, 4, 5)) //reduce聚合 val result = rdd1.reduce(_ + _) //第一_ 上次一个运算的结果,第二个_ 这一 次进来的元素
reducebykey	注意reducebykey是转换算子
repartition	改变分区数: 注意： repartition可以增加和减少rdd中的分区数， coalesce默认减少rdd分区数，增加rdd分区数不会生效。 不管增加还是减少分区数原rdd分区数不变	scala> val rdd1 = sc.parallelize(1 to 10,3) rdd1: org.apache.spark.rdd.RDD[Int] = ParallelCollectionRDD[124] at parallelize at <console>:24 scala> rdd1.repartition(2).partitions.length res52: Int = 2 scala> rdd1.partitions.length res53: Int = 3
collect	显示数据
count	求RDD中最外层元素的个数	scala> val rdd3 = sc.parallelize(List(List("a b c", "a b b"),List("e f g", "a f g"), List("h i j", "a a b"))) rdd3: org.apache.spark.rdd.RDD[List[String]] = ParallelCollectionRDD[129] at parallelize at <console>:24 scala> rdd3.count res54: Long = 3
distinct	去重	scala> val rdd = sc.parallelize(Array(1,2,3,4,5,5,6,7,8,1,2,3,4), 3) rdd: org.apache.spark.rdd.RDD[Int] = ParallelCollectionRDD[130] at parallelize at <console>:24 scala> rdd.distinct.collect res55: Array[Int] = Array(6, 3, 4, 1, 7, 8, 5, 2)
top	取出最大的前N个	scala>  sc.parallelize(List(3,6,1,2,4,5)).top(2) res56: Array[Int] = Array(6, 5)
take	//按照原来的顺序取前N个	scala>  sc.parallelize(List(3,6,1,2,4,5)).take(2) res57: Array[Int] = Array(3, 6)
first	按照原来的顺序取前第一个	scala>  sc.parallelize(List(3,6,1,2,4,5)).first res58: Int = 3