一零一二、Spark- RDD-DF-DS 相互转换

输入文件

代码

package example.spark.sql

import org.apache.log4j.{Level, Logger}
import org.apache.spark.SparkContext
import org.apache.spark.rdd.RDD
import org.apache.spark.sql.{DataFrame, Dataset, Row, SparkSession}

object RDD_DF_DS {
  def main(args: Array[String]): Unit = {
    val spark: SparkSession = SparkSession.builder().appName("sparksql").master("local[6]").config("spark.sql.warehouse.dir", "E:/").getOrCreate()
    val sc: SparkContext = spark.sparkContext

    Logger.getLogger("org").setLevel(Level.ERROR)
    val lines: RDD[String] = sc.textFile("data/input/csv")


    //    val personRDD: RDD[(Int,String,Int)] = lines.map(line => {
    val personRDD: RDD[Person] = lines.map(line => {

      val str: Array[String] = line.split(",")
      //      (str(0).toInt, str(1), str(2).toInt)
      Person(str(0).toInt, str(1), str(2).toInt)
    })
    //转换1: RDD->DF
    import spark.implicits._
    val personDF: DataFrame = personRDD.toDF("id", "name", "age")

    //转换2:RDD->DS
    val personDS: Dataset[Person] = personRDD.toDS()

    //转换3:DF->RDD   DF没有泛型
    val rdd: RDD[Row] = personDF.rdd

    //转换4:DS->RDD
    val rdd1: RDD[Person] = personDS.rdd

    //转换5:DF-->DS
    val ds: Dataset[Person] = personDF.as[Person]

    //转换6:DS-->DF
    val df: DataFrame = personDS.toDF()

    personDF.show()
    personDS.show()
    rdd.foreach(println)
    rdd1.foreach(println)

    //关闭资源
    spark.stop()
  }

  case class Person(id: Int, name: String, age: Int)
}

结果打印

+---+--------+---+
| id|    name|age|
+---+--------+---+
|  1|zhangsan| 20|
|  2|    lisi| 29|
|  3|  wangwu| 25|
|  4| zhaoliu| 30|
|  5|  tianqi| 35|
|  6|    kobe| 40|
+---+--------+---+

+---+--------+---+
| id|    name|age|
+---+--------+---+
|  1|zhangsan| 20|
|  2|    lisi| 29|
|  3|  wangwu| 25|
|  4| zhaoliu| 30|
|  5|  tianqi| 35|
|  6|    kobe| 40|
+---+--------+---+

[1,zhangsan,20]
[2,lisi,29]
[4,zhaoliu,30]
[3,wangwu,25]
[5,tianqi,35]
[6,kobe,40]
Person(1,zhangsan,20)
Person(4,zhaoliu,30)
Person(5,tianqi,35)
Person(6,kobe,40)
Person(2,lisi,29)
Person(3,wangwu,25)

1）DF/DS转RDD
Val Rdd = DF/DS.rdd
2) DS/RDD转DF
import spark.implicits._
调用 toDF（就是把一行数据封装成row类型）
3）RDD转DS
将RDD的每一行封装成样例类，再调用toDS方法
4）DF转DS
根据row字段定义样例类，再调用asDS方法[样例类]

特别注意：在使用一些特殊的操作时，一定要加上 import spark.implicits._ 不然toDF、toDS无法使用。