SparkSql行转列，groupBy算子用法

package bigdata
import java.util

import bean.Row2ColFieldBean
import com.alibaba.fastjson.JSON
import org.apache.spark.sql.{Column, DataFrame, SparkSession}
import utils.SparkUtils

import scala.collection.JavaConversions._


object Row2Column {

  def change(df: DataFrame,spark:SparkSession,json:String):Unit={

    //将json转换为可操作的java对象,字段实体类
    val field: Row2ColFieldBean = JSON.parseObject(json).toJavaObject(classOf[Row2ColFieldBean])
    //scala操作java对象需要导包 import scala.collection.JavaConversions._
    //获取分组字段集合
      val groupFieldList: util.List[String] = field.getGroupField

    //获取转换字段
    val changeField: String = field.getChangeField
    //获取取值字段集合
    val valueField: util.List[String] = field.getValueField

    //行转列，分组、转换、聚合
    //groupBy算子需要col类型参数
    df.groupBy(groupFieldList.map(str => df.col(str.trim)): _*)
      .pivot(changeField)
      .sum(valueField.map(str => str.trim): _*).show()

  }

  def main(args: Array[String]): Unit = {
    val spark: SparkSession = SparkUtils.getSparkSession()
//    val json = "{\n \"groupField\":\"name\",\n  \"changeField\":\"subject\",\n   \"valueField\":\"score\"\n}"
    val json = "{\n \"groupField\":[\"name\",\"id\"],\n  \"changeField\":\"subject\",\n   \"valueField\":[\"score\"]\n}"
    val df: DataFrame = SparkUtils.readMySQL(spark)
    change(df,spark,json:String)
  }
}

package utils

import org.apache.spark.sql.{DataFrame, SparkSession}

object SparkUtils {
  //获取SparkSession
  def getSparkSession():SparkSession={
    SparkSession.builder()
      .appName("Connect")
      .master("local[*]")
      .enableHiveSupport()
      .getOrCreate()
  }
  //获取mysql数据库表
  def readMySQL(spark:SparkSession):DataFrame={
    spark.read.format("jdbc")
      .option("url", "jdbc:mysql://localhost:3306/bigdata")
      .option("user", "root")
      .option("password", "root")
      .option("dbtable", "studentscore")
      .load()
  }
}