Spark本身的实现中位数不能用于groupBy的agg函数，下面代码实现在agg中调用

原生Spark计算中位数

df2.stat.approxQuantile("value", Array(0.5), 0)

1.定义方法

import org.apache.spark.sql.functions._
import org.apache.spark.sql.Column
import org.apache.spark.sql.catalyst.expressions.aggregate.ApproximatePercentile


object PercentileApprox {
  def percentile_approx(col: Column, percentage: Column, accuracy: Column): Column = {
    val expr = new ApproximatePercentile(
      col.expr,  percentage.expr, accuracy.expr
    ).toAggregateExpression
    new Column(expr)
  }
  def percentile_approx(col: Column, percentage: Column): Column = percentile_approx(
    col, percentage, lit(ApproximatePercentile.DEFAULT_PERCENTILE_ACCURACY)
  )
}

2.#调用方法

object AlgorithmTest {




  def main(args: Array[String]): Unit = {
    //System.setProperty("hadoop.home.dir", "D:\\hadoop-common-2.2.0-bin-master")

    val spark = SparkSession
      .builder
      .appName("InterfaceMonitor")
      .master("local[2]")
      .getOrCreate()

    import spark.implicits._


      val ds = Seq(
        ("20181102221610c07vy","10000011","10000032",20.0,1,20.0,0 ,"2019-04-19 22:16:10.0"),
        ("20181102221733dgvcv","10000011","10000032",20.0,1,20.0,0 ,"2019-04-19 22:17:34.0"),
        ("20181102222339oakpn","10000061","10000032",0.2 ,1,0.2 ,5 ,"2019-04-19 22:23:39.0"),
        ("20181102225503nhath","10000061","10000032",20.0,1,20.0,7 ,"2019-04-19 22:55:03.0"),
        ("201811030008236k9yy","10000061","10000032",0.2 ,1,0.2 ,5 ,"2019-04-19 00:08:23.0"),
        ("20181103005135do5zg","10000069","10000015",0.2 ,1,0.2 ,0 ,"2019-04-19 00:51:35.0"),
        ("20181103005148ptr7a","10000069","10000015",0.2 ,1,0.2 ,0 ,"2019-04-19 00:51:48.0"),
        ("20181103005148w9isk","10000069","10000015",0.2 ,1,0.2 ,5 ,"2019-04-19 00:51:48.0"),
        ("20181103005205b8gvm","10000069","10000015",0.2 ,1,0.2 ,0 ,"2019-04-19 00:52:05.0"),
        ("20181103015930m2cz0","10000011","10000063",30.0,1,30.0,0 ,"2019-04-19 01:59:30.0")
      ).toDS()
        .toDF("order_id","play_uid","god_uid","price","num","order_amount","order_status","create_time")



    val df = spark.read.format("json").json("file:///e:\\a.json")

    ds.groupBy($"god_uid")

      .agg(sum("num") as "total_sum",
        approx_count_distinct("order_id") as "order_num",
        mean("price"),
        percentile_approx($"price", lit(0.5)))
      .show(10)


  }

}

参考文章：https://stackoverflow.com/questions/53548964/how-to-use-approxquantile-by-group