spark UDAF

感谢我的同事 李震给我讲解UDAF

网上找到的大部分都只有代码，但是缺少讲解，官网的的API有讲解，但是看不太明白。我还是自己记录一下吧，或许对其他人有帮助。

接下来以一个求几何平均数的例子来说明如何实现一个自己的UDAF

首先需要导入这些包：

import org.apache.spark.sql.expressions.MutableAggregationBuffer
import org.apache.spark.sql.expressions.UserDefinedAggregateFunction
import org.apache.spark.sql.Row
import org.apache.spark.sql.types._

需要继承实现这个抽象类
class GeometricMean extends UserDefinedAggregateFunction {
  // This is the input fields for your aggregate function.
  就是需要输入的列的类型，可以有多个列，多个列的写法如下：
/*

StructType(StructField("slot",IntegerType) :: StructField("score",IntegerType)::Nil)

*/
  override def inputSchema: org.apache.spark.sql.types.StructType =
    StructType(StructField("value", DoubleType) :: Nil)

  存储聚合结果的中间buffer
  // This is the internal fields you keep for computing your aggregate.
  override def bufferSchema: StructType = StructType(
    StructField("count", LongType) ::
    StructField("product", DoubleType) :: Nil
  )

  // This is the output type of your aggregatation function.
  返回结果的类型，比如这个集合平均数就是返回一个double值
  override def dataType: DataType = DoubleType

  是每次运行结果都过一样，但是我也不太明白啊
  override def deterministic: Boolean = true

   初始化存储聚合结果的buffer
  // This is the initial value for your buffer schema.
  override def initialize(buffer: MutableAggregationBuffer): Unit = {
    buffer(0) = 0L
    buffer(1) = 1.0
  }

   每次更新怎么更新，比如新来了一行，如何加入更新聚合的结果
  // This is how to update your buffer schema given an input.
  override def update(buffer: MutableAggregationBuffer, input: Row): Unit = {
    buffer(0) = buffer.getAs[Long](0) + 1
    buffer(1) = buffer.getAs[Double](1) * input.getAs[Double](0)
  }

   spark会把数据划分成多个块，每个块都会进行处理，然后把每个块的结果进行合并处理
  // This is how to merge two objects with the bufferSchema type.
  override def merge(buffer1: MutableAggregationBuffer, buffer2: Row): Unit = {
    buffer1(0) = buffer1.getAs[Long](0) + buffer2.getAs[Long](0)
    buffer1(1) = buffer1.getAs[Double](1) * buffer2.getAs[Double](1)
  }

  返回结果
  // This is where you output the final value, given the final value of your bufferSchema.
  override def evaluate(buffer: Row): Any = {
    math.pow(buffer.getDouble(1), 1.toDouble / buffer.getLong(0))
  }
}

使用方法：

先注册

sqlContext.udf.register("gm", new GeometricMean)

使用自定义的UDAF

%sql
-- Use a group_by statement and call the UDAF.
select group_id, gm(id) from simple group by group_id

参考资料：

https://docs.databricks.com/spark/latest/spark-sql/udaf-scala.html