196、Spark 2.0之Dataset开发详解-untype

代码

object UntypedOperation {

  case class Employee(name: String, age: Long, depId: Long, gender: String, salary: Long)

  case class Department(id: Long, name: String)

  def main(args: Array[String]): Unit = {
    val sparkSession = SparkSession
      .builder()
      .appName("BasicOperation")
      .master("local")
      .getOrCreate()

    import sparkSession.implicits._
    import org.apache.spark.sql.functions._

    val employeePath = this.getClass.getClassLoader.getResource("employee.json").getPath
    val departmentPath = this.getClass.getClassLoader.getResource("department.json").getPath

    val employeeDF = sparkSession.read.json(employeePath)
    val departmentDF = sparkSession.read.json(departmentPath)

    val employeeDS = employeeDF.as[Employee]
    val departmentDS = departmentDF.as[Department]

    employeeDS.where("age > 20")
      .join(departmentDS, $"depId" === $"id")
      .groupBy(employeeDS("depId"))
      .agg(avg(employeeDS("salary")))
      .select("avg(salary)")
      .show()
  }
}