value toDF 不是 org.apache.spark.rdd.RDD 的成员

Question

我已经在其他SO帖子中读到了这个问题，但我仍然不知道我做错了什么。原则上，添加以下两行：

val sqlContext = new org.apache.spark.sql.SQLContext(sc)
import sqlContext.implicits._

应该已经完成了这个技巧，但错误仍然存在

这是我的构建.sbt：

name := "PickACustomer"
version := "1.0"
scalaVersion := "2.11.7"

libraryDependencies ++= Seq("com.databricks" %% "spark-avro" % "2.0.1",
"org.apache.spark" %% "spark-sql" % "1.6.0",
"org.apache.spark" %% "spark-core" % "1.6.0")

我的 scala 代码是：

import scala.collection.mutable.Map
import scala.collection.immutable.Vector
import org.apache.spark.rdd.RDD
import org.apache.spark.SparkContext
import org.apache.spark.SparkContext._
import org.apache.spark.SparkConf
import org.apache.spark.sql._

    object Foo{
    def reshuffle_rdd(rawText: RDD[String]): RDD[Map[String, (Vector[(Double, Double, String)], Map[String, Double])]]  = {...}
    def do_prediction(shuffled:RDD[Map[String, (Vector[(Double, Double, String)], Map[String, Double])]], prediction:(Vector[(Double, Double, String)] => Map[String, Double]) ) : RDD[Map[String, Double]] = {...}
    def get_match_rate_from_results(results : RDD[Map[String, Double]]) : Map[String, Double]  = {...}

    def retrieve_duid(element: Map[String,(Vector[(Double, Double, String)], Map[String,Double])]): Double = {...}


    def main(args: Array[String]){
        val conf = new SparkConf().setAppName(this.getClass.getSimpleName)
        if (!conf.getOption("spark.master").isDefined) conf.setMaster("local")
        val sc = new SparkContext(conf)
        //This should do the trick
        val sqlContext = new org.apache.spark.sql.SQLContext(sc)
        import sqlContext.implicits._
        val PATH_FILE = "/mnt/fast_export_file_clean.csv"
        val rawText = sc.textFile(PATH_FILE)
        val shuffled = reshuffle_rdd(rawText)
        // PREDICT AS A FUNCTION OF THE LAST SEEN UID
        val results = do_prediction(shuffled.filter(x => retrieve_duid(x) > 1) , predict_as_last_uid)
        results.cache()
        case class Summary(ismatch: Double, t_to_last:Double, nflips:Double,d_uid: Double, truth:Double, guess:Double)
        val summary = results.map(x => Summary(x("match"), x("t_to_last"), x("nflips"), x("d_uid"), x("truth"), x("guess")))

        //PROBLEMATIC LINE
        val sum_df = summary.toDF()
    }
    }

我总是得到：

value toDF 不是 org.apache.spark.rdd.RDD 的成员[摘要]

现在有点迷路了。有什么想法吗？

Answer 1

将案例类移出main：

object Foo {
  case class Summary(ismatch: Double, t_to_last:Double, nflips:Double,d_uid: Double, truth:Double, guess:Double)
  def main(args: Array[String]){
    ...
  }
}

关于它的作用域的一些事情阻止了Spark能够处理Summary模式的自动派生。仅供参考，我实际上收到了与sbt不同的错误：

没有可用于摘要的类型标记

Answer 2

太好了。 救救我一命

将您的案例类移到主类之外：

object Foo {
    case class Summary(ismatch: Double, t_to_last:Double, nflips:Double,d_uid: Double, truth:Double, guess:Double)
    def main(args: Array[String]){
...
    }
}

Answer 3

将案例类移到函数体之外。然后使用 import spark.implicits._ .