如何将DataFrame转换为RDD [Point]而不是RDD [ROW]?

时间:2019-04-19 11:03:41

标签: scala apache-spark magellan

我有一个数据列,其中包含许多列,这些列是根据定义模式的csv文件创建的。我唯一感兴趣的列是“ Point”列,其中定义了magellan Point(long,lat)。 我现在需要做的是根据该数据框创建一个RDD [Point]。

下面是我尝试过的代码,但是由于rdd是RDD [Row]而不是RDD [Point],所以它不起作用。

val schema = StructType(Array(
         StructField("vendorId", StringType, false),
         StructField("lpep_pickup_datetime", StringType, false),
         StructField("Lpep_dropoff_datetime", StringType, false),
         StructField("Store_and_fwd_flag",StringType, false),
         StructField("RateCodeID", IntegerType, false),
         StructField("Pickup_longitude", DoubleType, false),
         StructField("Pickup_latitude", DoubleType, false),
         StructField("Dropoff_longitude", DoubleType, false),
         StructField("Dropoff_latitude", DoubleType, false),
         StructField("Passenger_count", IntegerType, false),
         StructField("Trip_distance", DoubleType, false),
         StructField("Fare_amount", StringType, false),
         StructField("Extra", StringType, false),
         StructField("MTA_tax", StringType, false),
         StructField("Tip_amount", StringType, false),
         StructField("Tolls_amount", StringType, false),
         StructField("Ehail_fee", StringType, false),
         StructField("improvement_surcharge", StringType, false),
         StructField("Total_amount", DoubleType, false),
         StructField("Payment_type", IntegerType, false),
         StructField("Trip_type", IntegerType, false)))

    import spark.implicits._

    val points = spark.read.option("mode", "DROPMALFORMED")
     .schema(schema)
     .csv("/home/riccardo/Scrivania/Progetto/Materiale/NYC-taxi/")
     .withColumn("point", point($"Pickup_longitude",$"Pickup_latitude"))
     .limit(2000)

    val rdd = points.select("point").rdd

如何从数据框中获取RDD [Point]而不是RDD [Row]? 如果不可能,您会建议哪种解决方案?我需要一个RDD [Point]才能与以RDD [Point]作为输入的提供的库一起使用。

2 个答案:

答案 0 :(得分:1)

如果我理解正确,您希望结果为自定义类类型,即为Point而不是Row类型

这是我尝试过的:

我的输入数据示例是:

latitude,longitude
44.968046,-94.420307
44.968046,-94.420307
44.33328,-89.132008
33.755787,-116.359998
33.844843,-116.54911
44.92057,-93.44786
44.240309,-91.493619
44.968041,-94.419696
44.333304,-89.132027

我已经使用toString()

创建了自定义类
case class Pair(latitude: Double, longitude: Double) {
  override def toString: String = s"Pair($latitude, $longitude)"
}

现在,我使用spark作为DataFrame读取输入文件,并将其隐藏到RDD

val df = sparkSession.read.option("inferSchema", "true")
  .option("header", "true")
  .csv("/home/prasadkhode/sample_input.csv")

df.printSchema()
df.show()

val rdd = df.rdd.map(row => {
  Pair(row.getAs[Double]("latitude"), row.getAs[Double]("longitude"))
})

println(s"df count : ${df.count}")
println(s"rdd count : ${rdd.count}")

rdd.take(20).foreach(println)

最后的结果如下:

root
 |-- latitude: double (nullable = true)
 |-- longitude: double (nullable = true)

+---------+-----------+
| latitude|  longitude|
+---------+-----------+
|44.968046| -94.420307|
|44.968046| -94.420307|
| 44.33328| -89.132008|
|33.755787|-116.359998|
|33.844843| -116.54911|
| 44.92057|  -93.44786|
|44.240309| -91.493619|
|44.968041| -94.419696|
|44.333304| -89.132027|
+---------+-----------+

df count : 9
rdd count : 9

Pair(44.968046, -94.420307)
Pair(44.968046, -94.420307)
Pair(44.33328, -89.132008)
Pair(33.755787, -116.359998)
Pair(33.844843, -116.54911)
Pair(44.92057, -93.44786)
Pair(44.240309, -91.493619)
Pair(44.968041, -94.419696)
Pair(44.333304, -89.132027)

希望这对您有帮助...:-)

答案 1 :(得分:0)

“ as”和“ rdd”方法可以提供帮助:

case class Point(latitude: Double, longitude: Double)

val df = Seq((1.0, 2.0)).toDF("Pickup_longitude", "Pickup_latitude")

val rdd = df
  .select(
    $"Pickup_longitude".alias("latitude"),
    $"Pickup_latitude".alias("longitude"))
  .as[Point].rdd

rdd.foreach(println)

输出:

Point(1.0,2.0)