如何在Spark中使用Kryo序列化程序缓存DataFrame？

import org.apache.spark.sql.functions._
import org.apache.spark.sql.DataFrame

// Step 1: Configure Spark and Kryo serializer
spark.conf.set("spark.serializer", "org.apache.spark.serializer.KryoSerializer")
spark.conf.set("spark.kryo.registrator", "com.example.MyKryoRegistrator")

// Step 2: Create a custom KryoRegistrator
class MyKryoRegistrator extends KryoRegistrator {
  override def registerClasses(kryo: Kryo): Unit = {
    // Register your custom classes here
    kryo.register(classOf[MyCustomClass])
  }
}

// Step 3: Import necessary classes
import spark.implicits._
import org.apache.spark.sql.functions._
import org.apache.spark.sql.DataFrame

// Step 4: Create a SparkSession
val sparkSession = SparkSession.builder().appName("MyApp").getOrCreate()

// Step 5: Read data source and get a DataFrame
val df = sparkSession.read.format("csv").load("path/to/data.csv")

// Step 6: Broadcast the DataFrame
val broadcastDf = sparkSession.sparkContext.broadcast(df)

// Step 7: Use the broadcasted DataFrame
val result = sparkSession.sparkContext.parallelize(Seq(1, 2, 3)).mapPartitions { iter =>
  val df = broadcastDf.value
  iter.map { i =>
    // Perform operations on the DataFrame
    df.filter($"column" === i).count()
  }
}.collect()