基于spark中的join创建新的二进制列

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

// 创建SparkSession
val spark = SparkSession.builder()
  .appName("Join Example")
  .getOrCreate()

// 创建两个数据集
val dfA = spark.createDataFrame(Seq(
  (1, "A"),
  (2, "B"),
  (3, "C")
)).toDF("id", "value")

val dfB = spark.createDataFrame(Seq(
  (1, "X"),
  (2, "Y"),
  (4, "Z")
)).toDF("id", "value")

// 使用join操作连接两个数据集
val joinedDF = dfA.join(dfB, Seq("id"), "inner")

// 创建新的二进制列
val resultDF = joinedDF.withColumn("binary_column", when(col("value") === "X" || col("value") === "Y", 1).otherwise(0))

// 显示结果
resultDF.show()