将月份添加到Spark dataframe中的日期列

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

val extractMonth = udf((date: java.sql.Date) => {
  val cal = java.util.Calendar.getInstance()
  cal.setTime(date)
  cal.get(java.util.Calendar.MONTH) + 1
})

val dfWithDate = df.withColumn("date_column", col("date_column").cast(DateType))

val dfWithMonth = dfWithDate.withColumn("month", extractMonth(col("date_column")))

dfWithMonth.show()