spark自定义udf输入类型为array报错

定义udf如下

val list2string = udf { (style: Array[String], num: Array[Long]) =>
   style.zip(num).map(t => t._1 + ":" + t._2).mkString("<br>")
}

输入为两个数组，输出为string

 

报错如下

Caused by: java.lang.ClassCastException: scala.collection.mutable.WrappedArray$ofRef cannot be cast to [Ljava.lang.String;

 

打印输入字段类型

root

 |-- user_no: long (nullable = true)

 |-- style_list: array (nullable = true)

 |    |-- element: string (containsNull = true)

 |-- styleNum_list: array (nullable = true)

 |    |-- element: long (containsNull = true)

 

解决办法

更改udf输入为seq

val list2string = udf { (style: Seq[String], num: Seq[Long]) =>
    style.zip(num).map(t => t._1 + ":" + t._2).mkString("<br>")
 }

 

转载于:https://www.cnblogs.com/OS-BigData/p/8523607.html