[
https://issues.apache.org/jira/browse/SPARK-17668?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15802399#comment-15802399
]
Mathieu D commented on SPARK-17668:
-----------------------------------
I experience the same issue with a custom Aggregator having a case class as an
output type. The resulting aggregate is a Row, that I can't cast back to my
case class.
I provided an outputEncoder, using Encoders.product[T].
> Support representing structs with case classes and tuples in spark sql udf
> inputs
> ---------------------------------------------------------------------------------
>
> Key: SPARK-17668
> URL: https://issues.apache.org/jira/browse/SPARK-17668
> Project: Spark
> Issue Type: New Feature
> Components: SQL
> Affects Versions: 2.0.0
> Reporter: koert kuipers
> Priority: Minor
>
> after having gotten used to have case classes represent complex structures in
> Datasets, i am surprised to find out that when i work in DataFrames with udfs
> no such magic exists, and i have to fall back to manipulating Row objects,
> which is error prone and somewhat ugly.
> for example:
> {noformat}
> case class Person(name: String, age: Int)
> val df = Seq((Person("john", 33), 5), (Person("mike", 30), 6)).toDF("person",
> "id")
> val df1 = df.withColumn("person", udf({ (p: Person) => p.copy(age = p.age +
> 1) }).apply(col("person")))
> df1.printSchema
> df1.show
> {noformat}
> leads to:
> {noformat}
> java.lang.ClassCastException:
> org.apache.spark.sql.catalyst.expressions.GenericRowWithSchema cannot be cast
> to Person
> {noformat}
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
