[ 
https://issues.apache.org/jira/browse/SPARK-40624?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17612183#comment-17612183
 ] 

Hyukjin Kwon commented on SPARK-40624:
--------------------------------------

Does it work if you set {{spark.sql.ansi.enabled}} on?

> A DECIMAL value with division by 0 errors in DataFrame but evaluates to NULL 
> in SparkSQL
> ----------------------------------------------------------------------------------------
>
>                 Key: SPARK-40624
>                 URL: https://issues.apache.org/jira/browse/SPARK-40624
>             Project: Spark
>          Issue Type: Bug
>          Components: Spark Shell
>    Affects Versions: 3.2.1
>            Reporter: xsys
>            Priority: Major
>
> h3. Describe the bug
> Storing an invalid value (e.g. {{{}BigDecimal("1.0/0"){}}}) via 
> {{spark-shell}} errors out during RDD creation. However, {{1.0/0}} evaluates 
> to {{NULL}} if the value is inserted into a {{DECIMAL(20,10)}} column of a 
> table via {{{}spark-sql{}}}.
> h3. To Reproduce
> On Spark 3.2.1 (commit {{{}4f25b3f712{}}}), using {{{}spark-sql{}}}:
> {code:java}
> $SPARK_HOME/bin/spark-sql{code}
> Execute the following: (evaluated to {{{}NULL{}}})
> {code:java}
> spark-sql> create table decimal_vals(c1 DECIMAL(20,10)) stored as ORC;
> spark-sql> insert into decimal_vals select 1.0/0;
> spark-sql> select * from decimal_vals;
> NULL{code}
> Using {{{}spark-shell{}}}:
> {code:java}
> $SPARK_HOME/bin/spark-shell{code}
> Execute the following: (errors out during RDD creation)
> {code:java}
> scala> val rdd = sc.parallelize(Seq(Row(BigDecimal("1.0/0"))))
> java.lang.NumberFormatException
>   at java.math.BigDecimal.<init>(BigDecimal.java:497)
>   at java.math.BigDecimal.<init>(BigDecimal.java:383)
>   at java.math.BigDecimal.<init>(BigDecimal.java:809)
>   at scala.math.BigDecimal$.exact(BigDecimal.scala:126)
>   at scala.math.BigDecimal$.apply(BigDecimal.scala:284)
>   ... 49 elided{code}
> h3. Expected behavior
> We expect the two Spark interfaces ({{{}spark-sql{}}} & {{{}spark-shell{}}}) 
> to behave consistently for the same data type & input combination 
> ({{{}BigDecimal{}}}/{{{}DECIMAL(20,10){}}} and {{{}1.0/0{}}}).
>  



--
This message was sent by Atlassian Jira
(v8.20.10#820010)

---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]

Reply via email to