You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@spark.apache.org by "Mick Jermsurawong (JIRA)" <ji...@apache.org> on 2019/07/12 13:47:00 UTC
[jira] [Updated] (SPARK-28369) Check overflow in decimal UDF
[ https://issues.apache.org/jira/browse/SPARK-28369?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Mick Jermsurawong updated SPARK-28369:
--------------------------------------
Description:
Udf resulting in overflowing BigDecimal currently returns null. This is inconsistent with new behavior allow option to check and throw overflow introduced in https://issues.apache.org/jira/browse/SPARK-23179
{code:java}
import spark.implicits._
val tenFold: java.math.BigDecimal => java.math.BigDecimal =
_.multiply(new java.math.BigDecimal("10"))
val tenFoldUdf = udf(tenFold)
val ds = spark
.createDataset(Seq(BigDecimal("12345678901234567890.123")))
.select(tenFoldUdf(col("value")))
.as[BigDecimal]
ds.collect shouldEqual Seq(null){code}
The problem is at the {{CatalystTypeConverters}} where {{toPrecision}} gets converted to null
[https://github.com/apache/spark/blob/13ae9ebb38ba357aeb3f1e3fe497b322dff8eb35/sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/CatalystTypeConverters.scala#L344-L356]
was:
Udf that result in overflowing BigDecimal will return null. This is inconsistent with new behavior allow option to check and throw overflow introduced in https://issues.apache.org/jira/browse/SPARK-23179
{code:java}
import spark.implicits._
val tenFold: java.math.BigDecimal => java.math.BigDecimal =
_.multiply(new java.math.BigDecimal("10"))
val tenFoldUdf = udf(tenFold)
val ds = spark
.createDataset(Seq(BigDecimal("12345678901234567890.123")))
.select(tenFoldUdf(col("value")))
.as[BigDecimal]
ds.collect shouldEqual Seq(null){code}
The problem is at the {{CatalystTypeConverters}} where {{toPrecision}} gets converted to null
https://github.com/apache/spark/blob/13ae9ebb38ba357aeb3f1e3fe497b322dff8eb35/sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/CatalystTypeConverters.scala#L344-L356
> Check overflow in decimal UDF
> -----------------------------
>
> Key: SPARK-28369
> URL: https://issues.apache.org/jira/browse/SPARK-28369
> Project: Spark
> Issue Type: Bug
> Components: SQL
> Affects Versions: 3.0.0
> Reporter: Mick Jermsurawong
> Priority: Minor
>
> Udf resulting in overflowing BigDecimal currently returns null. This is inconsistent with new behavior allow option to check and throw overflow introduced in https://issues.apache.org/jira/browse/SPARK-23179
> {code:java}
> import spark.implicits._
> val tenFold: java.math.BigDecimal => java.math.BigDecimal =
> _.multiply(new java.math.BigDecimal("10"))
> val tenFoldUdf = udf(tenFold)
> val ds = spark
> .createDataset(Seq(BigDecimal("12345678901234567890.123")))
> .select(tenFoldUdf(col("value")))
> .as[BigDecimal]
> ds.collect shouldEqual Seq(null){code}
> The problem is at the {{CatalystTypeConverters}} where {{toPrecision}} gets converted to null
> [https://github.com/apache/spark/blob/13ae9ebb38ba357aeb3f1e3fe497b322dff8eb35/sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/CatalystTypeConverters.scala#L344-L356]
--
This message was sent by Atlassian JIRA
(v7.6.14#76016)
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@spark.apache.org
For additional commands, e-mail: issues-help@spark.apache.org