[jira] [Comment Edited] (SPARK-21160) Filtering rows with "not equal" operator yields unexpected result with null rows

["Hyukjin Kwon (JIRA)" <ji...@apache.org>]

    [ https://issues.apache.org/jira/browse/SPARK-21160?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16058537#comment-16058537 ] 

Hyukjin Kwon edited comment on SPARK-21160 at 6/22/17 12:40 AM:
----------------------------------------------------------------

There is null-safe equality comparison

{code}
scala> Seq(Some(1), Some(2), None).toDF("a").where("a != 1").show()
+---+
|  a|
+---+
|  2|
+---+


scala> Seq(Some(1), Some(2), None).toDF("a").where("not(a <=> 1)").show()
+----+
|   a|
+----+
|   2|
|null|
+----+
{code}

I am resolving this. Issuing warnings will mess the logs and I guess what you tested in RDB and R does not produce such warnings as well as references.


was (Author: hyukjin.kwon):
There is null-safe equality comparison

```
scala> Seq(Some(1), Some(2), None).toDF("a").where("a != 1").show()
+---+
|  a|
+---+
|  2|
+---+


scala> Seq(Some(1), Some(2), None).toDF("a").where("not(a <=> 1)").show()
+----+
|   a|
+----+
|   2|
|null|
+----+
```

I am resolving this. Issuing warnings will mess the logs and I guess what you tested in RDB and R does not produce such warnings as well as references.

> Filtering rows with "not equal" operator yields unexpected result with null rows
> --------------------------------------------------------------------------------
>
>                 Key: SPARK-21160
>                 URL: https://issues.apache.org/jira/browse/SPARK-21160
>             Project: Spark
>          Issue Type: Bug
>          Components: PySpark, Spark Core, SQL
>    Affects Versions: 2.0.2
>            Reporter: Edoardo Vivo
>            Priority: Minor
>
> ```
> schema = StructType([StructField("Test", DoubleType())])
> test2 = spark.createDataFrame([[1.0],[1.0],[2.0],[2.0],[None]], schema=schema)
> test2.where("Test != 1").show()
> ```
> This returns only the rows with the value 2, it does not return the null row. This should not be the expected behavior, IMO. 
> Thank you.



--
This message was sent by Atlassian JIRA
(v6.4.14#64029)

---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@spark.apache.org
For additional commands, e-mail: issues-help@spark.apache.org