spark-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Reynold Xin (JIRA)" <j...@apache.org>
Subject [jira] [Comment Edited] (SPARK-10868) monotonicallyIncreasingId() supports offset for indexing
Date Wed, 17 Aug 2016 23:33:20 GMT

    [ https://issues.apache.org/jira/browse/SPARK-10868?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15425598#comment-15425598
] 

Reynold Xin edited comment on SPARK-10868 at 8/17/16 11:33 PM:
---------------------------------------------------------------

I'm going to close this. I don't think adding an offset will actually address the use case,
due to how monotonically_increasing_id is implemented under the hood.

Also the proposed implementation is equivalent to 

{code}
offset + monotonically_increasing_id()
{code}

-- something the user can do trivially.



was (Author: rxin):
I'm going to close this. I don't think adding an offset will actually satisfy the use case,
due to how monotonically_increasing_id is implemented under the hood.

Also the proposed implementation is equivalent to 

{code}
offset + monotonically_increasing_id()
{code}

-- something the user can do trivially.


> monotonicallyIncreasingId() supports offset for indexing
> --------------------------------------------------------
>
>                 Key: SPARK-10868
>                 URL: https://issues.apache.org/jira/browse/SPARK-10868
>             Project: Spark
>          Issue Type: New Feature
>          Components: SQL
>    Affects Versions: 1.5.0
>            Reporter: Martin Senne
>
> With SPARK-7135 and https://github.com/apache/spark/pull/5709 `monotonicallyIncreasingID()`
allows to create an index column with unique ids. The indexing always starts at 0 (no offset).
> *Feature wish*
> Having a parameter `offset`, such that the function can be used as
>     {{monotonicallyIncreasingID( offset )}}
> and indexing _starts at *offset* instead of 0_.
> *Use-case* 
> Add rows to a DataFrame that is already written to a DB (via _.write.jdbc(...)_).
> In detail:
> - A DataFrame *A* (containing an ID column) and having indices from 0 to 199 in that
column is existent in DB.
> - New rows need to be added to *A*. This included
> -- Creating a DataFrame *A'* with new rows, but without id column
> -- Add the index column to *A'* - this time starting at *200*, as there are already entries
with id's from 0 to 199 (*here, monotonicallyInreasingID( 200 ) is required.*)
> -- union *A* and *A'*
> -- store into DB



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)

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


Mime
View raw message