spark-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Joseph K. Bradley (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (SPARK-14817) ML, Graph, R 2.0 QA: Programming guide update and migration guide
Date Thu, 14 Jul 2016 22:25:20 GMT

    [ https://issues.apache.org/jira/browse/SPARK-14817?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15378505#comment-15378505
] 

Joseph K. Bradley commented on SPARK-14817:
-------------------------------------------

The initial PR does need to include 1 major decision: Should the primary MLlib documentation
page be "ml-guide.html" (currently the DataFrame-based API) or "mllib-guide.html" (currently
the RDD-based API).  I propose:
* Spark doc menu bar links to ml-guide.html, which will be the primary doc page.  It will
mimic the current mllib-guide.html page in terms of content: overview, migration guide, etc.
** Move Pipeline description into ml-pipeline.html
** Move tuning into ml-tuning.html
* mllib-guide.html keeps RDD-specific stuff, with a link at the top redirecting people to
ml-guide.html
* ml-guide.html will include a "maintenance mode" announcement about the RDD-based API
* Sidebar remains the same


> ML, Graph, R 2.0 QA: Programming guide update and migration guide
> -----------------------------------------------------------------
>
>                 Key: SPARK-14817
>                 URL: https://issues.apache.org/jira/browse/SPARK-14817
>             Project: Spark
>          Issue Type: Sub-task
>          Components: Documentation, GraphX, ML, MLlib, SparkR
>            Reporter: Joseph K. Bradley
>            Assignee: Joseph K. Bradley
>            Priority: Critical
>
> Before the release, we need to update the MLlib, GraphX, and SparkR Programming Guides.
 Updates will include:
> * Add migration guide subsection.
> ** Use the results of the QA audit JIRAs and [SPARK-13448].
> * Check phrasing, especially in main sections (for outdated items such as "In this release,
...")
> For MLlib, we will make the DataFrame-based API (spark.ml) front-and-center, to make
it clear the RDD-based API is the older, maintenance-mode one.
> * No docs for spark.mllib will be deleted; they will just be reorganized and put in a
subsection.
> * If spark.ml docs are less complete, or if spark.ml docs say "refer to the spark.mllib
docs for details," then we should copy those details to the spark.ml docs.  This per-feature
work can happen under [SPARK-14815].
> * This big reorganization should be done *after* docs are added for each feature (to
minimize merge conflicts).



--
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