You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@madlib.apache.org by "Frank McQuillan (JIRA)" <ji...@apache.org> on 2017/03/15 17:24:41 UTC

[jira] [Commented] (MADLIB-1070) Graph - visualization

    [ https://issues.apache.org/jira/browse/MADLIB-1070?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15926596#comment-15926596 ] 

Frank McQuillan commented on MADLIB-1070:
-----------------------------------------

After looking into this a bit:

* like the measures Gephi has but you have to load the data into Gephi to get those measures, so that is restrictive, due to the size limitations of Gephi
* also, you can’t loop in Gephi so you need to load one (sub)graph at a time and get the measures you need, which is time consuming
* all this to say that MADlib-Gephi integration does not seem to be a requirement from a visualization perspective.   Plus, many real-world graphs are too big to visualize anyways

Removing from 1.11 list for now


> Graph - visualization
> ---------------------
>
>                 Key: MADLIB-1070
>                 URL: https://issues.apache.org/jira/browse/MADLIB-1070
>             Project: Apache MADlib
>          Issue Type: New Feature
>          Components: Module: Graph
>            Reporter: Frank McQuillan
>
> Story 
> As a data scientist, I want to export a graph in a form suitable for visualization, so that it makes my analysis easier and I can communicate findings more easily to others.
> This is also useful for MADlib demos.
> Export format TBD. Pivotal data science team likes Gephi [1].  Another option is GraphViz [2] which we support for decision trees.
> References
> [1] Gephi
> https://gephi.org/
> [2] GraphViz
> http://www.graphviz.org/



--
This message was sent by Atlassian JIRA
(v6.3.15#6346)