Re: [Jung-support] Does PersonalizedPageRank class have bug?

[Joshua O'M. <jos...@gm...>]

Parisa:

(0) StackOverflow, these days, is a better place to post questions like
this.

(1) The name of the class is "PageRankWithPriors".  Using the correct class
name in your questions will speed things up, generally.  :)

(2) I'm not sure why you're calling setHyperEdgesAreSelfLoops()--you're not
working with a hypergraph.  But the fact that the resultant numbers are
weird suggests that yes, there might be a bug there.  (It looks like we're
not guarding against the possibility that the graph is not a hypergraph
when that method is called.  :P )

(3) I think that this may be the problem:

  g.addVertex((Integer)0);

I believe that casting 0 (or any other integer) to Integer is not the same
thing as 'new Integer(0)' or 'Integer.valueOf(0)'.  What you're doing is
actually casting the number 0 to a Integer reference.  So I'll bet that if
you ask your graph how many vertices it has, it will report '6' rather than
'3'.

The good news is that addEdge() will automatically add the vertex arguments
to the graph as needed, so you can just remove the addVertex() calls
entirely.

Let us know if that works for you.

Good luck!

Joshua O'Madadhain



On Mon, Oct 14, 2013 at 1:00 AM, parisa berangi <par...@gm...>wrote:

> Hi ,
> I am using JUNG library for its "PersonalizedPageRank class". I used a
> very small graph (with just 3 nodes) to test this class. when there isn't
> any node with self-edge the output is OK,  but if the graph contains a node
> with a self-edge the output becomes incorrect (the final rank of nodes is
> not in the form of probability value and their sum is not equal to 1). the
> output is :
> 0.2997601918465228
> 0.1247002398081535
> 0.1918465227817746
>
> i used "setHyperedgesAreSelfLoops(true)" but the outputs are still
> incorrect:
> 320751.99531359226
> 609574.2619040733
> 2554381.251484884
>
> my code is attached to this email.
> why "PersonalizedPageRank class" does  not work correctly? Is there a bug
> inside it or i do something wrong?
>
> your help is greatly appreciated.
>
> --
> Parisa Berangi
> M.Sc. Student, Natural Language Processing Lab,
> School of Electrical and Computer Engineering
> University Of Tehran
> Tehran ,Iran
>
>
> ------------------------------------------------------------------------------
> October Webinars: Code for Performance
> Free Intel webinars can help you accelerate application performance.
> Explore tips for MPI, OpenMP, advanced profiling, and more. Get the most
> from
> the latest Intel processors and coprocessors. See abstracts and register >
> http://pubads.g.doubleclick.net/gampad/clk?id=60134071&iu=/4140/ostg.clktrk
> _______________________________________________
> Jung-support mailing list
> Jun...@li...
> https://lists.sourceforge.net/lists/listinfo/jung-support
>
>


-- 
jos...@gm.........sites.google.com/site/joshuaomadadhain/
   Joshua O'Madadhain: Information Scientist, Musician, Philosopher-At-Tall
It's that moment of dawning comprehension that I live for.  -- Bill
Watterson
 My opinions are too rational and insightful to be those of any
organization.