You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@hama.apache.org by "Edward J. Yoon" <ed...@apache.org> on 2009/09/19 09:36:23 UTC

Re: question about hamburg example

Hi,

Thanks for your review. We'll check it out. FYI, This community will
be moved to apache Hama project.

On Fri, Sep 18, 2009 at 11:20 PM, Ilja Livenson <il...@gmail.com> wrote:
>
> Oops :) As it sometimes happens the problem got solved right after
> sending out the mail.
>
> The problem was that one of the neighbors was larger than the number
> of vertices.
>
> Sorry for disturbance! Will continue digging your great project!
>
> Ilja
>
> 2009/9/18 Ilja Livenson <il...@gmail.com>:
>> Hi, guys
>>
>> I'm playing with hamburg framework, but I've got a very stupid problem
>> with the input format. I saw an example data in this list and from
>> what I understand the expected graph file format is essentially:
>> for i in (number of vertices):
>>    i neighbour_of_i_1 neighbour_of_i_2 ...
>>
>> I wrote a small generator, but hamburg fails to process it with these
>> kind of errors:
>> ...
>> == [Thread-0] 0 round traverse begin ==
>> == [Thread-2] 0 round traverse begin ==
>> == [Thread-1] 0 round traverse begin ==
>> [Thread-2] Enter the sync barrier
>> Exception in thread "Thread-0" Exception in thread "Thread-1"
>> java.lang.IndexOutOfBoundsException: Index: 3, Size: 3
>>        at java.util.ArrayList.RangeCheck(ArrayList.java:547)
>>        at java.util.ArrayList.get(ArrayList.java:322)
>>        at org.apache.hadoop.hamburg.examples.ShortestPath$BFSearcher.synchronize(ShortestPath.java:129)
>>        at org.apache.hadoop.hamburg.Walker.run(Walker.java:107)
>>
>> I'm a bit confused - is it a format error? Or is it rather that the
>> structure of graph is "unacceptable"? The graph is connected, but some
>> vertices might not be reachable.
>>
>> Ilja
>>
>



-- 
Best Regards, Edward J. Yoon @ NHN, corp.
edwardyoon@apache.org
http://blog.udanax.org