[
https://issues.apache.org/jira/browse/HAMA-941?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15273213#comment-15273213
]
Behroz Sikander edited comment on HAMA-941 at 5/5/16 10:20 PM:
---------------------------------------------------------------
I then tried to run a simple example against the algorithm. Here is the input
{code}
1 2-1.0,3-1.0
2 1-1.0,3-2.0
3 1-1.0,2-2.0,4-2.0,5-1.0
4 3-2.0,5-1.0
5 3-1.0,4-1.0
{code}
and here is the output
{code}
C2532 = [4, 5, 2, 3, 1]
C2531 = [4, 5, 2, 3, 1]
{code}
which again seems to be wrong. So, my conclusion is that there are more bugs in
semi-clustering algorithm. To figure them out I needed to play with an example
of semi-clustering. So, I configured okapi which is a package on top of Apache
Giraph.
https://github.com/grafos-ml/okapi/blob/master/src/test/java/ml/grafos/okapi/graphs/SemiClusteringTest.java
https://github.com/grafos-ml/okapi/blob/master/src/main/java/ml/grafos/okapi/graphs/SemiClustering.java
After some issues, I was able to run and debug the semi-clustering example.
There algorithm seem to be working fine because I changed different
configuration and output seems to make sense. While debugging thorough the code
of Giraph, I found the ouput of first superstep to be different then my
understanding. Currently, I am trying to make sense of it. Once I am done with
it, I will find the issue in our implementation, update a test case to include
a simple example and apply the fixes that you provided.
was (Author: bsikander):
I then tried to run a simple example against the algorithm. Here is the input
{code}
1 2-1.0,3-1.0
2 1-1.0,3-2.0
3 1-1.0,2-2.0,4-2.0,5-1.0
4 3-2.0,5-1.0
5 3-1.0,4-1.0
{code}
> Semiclustering Termination
> --------------------------
>
> Key: HAMA-941
> URL: https://issues.apache.org/jira/browse/HAMA-941
> Project: Hama
> Issue Type: Improvement
> Components: examples, graph
> Reporter: Edward J. Yoon
> Priority: Minor
>
> Currently Semiclustering example will be terminated when the number of
> iterations exceeded the predefined threshold max iteration.
> App should be stopped if there's no cluster changes (I guess). Please check
> and improve it.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
