You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@hama.apache.org by "praveen sripati (Created) (JIRA)" <ji...@apache.org> on 2012/04/08 13:28:18 UTC
[jira] [Created] (HAMA-547) Submit the K-Means code after mimicking
the Mahout API for K-Means
Submit the K-Means code after mimicking the Mahout API for K-Means
------------------------------------------------------------------
Key: HAMA-547
URL: https://issues.apache.org/jira/browse/HAMA-547
Project: Hama
Issue Type: Improvement
Components: bsp
Affects Versions: 0.4.0
Reporter: praveen sripati
Priority: Minor
Fix For: 0.5.0, 0.6.0
Submit the K-Means code (1) after mimicking the Mahout API. Existing users of Mahout will be able to use Hama easily and also existing Mahout programs can be ported to Hama with minimal changes.
org.apache.mahout.clustering.kmeans.KMeansClusterer
org.apache.mahout.clustering.kmeans.KMeansDriver
http://codingwiththomas.blogspot.in/2011/12/k-means-clustering-with-bsp-intuition.html
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HAMA-547) Add K-Means to ML package
Posted by "Thomas Jungblut (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HAMA-547?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13432926#comment-13432926 ]
Thomas Jungblut commented on HAMA-547:
--------------------------------------
I have written the patch, but it remains untested. Stay tuned, 2-3 days and you get a patch ;)
bq.how is a input format different from mahout? if possible, let's compare.
I have no idea what the input format of Mahout is. But I can work it out, maybe not before the weekend.
> Add K-Means to ML package
> -------------------------
>
> Key: HAMA-547
> URL: https://issues.apache.org/jira/browse/HAMA-547
> Project: Hama
> Issue Type: New Feature
> Components: machine learning
> Reporter: praveen sripati
> Assignee: Thomas Jungblut
>
> Based on my blog post:
> http://codingwiththomas.blogspot.in/2011/12/k-means-clustering-with-bsp-intuition.html
> https://github.com/thomasjungblut/thomasjungblut-common/blob/master/src/de/jungblut/clustering/KMeansBSP.java
> We need a math lib for that, let's wait until Mikalai finished with HAMA-524.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Updated] (HAMA-547) Add K-Means to ML package
Posted by "Thomas Jungblut (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HAMA-547?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Thomas Jungblut updated HAMA-547:
---------------------------------
Attachment: HAMA-547_2.patch
No idea whats wrong when reading the center sequencefiles...:
{noformat}
Caused by: java.io.IOException: [] read 4 bytes, should read 22
at org.apache.hadoop.io.SequenceFile$Reader.next(SequenceFile.java:1897)
at org.apache.hadoop.io.SequenceFile$Reader.next(SequenceFile.java:1934)
{noformat}
Will fix this soon...
> Add K-Means to ML package
> -------------------------
>
> Key: HAMA-547
> URL: https://issues.apache.org/jira/browse/HAMA-547
> Project: Hama
> Issue Type: New Feature
> Components: machine learning
> Reporter: praveen sripati
> Assignee: Thomas Jungblut
> Attachments: HAMA-547_1.patch, HAMA-547_2.patch
>
>
> Based on my blog post:
> http://codingwiththomas.blogspot.in/2011/12/k-means-clustering-with-bsp-intuition.html
> https://github.com/thomasjungblut/thomasjungblut-common/blob/master/src/de/jungblut/clustering/KMeansBSP.java
> We need a math lib for that, let's wait until Mikalai finished with HAMA-524.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Updated] (HAMA-547) Add K-Means to ML package
Posted by "Thomas Jungblut (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HAMA-547?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Thomas Jungblut updated HAMA-547:
---------------------------------
Attachment: HAMA-547.patch
not tested yet, but should work. also adds a few distance metrics for clustering.
> Add K-Means to ML package
> -------------------------
>
> Key: HAMA-547
> URL: https://issues.apache.org/jira/browse/HAMA-547
> Project: Hama
> Issue Type: New Feature
> Components: machine learning
> Reporter: praveen sripati
> Assignee: Thomas Jungblut
> Attachments: HAMA-547.patch
>
>
> Based on my blog post:
> http://codingwiththomas.blogspot.in/2011/12/k-means-clustering-with-bsp-intuition.html
> https://github.com/thomasjungblut/thomasjungblut-common/blob/master/src/de/jungblut/clustering/KMeansBSP.java
> We need a math lib for that, let's wait until Mikalai finished with HAMA-524.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Updated] (HAMA-547) Add K-Means to a new ML package
Posted by "Thomas Jungblut (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HAMA-547?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Thomas Jungblut updated HAMA-547:
---------------------------------
Description:
Based on my blog post:
http://codingwiththomas.blogspot.in/2011/12/k-means-clustering-with-bsp-intuition.html
https://github.com/thomasjungblut/thomasjungblut-common/blob/master/src/de/jungblut/clustering/KMeansBSP.java
We need a math lib for that, let's wait until Mikalai finished with HAMA-524.
was:
Submit the K-Means code (1) after mimicking the Mahout API. Existing users of Mahout will be able to use Hama easily and also existing Mahout programs can be ported to Hama with minimal changes.
org.apache.mahout.clustering.kmeans.KMeansClusterer
org.apache.mahout.clustering.kmeans.KMeansDriver
http://codingwiththomas.blogspot.in/2011/12/k-means-clustering-with-bsp-intuition.html
Priority: Major (was: Minor)
Affects Version/s: (was: 0.5.0)
Issue Type: New Feature (was: Improvement)
Summary: Add K-Means to a new ML package (was: Submit the K-Means code after mimicking the Mahout API for K-Means)
> Add K-Means to a new ML package
> -------------------------------
>
> Key: HAMA-547
> URL: https://issues.apache.org/jira/browse/HAMA-547
> Project: Hama
> Issue Type: New Feature
> Components: bsp core
> Reporter: praveen sripati
> Fix For: 0.6.0
>
>
> Based on my blog post:
> http://codingwiththomas.blogspot.in/2011/12/k-means-clustering-with-bsp-intuition.html
> https://github.com/thomasjungblut/thomasjungblut-common/blob/master/src/de/jungblut/clustering/KMeansBSP.java
> We need a math lib for that, let's wait until Mikalai finished with HAMA-524.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Updated] (HAMA-547) Add K-Means to ML package
Posted by "Thomas Jungblut (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HAMA-547?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Thomas Jungblut updated HAMA-547:
---------------------------------
Affects Version/s: 0.5.0
> Add K-Means to ML package
> -------------------------
>
> Key: HAMA-547
> URL: https://issues.apache.org/jira/browse/HAMA-547
> Project: Hama
> Issue Type: New Feature
> Components: machine learning
> Affects Versions: 0.5.0
> Reporter: praveen sripati
> Assignee: Thomas Jungblut
> Fix For: 0.6.0
>
> Attachments: HAMA-547_1.patch, HAMA-547_2.patch, HAMA-547_3.patch, HAMA-547_4.patch
>
>
> Based on my blog post:
> http://codingwiththomas.blogspot.in/2011/12/k-means-clustering-with-bsp-intuition.html
> https://github.com/thomasjungblut/thomasjungblut-common/blob/master/src/de/jungblut/clustering/KMeansBSP.java
> We need a math lib for that, let's wait until Mikalai finished with HAMA-524.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HAMA-547) Add K-Means to ML package
Posted by "Edward J. Yoon (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HAMA-547?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13432291#comment-13432291 ]
Edward J. Yoon commented on HAMA-547:
-------------------------------------
NP, take all the time you need. and if you have good idea about effective way to advertise your world's best parallel k-means clustering, please let me know.
> Add K-Means to ML package
> -------------------------
>
> Key: HAMA-547
> URL: https://issues.apache.org/jira/browse/HAMA-547
> Project: Hama
> Issue Type: New Feature
> Components: machine learning
> Reporter: praveen sripati
> Assignee: Thomas Jungblut
>
> Based on my blog post:
> http://codingwiththomas.blogspot.in/2011/12/k-means-clustering-with-bsp-intuition.html
> https://github.com/thomasjungblut/thomasjungblut-common/blob/master/src/de/jungblut/clustering/KMeansBSP.java
> We need a math lib for that, let's wait until Mikalai finished with HAMA-524.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Resolved] (HAMA-547) Add K-Means to ML package
Posted by "Thomas Jungblut (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HAMA-547?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Thomas Jungblut resolved HAMA-547.
----------------------------------
Resolution: Fixed
Fix Version/s: 0.6.0
> Add K-Means to ML package
> -------------------------
>
> Key: HAMA-547
> URL: https://issues.apache.org/jira/browse/HAMA-547
> Project: Hama
> Issue Type: New Feature
> Components: machine learning
> Reporter: praveen sripati
> Assignee: Thomas Jungblut
> Fix For: 0.6.0
>
> Attachments: HAMA-547_1.patch, HAMA-547_2.patch, HAMA-547_3.patch, HAMA-547_4.patch
>
>
> Based on my blog post:
> http://codingwiththomas.blogspot.in/2011/12/k-means-clustering-with-bsp-intuition.html
> https://github.com/thomasjungblut/thomasjungblut-common/blob/master/src/de/jungblut/clustering/KMeansBSP.java
> We need a math lib for that, let's wait until Mikalai finished with HAMA-524.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HAMA-547) Add K-Means to ML package
Posted by "Thomas Jungblut (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HAMA-547?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13431977#comment-13431977 ]
Thomas Jungblut commented on HAMA-547:
--------------------------------------
Will do on the weekend, sorry for the delay. You're all on vacation anyway :p
> Add K-Means to ML package
> -------------------------
>
> Key: HAMA-547
> URL: https://issues.apache.org/jira/browse/HAMA-547
> Project: Hama
> Issue Type: New Feature
> Components: machine learning
> Reporter: praveen sripati
> Assignee: Thomas Jungblut
>
> Based on my blog post:
> http://codingwiththomas.blogspot.in/2011/12/k-means-clustering-with-bsp-intuition.html
> https://github.com/thomasjungblut/thomasjungblut-common/blob/master/src/de/jungblut/clustering/KMeansBSP.java
> We need a math lib for that, let's wait until Mikalai finished with HAMA-524.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Updated] (HAMA-547) Add K-Means to ML package
Posted by "Thomas Jungblut (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HAMA-547?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Thomas Jungblut updated HAMA-547:
---------------------------------
Attachment: HAMA-547_3.patch
Ah fail the read/write methods were outdated.
But the result from clustering is wrong: {0=[Infinity, Infinity], 1=[7.625, 0.25]}
> Add K-Means to ML package
> -------------------------
>
> Key: HAMA-547
> URL: https://issues.apache.org/jira/browse/HAMA-547
> Project: Hama
> Issue Type: New Feature
> Components: machine learning
> Reporter: praveen sripati
> Assignee: Thomas Jungblut
> Attachments: HAMA-547_1.patch, HAMA-547_2.patch, HAMA-547_3.patch
>
>
> Based on my blog post:
> http://codingwiththomas.blogspot.in/2011/12/k-means-clustering-with-bsp-intuition.html
> https://github.com/thomasjungblut/thomasjungblut-common/blob/master/src/de/jungblut/clustering/KMeansBSP.java
> We need a math lib for that, let's wait until Mikalai finished with HAMA-524.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HAMA-547) Add K-Means to ML package
Posted by "Edward J. Yoon (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HAMA-547?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13431620#comment-13431620 ]
Edward J. Yoon commented on HAMA-547:
-------------------------------------
how is a input format different from mahout? if possible, let's compare.
> Add K-Means to ML package
> -------------------------
>
> Key: HAMA-547
> URL: https://issues.apache.org/jira/browse/HAMA-547
> Project: Hama
> Issue Type: New Feature
> Components: machine learning
> Reporter: praveen sripati
> Assignee: Thomas Jungblut
>
> Based on my blog post:
> http://codingwiththomas.blogspot.in/2011/12/k-means-clustering-with-bsp-intuition.html
> https://github.com/thomasjungblut/thomasjungblut-common/blob/master/src/de/jungblut/clustering/KMeansBSP.java
> We need a math lib for that, let's wait until Mikalai finished with HAMA-524.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HAMA-547) Add K-Means to ML package
Posted by "Hudson (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HAMA-547?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13439134#comment-13439134 ]
Hudson commented on HAMA-547:
-----------------------------
Integrated in Hama-Nightly #649 (See [https://builds.apache.org/job/Hama-Nightly/649/])
[HAMA-547]: Add K-Means to ML package (Revision 1375724)
Result = FAILURE
tjungblut :
Files :
* /hama/trunk/CHANGES.txt
* /hama/trunk/examples/pom.xml
* /hama/trunk/examples/src/main/java/org/apache/hama/examples/ExampleDriver.java
* /hama/trunk/examples/src/main/java/org/apache/hama/examples/Kmeans.java
* /hama/trunk/ml/src/main/java/org/apache/hama/ml/CenterMessage.java
* /hama/trunk/ml/src/main/java/org/apache/hama/ml/KMeansBSP.java
* /hama/trunk/ml/src/main/java/org/apache/hama/ml/distance
* /hama/trunk/ml/src/main/java/org/apache/hama/ml/distance/CosineDistance.java
* /hama/trunk/ml/src/main/java/org/apache/hama/ml/distance/DistanceMeasurer.java
* /hama/trunk/ml/src/main/java/org/apache/hama/ml/distance/EuclidianDistance.java
* /hama/trunk/ml/src/main/java/org/apache/hama/ml/writable/VectorWritable.java
* /hama/trunk/ml/src/test/java/org/apache/hama/ml/TestKMeansBSP.java
> Add K-Means to ML package
> -------------------------
>
> Key: HAMA-547
> URL: https://issues.apache.org/jira/browse/HAMA-547
> Project: Hama
> Issue Type: New Feature
> Components: machine learning
> Affects Versions: 0.5.0
> Reporter: praveen sripati
> Assignee: Thomas Jungblut
> Fix For: 0.6.0
>
> Attachments: HAMA-547_1.patch, HAMA-547_2.patch, HAMA-547_3.patch, HAMA-547_4.patch
>
>
> Based on my blog post:
> http://codingwiththomas.blogspot.in/2011/12/k-means-clustering-with-bsp-intuition.html
> https://github.com/thomasjungblut/thomasjungblut-common/blob/master/src/de/jungblut/clustering/KMeansBSP.java
> We need a math lib for that, let's wait until Mikalai finished with HAMA-524.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Updated] (HAMA-547) Add K-Means to ML package
Posted by "Thomas Jungblut (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HAMA-547?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Thomas Jungblut updated HAMA-547:
---------------------------------
Attachment: HAMA-547_1.patch
just another untested patch
> Add K-Means to ML package
> -------------------------
>
> Key: HAMA-547
> URL: https://issues.apache.org/jira/browse/HAMA-547
> Project: Hama
> Issue Type: New Feature
> Components: machine learning
> Reporter: praveen sripati
> Assignee: Thomas Jungblut
> Attachments: HAMA-547_1.patch
>
>
> Based on my blog post:
> http://codingwiththomas.blogspot.in/2011/12/k-means-clustering-with-bsp-intuition.html
> https://github.com/thomasjungblut/thomasjungblut-common/blob/master/src/de/jungblut/clustering/KMeansBSP.java
> We need a math lib for that, let's wait until Mikalai finished with HAMA-524.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Updated] (HAMA-547) Add K-Means to ML package
Posted by "Thomas Jungblut (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HAMA-547?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Thomas Jungblut updated HAMA-547:
---------------------------------
Attachment: HAMA-547_4.patch
another patch with testcase
> Add K-Means to ML package
> -------------------------
>
> Key: HAMA-547
> URL: https://issues.apache.org/jira/browse/HAMA-547
> Project: Hama
> Issue Type: New Feature
> Components: machine learning
> Reporter: praveen sripati
> Assignee: Thomas Jungblut
> Attachments: HAMA-547_1.patch, HAMA-547_2.patch, HAMA-547_3.patch, HAMA-547_4.patch
>
>
> Based on my blog post:
> http://codingwiththomas.blogspot.in/2011/12/k-means-clustering-with-bsp-intuition.html
> https://github.com/thomasjungblut/thomasjungblut-common/blob/master/src/de/jungblut/clustering/KMeansBSP.java
> We need a math lib for that, let's wait until Mikalai finished with HAMA-524.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Assigned] (HAMA-547) Add K-Means to ML package
Posted by "Thomas Jungblut (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HAMA-547?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Thomas Jungblut reassigned HAMA-547:
------------------------------------
Assignee: Thomas Jungblut
> Add K-Means to ML package
> -------------------------
>
> Key: HAMA-547
> URL: https://issues.apache.org/jira/browse/HAMA-547
> Project: Hama
> Issue Type: New Feature
> Components: machine learning
> Reporter: praveen sripati
> Assignee: Thomas Jungblut
>
> Based on my blog post:
> http://codingwiththomas.blogspot.in/2011/12/k-means-clustering-with-bsp-intuition.html
> https://github.com/thomasjungblut/thomasjungblut-common/blob/master/src/de/jungblut/clustering/KMeansBSP.java
> We need a math lib for that, let's wait until Mikalai finished with HAMA-524.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HAMA-547) Add K-Means to ML package
Posted by "Tommaso Teofili (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HAMA-547?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13431749#comment-13431749 ]
Tommaso Teofili commented on HAMA-547:
--------------------------------------
Hi Thomas, that looks good, I think you can commit it and then we start improving / testing it.
> Add K-Means to ML package
> -------------------------
>
> Key: HAMA-547
> URL: https://issues.apache.org/jira/browse/HAMA-547
> Project: Hama
> Issue Type: New Feature
> Components: machine learning
> Reporter: praveen sripati
> Assignee: Thomas Jungblut
>
> Based on my blog post:
> http://codingwiththomas.blogspot.in/2011/12/k-means-clustering-with-bsp-intuition.html
> https://github.com/thomasjungblut/thomasjungblut-common/blob/master/src/de/jungblut/clustering/KMeansBSP.java
> We need a math lib for that, let's wait until Mikalai finished with HAMA-524.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HAMA-547) Add K-Means to ML package
Posted by "Thomas Jungblut (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HAMA-547?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13431751#comment-13431751 ]
Thomas Jungblut commented on HAMA-547:
--------------------------------------
I wanted to make it compatible to the text input and the other examples. I will do tonight.
> Add K-Means to ML package
> -------------------------
>
> Key: HAMA-547
> URL: https://issues.apache.org/jira/browse/HAMA-547
> Project: Hama
> Issue Type: New Feature
> Components: machine learning
> Reporter: praveen sripati
> Assignee: Thomas Jungblut
>
> Based on my blog post:
> http://codingwiththomas.blogspot.in/2011/12/k-means-clustering-with-bsp-intuition.html
> https://github.com/thomasjungblut/thomasjungblut-common/blob/master/src/de/jungblut/clustering/KMeansBSP.java
> We need a math lib for that, let's wait until Mikalai finished with HAMA-524.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Updated] (HAMA-547) Add K-Means to ML package
Posted by "Thomas Jungblut (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HAMA-547?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Thomas Jungblut updated HAMA-547:
---------------------------------
Attachment: (was: HAMA-547.patch)
> Add K-Means to ML package
> -------------------------
>
> Key: HAMA-547
> URL: https://issues.apache.org/jira/browse/HAMA-547
> Project: Hama
> Issue Type: New Feature
> Components: machine learning
> Reporter: praveen sripati
> Assignee: Thomas Jungblut
>
> Based on my blog post:
> http://codingwiththomas.blogspot.in/2011/12/k-means-clustering-with-bsp-intuition.html
> https://github.com/thomasjungblut/thomasjungblut-common/blob/master/src/de/jungblut/clustering/KMeansBSP.java
> We need a math lib for that, let's wait until Mikalai finished with HAMA-524.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Updated] (HAMA-547) Add K-Means to ML package
Posted by "Edward J. Yoon (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HAMA-547?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Edward J. Yoon updated HAMA-547:
--------------------------------
Component/s: (was: bsp core)
machine learning
Fix Version/s: (was: 0.6.0)
Unlinking. Because HAMA-524 goal is just implementation of SpMV algorithm, not a math lib, GSoC has few deadlines, I don't want to give other load to him.
We can extract or merge reusable code later.
> Add K-Means to ML package
> -------------------------
>
> Key: HAMA-547
> URL: https://issues.apache.org/jira/browse/HAMA-547
> Project: Hama
> Issue Type: New Feature
> Components: machine learning
> Reporter: praveen sripati
>
> Based on my blog post:
> http://codingwiththomas.blogspot.in/2011/12/k-means-clustering-with-bsp-intuition.html
> https://github.com/thomasjungblut/thomasjungblut-common/blob/master/src/de/jungblut/clustering/KMeansBSP.java
> We need a math lib for that, let's wait until Mikalai finished with HAMA-524.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HAMA-547) Add K-Means to ML package
Posted by "Thomas Jungblut (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HAMA-547?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13438939#comment-13438939 ]
Thomas Jungblut commented on HAMA-547:
--------------------------------------
I'm not through testing edge cases, especially on clustered enviroments. But let's do it later by follow-ups if issues arise.
Thanks for the upvote, going to commit this then.
> Add K-Means to ML package
> -------------------------
>
> Key: HAMA-547
> URL: https://issues.apache.org/jira/browse/HAMA-547
> Project: Hama
> Issue Type: New Feature
> Components: machine learning
> Reporter: praveen sripati
> Assignee: Thomas Jungblut
> Attachments: HAMA-547_1.patch, HAMA-547_2.patch, HAMA-547_3.patch, HAMA-547_4.patch
>
>
> Based on my blog post:
> http://codingwiththomas.blogspot.in/2011/12/k-means-clustering-with-bsp-intuition.html
> https://github.com/thomasjungblut/thomasjungblut-common/blob/master/src/de/jungblut/clustering/KMeansBSP.java
> We need a math lib for that, let's wait until Mikalai finished with HAMA-524.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Updated] (HAMA-547) Submit the K-Means code after mimicking
the Mahout API for K-Means
Posted by "Edward J. Yoon (Updated) (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HAMA-547?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Edward J. Yoon updated HAMA-547:
--------------------------------
Affects Version/s: (was: 0.4.0)
0.5.0
Fix Version/s: (was: 0.5.0)
> Submit the K-Means code after mimicking the Mahout API for K-Means
> ------------------------------------------------------------------
>
> Key: HAMA-547
> URL: https://issues.apache.org/jira/browse/HAMA-547
> Project: Hama
> Issue Type: Improvement
> Components: bsp
> Affects Versions: 0.5.0
> Reporter: praveen sripati
> Priority: Minor
> Fix For: 0.6.0
>
>
> Submit the K-Means code (1) after mimicking the Mahout API. Existing users of Mahout will be able to use Hama easily and also existing Mahout programs can be ported to Hama with minimal changes.
> org.apache.mahout.clustering.kmeans.KMeansClusterer
> org.apache.mahout.clustering.kmeans.KMeansDriver
> http://codingwiththomas.blogspot.in/2011/12/k-means-clustering-with-bsp-intuition.html
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HAMA-547) Add K-Means to ML package
Posted by "Edward J. Yoon (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HAMA-547?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13438520#comment-13438520 ]
Edward J. Yoon commented on HAMA-547:
-------------------------------------
+1 let's commit this.
> Add K-Means to ML package
> -------------------------
>
> Key: HAMA-547
> URL: https://issues.apache.org/jira/browse/HAMA-547
> Project: Hama
> Issue Type: New Feature
> Components: machine learning
> Reporter: praveen sripati
> Assignee: Thomas Jungblut
> Attachments: HAMA-547_1.patch, HAMA-547_2.patch, HAMA-547_3.patch, HAMA-547_4.patch
>
>
> Based on my blog post:
> http://codingwiththomas.blogspot.in/2011/12/k-means-clustering-with-bsp-intuition.html
> https://github.com/thomasjungblut/thomasjungblut-common/blob/master/src/de/jungblut/clustering/KMeansBSP.java
> We need a math lib for that, let's wait until Mikalai finished with HAMA-524.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Updated] (HAMA-547) Add K-Means to ML package
Posted by "Thomas Jungblut (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HAMA-547?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Thomas Jungblut updated HAMA-547:
---------------------------------
Summary: Add K-Means to ML package (was: Add K-Means to a new ML package)
> Add K-Means to ML package
> -------------------------
>
> Key: HAMA-547
> URL: https://issues.apache.org/jira/browse/HAMA-547
> Project: Hama
> Issue Type: New Feature
> Components: bsp core
> Reporter: praveen sripati
> Fix For: 0.6.0
>
>
> Based on my blog post:
> http://codingwiththomas.blogspot.in/2011/12/k-means-clustering-with-bsp-intuition.html
> https://github.com/thomasjungblut/thomasjungblut-common/blob/master/src/de/jungblut/clustering/KMeansBSP.java
> We need a math lib for that, let's wait until Mikalai finished with HAMA-524.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira