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 (JIRA)" <ji...@apache.org> on 2008/10/20 08:42:44 UTC
[jira] Created: (HAMA-83) 2D sqaure blocking for matrix
multiplication
2D sqaure blocking for matrix multiplication
--------------------------------------------
Key: HAMA-83
URL: https://issues.apache.org/jira/browse/HAMA-83
Project: Hama
Issue Type: New Feature
Components: implementation
Affects Versions: 0.1.0
Reporter: Edward J. Yoon
Assignee: Edward J. Yoon
Fix For: 0.1.0
The matrix multiplication of the original arrays can be transformed into matrix multiplication of blocks. For example,
C_block(1,1)=A_block(1,1)*B_block(1,1) + A_block(1,2)*B_block(2,1)
emit <blockKey, subMatrix> pairs instead of current <Int, Vector> input formatter, and have a reducer sum up the sub matrices as describe below.
{code}
Mapper<BlockKey, subMatrix> {
sub-matrix multiplication of block
}
Reducer<BlockKey, subMatrix> {
summation all sub-matrices
}
{code}
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
Re: [jira] Commented: (HAMA-83) 2D sqaure blocking for dense matrix multiplication
Posted by "Edward J. Yoon" <ed...@apache.org>.
Hmm.. BTW, How to synchronize them? Should we blocking when operate something?
On Sat, Nov 15, 2008 at 11:22 AM, Edward J. Yoon (JIRA) <ji...@apache.org> wrote:
>
> [ https://issues.apache.org/jira/browse/HAMA-83?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12647792#action_12647792 ]
>
> Edward J. Yoon commented on HAMA-83:
> ------------------------------------
>
> oh. counter also should be changed.
>
>> 2D sqaure blocking for dense matrix multiplication
>> --------------------------------------------------
>>
>> Key: HAMA-83
>> URL: https://issues.apache.org/jira/browse/HAMA-83
>> Project: Hama
>> Issue Type: New Feature
>> Components: implementation
>> Affects Versions: 0.1.0
>> Reporter: Edward J. Yoon
>> Assignee: Edward J. Yoon
>> Fix For: 0.1.0
>>
>> Attachments: HAMA-83.patch
>>
>>
>> The matrix multiplication of the original arrays can be transformed into matrix multiplication of blocks. For example,
>> C_block(1,1)=A_block(1,1)*B_block(1,1) + A_block(1,2)*B_block(2,1)
>> emit <blockKey, subMatrix> pairs instead of current <Int, Vector> input formatter, and have a reducer sum up the sub matrices as describe below.
>> {code}
>> Mapper<BlockKey, subMatrix> {
>> sub-matrix multiplication of block
>> }
>> Reducer<BlockKey, subMatrix> {
>> summation all sub-matrices
>> }
>> {code}
>
> --
> This message is automatically generated by JIRA.
> -
> You can reply to this email to add a comment to the issue online.
>
>
--
Best Regards, Edward J. Yoon @ NHN, corp.
edwardyoon@apache.org
http://blog.udanax.org
[jira] Commented: (HAMA-83) 2D sqaure blocking for dense matrix
multiplication
Posted by "Edward J. Yoon (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HAMA-83?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12646816#action_12646816 ]
Edward J. Yoon commented on HAMA-83:
------------------------------------
This is my plan.
1. Generate formula
2. To avoid a lot of full scan, collect the data to new structure
3. Calculate them on map/reduce
4. Put result to table
> 2D sqaure blocking for dense matrix multiplication
> --------------------------------------------------
>
> Key: HAMA-83
> URL: https://issues.apache.org/jira/browse/HAMA-83
> Project: Hama
> Issue Type: New Feature
> Components: implementation
> Affects Versions: 0.1.0
> Reporter: Edward J. Yoon
> Assignee: Edward J. Yoon
> Fix For: 0.1.0
>
>
> The matrix multiplication of the original arrays can be transformed into matrix multiplication of blocks. For example,
> C_block(1,1)=A_block(1,1)*B_block(1,1) + A_block(1,2)*B_block(2,1)
> emit <blockKey, subMatrix> pairs instead of current <Int, Vector> input formatter, and have a reducer sum up the sub matrices as describe below.
> {code}
> Mapper<BlockKey, subMatrix> {
> sub-matrix multiplication of block
> }
> Reducer<BlockKey, subMatrix> {
> summation all sub-matrices
> }
> {code}
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Assigned: (HAMA-83) 2D sqaure blocking for dense matrix
multiplication
Posted by "Edward J. Yoon (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HAMA-83?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Edward J. Yoon reassigned HAMA-83:
----------------------------------
Assignee: (was: Edward J. Yoon)
> 2D sqaure blocking for dense matrix multiplication
> --------------------------------------------------
>
> Key: HAMA-83
> URL: https://issues.apache.org/jira/browse/HAMA-83
> Project: Hama
> Issue Type: New Feature
> Components: implementation
> Affects Versions: 0.1.0
> Reporter: Edward J. Yoon
> Fix For: 0.1.0
>
>
> The matrix multiplication of the original arrays can be transformed into matrix multiplication of blocks. For example,
> C_block(1,1)=A_block(1,1)*B_block(1,1) + A_block(1,2)*B_block(2,1)
> emit <blockKey, subMatrix> pairs instead of current <Int, Vector> input formatter, and have a reducer sum up the sub matrices as describe below.
> {code}
> Mapper<BlockKey, subMatrix> {
> sub-matrix multiplication of block
> }
> Reducer<BlockKey, subMatrix> {
> summation all sub-matrices
> }
> {code}
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Commented: (HAMA-83) 2D sqaure blocking for matrix
multiplication
Posted by "Edward J. Yoon (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HAMA-83?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12640936#action_12640936 ]
Edward J. Yoon commented on HAMA-83:
------------------------------------
Let's re-arrange a name of BlockKey to BlockID.
> 2D sqaure blocking for matrix multiplication
> --------------------------------------------
>
> Key: HAMA-83
> URL: https://issues.apache.org/jira/browse/HAMA-83
> Project: Hama
> Issue Type: New Feature
> Components: implementation
> Affects Versions: 0.1.0
> Reporter: Edward J. Yoon
> Assignee: Edward J. Yoon
> Fix For: 0.1.0
>
>
> The matrix multiplication of the original arrays can be transformed into matrix multiplication of blocks. For example,
> C_block(1,1)=A_block(1,1)*B_block(1,1) + A_block(1,2)*B_block(2,1)
> emit <blockKey, subMatrix> pairs instead of current <Int, Vector> input formatter, and have a reducer sum up the sub matrices as describe below.
> {code}
> Mapper<BlockKey, subMatrix> {
> sub-matrix multiplication of block
> }
> Reducer<BlockKey, subMatrix> {
> summation all sub-matrices
> }
> {code}
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Updated: (HAMA-83) 2D sqaure blocking for dense matrix
multiplication
Posted by "Edward J. Yoon (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HAMA-83?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Edward J. Yoon updated HAMA-83:
-------------------------------
Attachment: HAMA-83_v01.patch
attach a new patch.
> 2D sqaure blocking for dense matrix multiplication
> --------------------------------------------------
>
> Key: HAMA-83
> URL: https://issues.apache.org/jira/browse/HAMA-83
> Project: Hama
> Issue Type: New Feature
> Components: implementation
> Affects Versions: 0.1.0
> Reporter: Edward J. Yoon
> Assignee: Edward J. Yoon
> Fix For: 0.1.0
>
> Attachments: HAMA-83.patch, HAMA-83_v01.patch
>
>
> The matrix multiplication of the original arrays can be transformed into matrix multiplication of blocks. For example,
> C_block(1,1)=A_block(1,1)*B_block(1,1) + A_block(1,2)*B_block(2,1)
> emit <blockKey, subMatrix> pairs instead of current <Int, Vector> input formatter, and have a reducer sum up the sub matrices as describe below.
> {code}
> Mapper<BlockKey, subMatrix> {
> sub-matrix multiplication of block
> }
> Reducer<BlockKey, subMatrix> {
> summation all sub-matrices
> }
> {code}
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Commented: (HAMA-83) 2D sqaure blocking for dense matrix
multiplication
Posted by "Edward J. Yoon (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HAMA-83?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12647792#action_12647792 ]
Edward J. Yoon commented on HAMA-83:
------------------------------------
oh. counter also should be changed.
> 2D sqaure blocking for dense matrix multiplication
> --------------------------------------------------
>
> Key: HAMA-83
> URL: https://issues.apache.org/jira/browse/HAMA-83
> Project: Hama
> Issue Type: New Feature
> Components: implementation
> Affects Versions: 0.1.0
> Reporter: Edward J. Yoon
> Assignee: Edward J. Yoon
> Fix For: 0.1.0
>
> Attachments: HAMA-83.patch
>
>
> The matrix multiplication of the original arrays can be transformed into matrix multiplication of blocks. For example,
> C_block(1,1)=A_block(1,1)*B_block(1,1) + A_block(1,2)*B_block(2,1)
> emit <blockKey, subMatrix> pairs instead of current <Int, Vector> input formatter, and have a reducer sum up the sub matrices as describe below.
> {code}
> Mapper<BlockKey, subMatrix> {
> sub-matrix multiplication of block
> }
> Reducer<BlockKey, subMatrix> {
> summation all sub-matrices
> }
> {code}
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Updated: (HAMA-83) 2D sqaure blocking for dense matrix
multiplication
Posted by "Edward J. Yoon (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HAMA-83?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Edward J. Yoon updated HAMA-83:
-------------------------------
Status: Open (was: Patch Available)
> 2D sqaure blocking for dense matrix multiplication
> --------------------------------------------------
>
> Key: HAMA-83
> URL: https://issues.apache.org/jira/browse/HAMA-83
> Project: Hama
> Issue Type: New Feature
> Components: implementation
> Affects Versions: 0.1.0
> Reporter: Edward J. Yoon
> Assignee: Edward J. Yoon
> Fix For: 0.1.0
>
> Attachments: HAMA-83.patch, HAMA-83_v01.patch
>
>
> The matrix multiplication of the original arrays can be transformed into matrix multiplication of blocks. For example,
> C_block(1,1)=A_block(1,1)*B_block(1,1) + A_block(1,2)*B_block(2,1)
> emit <blockKey, subMatrix> pairs instead of current <Int, Vector> input formatter, and have a reducer sum up the sub matrices as describe below.
> {code}
> Mapper<BlockKey, subMatrix> {
> sub-matrix multiplication of block
> }
> Reducer<BlockKey, subMatrix> {
> summation all sub-matrices
> }
> {code}
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Commented: (HAMA-83) 2D sqaure blocking for dense matrix
multiplication
Posted by "Hudson (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HAMA-83?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12647603#action_12647603 ]
Hudson commented on HAMA-83:
----------------------------
-1 overall. Here are the results of testing the latest attachment
http://issues.apache.org/jira/secure/attachment/12393927/HAMA-83.patch
against trunk revision 713666.
@author +1. The patch does not contain any @author tags.
tests included +1. The patch appears to include 9 new or modified tests.
javadoc +1. The javadoc tool did not generate any warning messages.
javac +1. The applied patch does not generate any new javac compiler warnings.
release audit +1. The applied patch does not generate any new release audit warnings.
findbugs +1. The patch does not introduce any new Findbugs warnings.
core tests -1. The patch failed core unit tests.
Test results: http://hudson.zones.apache.org/hudson/job/Hama-Patch/113/testReport/
Findbugs warnings: http://hudson.zones.apache.org/hudson/job/Hama-Patch/113/artifact/trunk/build/reports/findbugs/newPatchFindbugsWarnings.html
Checkstyle results: http://hudson.zones.apache.org/hudson/job/Hama-Patch/113/artifact/trunk/build/test/checkstyle-errors.html
Console output: http://hudson.zones.apache.org/hudson/job/Hama-Patch/113/console
This message is automatically generated.
> 2D sqaure blocking for dense matrix multiplication
> --------------------------------------------------
>
> Key: HAMA-83
> URL: https://issues.apache.org/jira/browse/HAMA-83
> Project: Hama
> Issue Type: New Feature
> Components: implementation
> Affects Versions: 0.1.0
> Reporter: Edward J. Yoon
> Assignee: Edward J. Yoon
> Fix For: 0.1.0
>
> Attachments: HAMA-83.patch
>
>
> The matrix multiplication of the original arrays can be transformed into matrix multiplication of blocks. For example,
> C_block(1,1)=A_block(1,1)*B_block(1,1) + A_block(1,2)*B_block(2,1)
> emit <blockKey, subMatrix> pairs instead of current <Int, Vector> input formatter, and have a reducer sum up the sub matrices as describe below.
> {code}
> Mapper<BlockKey, subMatrix> {
> sub-matrix multiplication of block
> }
> Reducer<BlockKey, subMatrix> {
> summation all sub-matrices
> }
> {code}
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Updated: (HAMA-83) 2D sqaure blocking for dense matrix
multiplication
Posted by "Edward J. Yoon (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HAMA-83?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Edward J. Yoon updated HAMA-83:
-------------------------------
Status: Patch Available (was: Open)
Local test passed. Rescheduling.
> 2D sqaure blocking for dense matrix multiplication
> --------------------------------------------------
>
> Key: HAMA-83
> URL: https://issues.apache.org/jira/browse/HAMA-83
> Project: Hama
> Issue Type: New Feature
> Components: implementation
> Affects Versions: 0.1.0
> Reporter: Edward J. Yoon
> Assignee: Edward J. Yoon
> Fix For: 0.1.0
>
> Attachments: HAMA-83.patch, HAMA-83_v01.patch
>
>
> The matrix multiplication of the original arrays can be transformed into matrix multiplication of blocks. For example,
> C_block(1,1)=A_block(1,1)*B_block(1,1) + A_block(1,2)*B_block(2,1)
> emit <blockKey, subMatrix> pairs instead of current <Int, Vector> input formatter, and have a reducer sum up the sub matrices as describe below.
> {code}
> Mapper<BlockKey, subMatrix> {
> sub-matrix multiplication of block
> }
> Reducer<BlockKey, subMatrix> {
> summation all sub-matrices
> }
> {code}
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Updated: (HAMA-83) 2D sqaure blocking for dense matrix
multiplication
Posted by "Edward J. Yoon (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HAMA-83?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Edward J. Yoon updated HAMA-83:
-------------------------------
Attachment: HAMA-83.patch
After apply this patch, 300 * 300 mat-mat multiplication is only need 12 seconds. : )
> 2D sqaure blocking for dense matrix multiplication
> --------------------------------------------------
>
> Key: HAMA-83
> URL: https://issues.apache.org/jira/browse/HAMA-83
> Project: Hama
> Issue Type: New Feature
> Components: implementation
> Affects Versions: 0.1.0
> Reporter: Edward J. Yoon
> Assignee: Edward J. Yoon
> Fix For: 0.1.0
>
> Attachments: HAMA-83.patch
>
>
> The matrix multiplication of the original arrays can be transformed into matrix multiplication of blocks. For example,
> C_block(1,1)=A_block(1,1)*B_block(1,1) + A_block(1,2)*B_block(2,1)
> emit <blockKey, subMatrix> pairs instead of current <Int, Vector> input formatter, and have a reducer sum up the sub matrices as describe below.
> {code}
> Mapper<BlockKey, subMatrix> {
> sub-matrix multiplication of block
> }
> Reducer<BlockKey, subMatrix> {
> summation all sub-matrices
> }
> {code}
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Updated: (HAMA-83) 2D sqaure blocking for dense matrix
multiplication
Posted by "Edward J. Yoon (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HAMA-83?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Edward J. Yoon updated HAMA-83:
-------------------------------
Status: Patch Available (was: Open)
Submit to hudson.
> 2D sqaure blocking for dense matrix multiplication
> --------------------------------------------------
>
> Key: HAMA-83
> URL: https://issues.apache.org/jira/browse/HAMA-83
> Project: Hama
> Issue Type: New Feature
> Components: implementation
> Affects Versions: 0.1.0
> Reporter: Edward J. Yoon
> Assignee: Edward J. Yoon
> Fix For: 0.1.0
>
> Attachments: HAMA-83.patch
>
>
> The matrix multiplication of the original arrays can be transformed into matrix multiplication of blocks. For example,
> C_block(1,1)=A_block(1,1)*B_block(1,1) + A_block(1,2)*B_block(2,1)
> emit <blockKey, subMatrix> pairs instead of current <Int, Vector> input formatter, and have a reducer sum up the sub matrices as describe below.
> {code}
> Mapper<BlockKey, subMatrix> {
> sub-matrix multiplication of block
> }
> Reducer<BlockKey, subMatrix> {
> summation all sub-matrices
> }
> {code}
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Updated: (HAMA-83) 2D sqaure blocking for dense matrix
multiplication
Posted by "Edward J. Yoon (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HAMA-83?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Edward J. Yoon updated HAMA-83:
-------------------------------
Resolution: Fixed
Status: Resolved (was: Patch Available)
Test passed. I just committed this.
----
+1 overall. Here are the results of testing the latest attachment
http://issues.apache.org/jira/secure/attachment/12394030/HAMA-83_v01.patch
against trunk revision 713666.
@author +1. The patch does not contain any @author tags.
tests included +1. The patch appears to include 12 new or modified tests.
javadoc +1. The javadoc tool did not generate any warning messages.
javac +1. The applied patch does not generate any new javac compiler warnings.
release audit +1. The applied patch does not generate any new release audit warnings.
findbugs +1. The patch does not introduce any new Findbugs warnings.
core tests +1. The patch passed core unit tests.
Test results: http://hudson.zones.apache.org/hudson/job/Hama-Patch/114/testReport/
Findbugs warnings: http://hudson.zones.apache.org/hudson/job/Hama-Patch/114/artifact/trunk/build/reports/findbugs/newPatchFindbugsWarnings.html
Checkstyle results: http://hudson.zones.apache.org/hudson/job/Hama-Patch/114/artifact/trunk/build/test/checkstyle-errors.html
Console output: http://hudson.zones.apache.org/hudson/job/Hama-Patch/114/console
This message is automatically generated.
Login failed
<Fault soapenv:Server.userException: com.atlassian.jira.rpc.exception.RemoteAuthenticationException: Invalid username or password.: <SOAPpy.Types.structType detail at 147023916>: {'hostname': 'brutus', 'com.atlassian.jira.rpc.exception.RemoteAuthenticationException': ''}>
Previous login is invalid or has expired
Login failed
<Fault soapenv:Server.userException: com.atlassian.jira.rpc.exception.RemoteAuthenticationException: Invalid username or password.: <SOAPpy.Types.structType detail at 147362988>: {'hostname': 'brutus', 'com.atlassian.jira.rpc.exception.RemoteAuthenticationException': ''}>
Previous login is invalid or has expired
> 2D sqaure blocking for dense matrix multiplication
> --------------------------------------------------
>
> Key: HAMA-83
> URL: https://issues.apache.org/jira/browse/HAMA-83
> Project: Hama
> Issue Type: New Feature
> Components: implementation
> Affects Versions: 0.1.0
> Reporter: Edward J. Yoon
> Assignee: Edward J. Yoon
> Fix For: 0.1.0
>
> Attachments: HAMA-83.patch, HAMA-83_v01.patch
>
>
> The matrix multiplication of the original arrays can be transformed into matrix multiplication of blocks. For example,
> C_block(1,1)=A_block(1,1)*B_block(1,1) + A_block(1,2)*B_block(2,1)
> emit <blockKey, subMatrix> pairs instead of current <Int, Vector> input formatter, and have a reducer sum up the sub matrices as describe below.
> {code}
> Mapper<BlockKey, subMatrix> {
> sub-matrix multiplication of block
> }
> Reducer<BlockKey, subMatrix> {
> summation all sub-matrices
> }
> {code}
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Assigned: (HAMA-83) 2D sqaure blocking for dense matrix
multiplication
Posted by "Edward J. Yoon (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HAMA-83?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Edward J. Yoon reassigned HAMA-83:
----------------------------------
Assignee: Edward J. Yoon
> 2D sqaure blocking for dense matrix multiplication
> --------------------------------------------------
>
> Key: HAMA-83
> URL: https://issues.apache.org/jira/browse/HAMA-83
> Project: Hama
> Issue Type: New Feature
> Components: implementation
> Affects Versions: 0.1.0
> Reporter: Edward J. Yoon
> Assignee: Edward J. Yoon
> Fix For: 0.1.0
>
>
> The matrix multiplication of the original arrays can be transformed into matrix multiplication of blocks. For example,
> C_block(1,1)=A_block(1,1)*B_block(1,1) + A_block(1,2)*B_block(2,1)
> emit <blockKey, subMatrix> pairs instead of current <Int, Vector> input formatter, and have a reducer sum up the sub matrices as describe below.
> {code}
> Mapper<BlockKey, subMatrix> {
> sub-matrix multiplication of block
> }
> Reducer<BlockKey, subMatrix> {
> summation all sub-matrices
> }
> {code}
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Updated: (HAMA-83) 2D sqaure blocking for dense matrix
multiplication
Posted by "Edward J. Yoon (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HAMA-83?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Edward J. Yoon updated HAMA-83:
-------------------------------
Summary: 2D sqaure blocking for dense matrix multiplication (was: 2D sqaure blocking for matrix multiplication)
Current mult operation algorithm is too bad. it should be removed.
> 2D sqaure blocking for dense matrix multiplication
> --------------------------------------------------
>
> Key: HAMA-83
> URL: https://issues.apache.org/jira/browse/HAMA-83
> Project: Hama
> Issue Type: New Feature
> Components: implementation
> Affects Versions: 0.1.0
> Reporter: Edward J. Yoon
> Assignee: Edward J. Yoon
> Fix For: 0.1.0
>
>
> The matrix multiplication of the original arrays can be transformed into matrix multiplication of blocks. For example,
> C_block(1,1)=A_block(1,1)*B_block(1,1) + A_block(1,2)*B_block(2,1)
> emit <blockKey, subMatrix> pairs instead of current <Int, Vector> input formatter, and have a reducer sum up the sub matrices as describe below.
> {code}
> Mapper<BlockKey, subMatrix> {
> sub-matrix multiplication of block
> }
> Reducer<BlockKey, subMatrix> {
> summation all sub-matrices
> }
> {code}
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.