You are viewing a plain text version of this content. The canonical link for it is here.
Posted to user@hbase.apache.org by as...@gmail.com on 2008/09/22 15:55:23 UTC

Doing MapReduce on sparse matrices in hbase

Hi,

I need to use hbase to 
1) store two sparse matrices.
2) Write a mapreduce program to multiply them.

Problem 1: How do I create a table for sparse matrices in hbase and then pass it to a mapreduce application?

Problem 2:
The problem that I face is that the examples of mapper functions that I have seen - wordcount etc only look at one data item at a time and produce the key-value pair.
How do I look at more than one item at a time to produce a key-value pair?


Re: Doing MapReduce on sparse matrices in hbase

Posted by Cosmin Lehene <cl...@adobe.com>.
Hi,

Not directly related to your problem, but...

Take a look at this project: http://incubator.apache.org/hama/
I'm not sure what stage it's in, but it's supposed to take care of some problems in the same domain.

Regards,
Cosmin



On 9/22/08 4:55 PM, "ashishgandhe@gmail.com" <as...@gmail.com> wrote:

Hi,

I need to use hbase to
1) store two sparse matrices.
2) Write a mapreduce program to multiply them.

Problem 1: How do I create a table for sparse matrices in hbase and then pass it to a mapreduce application?

Problem 2:
The problem that I face is that the examples of mapper functions that I have seen - wordcount etc only look at one data item at a time and produce the key-value pair.
How do I look at more than one item at a time to produce a key-value pair?