[jira] [Closed] (SYSTEMML-381) Data converters sparse matrix blocks

["Matthias Boehm (JIRA)" <ji...@apache.org>]

     [ https://issues.apache.org/jira/browse/SYSTEMML-381?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Matthias Boehm closed SYSTEMML-381.
-----------------------------------

> Data converters sparse matrix blocks
> ------------------------------------
>
>                 Key: SYSTEMML-381
>                 URL: https://issues.apache.org/jira/browse/SYSTEMML-381
>             Project: SystemML
>          Issue Type: Task
>            Reporter: Matthias Boehm
>            Assignee: Matthias Boehm
>             Fix For: SystemML 0.10
>
>
> In order to allow for a seamless integration into our runtime and the ability to work with different representations of sparse matrices in a single ML program, we need to provide data converters between the different formats. We aim to create SparseBlock->SparseBlockXXX converters which convert - based on the sparse block abstraction created in [SYSTEMML-378] - arbitrary sparse block formats into a specific SparseBlock implementation. This reduces the number of combinations from O(n^2) to O( n) and simplifies the process of adding new sparse block representations. 



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)