[
https://issues.apache.org/jira/browse/MAHOUT-1574?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14020709#comment-14020709
]
Dmitriy Lyubimov commented on MAHOUT-1574:
------------------------------------------
of course i read it. Which is why i said it is a good first step rather than
something else. Still it misses cases of transpose views and symmetry cases of
SCM and SRM (for once, at the very least with this approach we need something
symmetrical in the SparseColumnMatrix).
As for complexity, well, we don't have to nail all complexity, i'd be happy if
we just nail SRM and SCM and SM and view transpose. I think it would capture
most of what is Sebastian has been referring to. Traversal locality probably
will help a lot, too, in some cases, because it would reduce need for number of
generated view POJOs.
> SparseRowMatrix needs performance improvement for times()
> ---------------------------------------------------------
>
> Key: MAHOUT-1574
> URL: https://issues.apache.org/jira/browse/MAHOUT-1574
> Project: Mahout
> Issue Type: Bug
> Reporter: Ted Dunning
> Assignee: Ted Dunning
>
> According to ssc,
> > * SparseRowMatrix with sequential vectors times SparseRowMatrix with
> > sequential vectors is totally broken, it uses three nested loops and uses
> > get(row, col) on the matrices, which internally uses binary search...
--
This message was sent by Atlassian JIRA
(v6.2#6252)
