[
https://issues.apache.org/jira/browse/MAHOUT-376?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12966394#action_12966394
]
Dmitriy Lyubimov commented on MAHOUT-376:
-----------------------------------------
{quote}You don't get a choice. The frame will run the combiner as many times as
it feels it wants to. It will run in the mapper or reducer or both or
neither.{quote}
Ted, thank you for pointing it. As i mentioned couple of comments above, this
is very well understood, was a concern from the beginning, never came up as a
problem in tests, but I will do a small patch per my comment above that will
make processing complexity even faster albeit code might become a little
uglier. I am very well aware mapper can spill some records past combiner and
send them to sort in the reducer per spec. I
In fact, i had version that does just that on another branch. i just need to
yank it and bring in sync with this branch.
Thank you.
> Implement Map-reduce version of stochastic SVD
> ----------------------------------------------
>
> Key: MAHOUT-376
> URL: https://issues.apache.org/jira/browse/MAHOUT-376
> Project: Mahout
> Issue Type: Improvement
> Components: Math
> Reporter: Ted Dunning
> Assignee: Ted Dunning
> Fix For: 0.5
>
> Attachments: MAHOUT-376.patch, Modified stochastic svd algorithm for
> mapreduce.pdf, QR decomposition for Map.pdf, QR decomposition for Map.pdf, QR
> decomposition for Map.pdf, sd-bib.bib, sd.pdf, sd.pdf, sd.pdf, sd.pdf,
> sd.tex, sd.tex, sd.tex, sd.tex, SSVD working notes.pdf, SSVD working
> notes.pdf, SSVD working notes.pdf, ssvd-CDH3-or-0.21.patch.gz,
> ssvd-CDH3-or-0.21.patch.gz, ssvd-m1.patch.gz, ssvd-m2.patch.gz,
> ssvd-m3.patch.gz, Stochastic SVD using eigensolver trick.pdf
>
>
> See attached pdf for outline of proposed method.
> All comments are welcome.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
