You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@mahout.apache.org by "Dmitriy Lyubimov (Issue Comment Edited) (JIRA)" <ji...@apache.org> on 2011/12/12 09:56:30 UTC
[jira] [Issue Comment Edited] (MAHOUT-797) MapReduce SSVD: provide
alternative B-pipeline per B=R' ^{-1} Y'A
[ https://issues.apache.org/jira/browse/MAHOUT-797?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13167396#comment-13167396 ]
Dmitriy Lyubimov edited comment on MAHOUT-797 at 12/12/11 8:54 AM:
-------------------------------------------------------------------
Thank you.
rank deficiency of the input, yes, that's what i thought. but i can't figure how it may end up being singular, given that i generated it randomly. random input is virtually guaranteed to be non-singular. Ok must be something with my test program.
I guess i am also a little bit confused about this whole positive definite thing. Y'Y is guaranteed to be Hermitian but is it guaranteed to be positive definite? At least in my experiments Y'Y doesn't seem to ever end up positive definite.
You are right, some of singular values were zero. So now it runs ok if i step thru all steps it but still gives me degenerate singular system error if i call it as a routine. I am still missing something.
was (Author: dlyubimov):
Thank you.
rank deficiency of the input, yes, that's what i thought. but i can't figure how it may end up being singular, given that i generated it randomly. random input is virtually guaranteed to be non-singular. Ok must be something with my test program.
I guess i am also a little bit confused about this whole positive definite thing. Y'Y is guaranteed to be Hermitian but is it guaranteed to be positive definite? At least in my experiments Y'Y doesn't seem to ever end up positive definite.
> MapReduce SSVD: provide alternative B-pipeline per B=R' ^{-1} Y'A
> -----------------------------------------------------------------
>
> Key: MAHOUT-797
> URL: https://issues.apache.org/jira/browse/MAHOUT-797
> Project: Mahout
> Issue Type: Improvement
> Affects Versions: 0.5, 0.6
> Reporter: Dmitriy Lyubimov
> Assignee: Dmitriy Lyubimov
> Fix For: Backlog
>
> Attachments: MAHOUT-797.pdf
>
> Original Estimate: 48h
> Remaining Estimate: 48h
>
> Since alternative flow using Cholesky decomposition is extremely easy to add to existing computations of BB', I am thinking of just adding an option that chooses between B-pipeline with QR step and B-pipeline with Y'Y-Cholesky step.
> Ongoing work and some initial code (Y'Y step) is here https://github.com/dlyubimov/mahout-commits/tree/MAHOUT-797.
> I also want to fix what's left unfixed in MAHOUT-638.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira