[
https://issues.apache.org/jira/browse/MATH-1482?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16835942#comment-16835942
]
Gilles commented on MATH-1482:
------------------------------
Thanks for your proposal.
It just happens that this part of the code is going to be refactored and ported
to a new ["Commons
Statistics"|http://commons.apache.org/proper/commons-statistics/] component.
This a GSoC project being discussed right now on the "dev" mailing list. And
your use-case is certainly welcome in order to shape the new design. You can
make sure that it will be taken into account by subscribing to the ML and start
a discussion over there.
> Pull request for GLSMultipleLinearRegression
> --------------------------------------------
>
> Key: MATH-1482
> URL: https://issues.apache.org/jira/browse/MATH-1482
> Project: Commons Math
> Issue Type: Improvement
> Reporter: Elena Kartysheva
> Priority: Trivial
>
> I would like to propose a pull request implementing an option to use variance
> vector instead of covariance matrix. It allows users to avoid unnecessary
> memory usage and excessive computation in case of uncorrelated but
> heteroscedastic errors thus making it possible to work with huge input
> matrices. Using variance vector in such cases allows to reduce time
> complexity from O(n^2) to just O(n) (where n is a number of observations) and
> dramatically reduce memory usage. For example, in my practice arose a need to
> train generalized linear model. Usage of Iteratively reweighted least squares
> algorithm requires weighted regression with more than a million observations.
> Current implementation would require approximately 12 terabytes of memory
> while patched version needs only 8 megabytes. Since IRLS is iterative
> algorithm a million-times complexity reduction is also pretty handy.
> https://github.com/apache/commons-math/pull/106
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
