[mlpack-svn] [mlpack] don't hardcore Evaluate for the lovasz theta objective function (#371)

[Ryan Curtin]

My understanding is that this commit corrects the objective returned by LRSDPFunction::Evaluate(), which before was incorrect, but in the scheme of things this doesn't really matter much since AugLagrangian uses a specialization of Evaluate() anyway.  If that understanding is wrong, let me know, but that's what it looks like to me.  Maybe when I wrote it I assumed that C = ones(n, n), which I think is true for the problem I was writing it for at the time, but I don't feel like bothering to look it up since it's been so long...

Thanks for the fix!  This will make someone's life easier when they decide to solve LRSDPFunction with something that's not the Augmented Lagrangian method. :)

---
Reply to this email directly or view it on GitHub:
https://github.com/mlpack/mlpack/pull/371#issuecomment-68584290
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://mailman.cc.gatech.edu/pipermail/mlpack-git/attachments/20150102/d6518a82/attachment.html>