John Iacona wrote: > So did anything ever happen with the project in the thread referenced by > Luc? It seems like that was quite a while ago and I have not seen any > activity in the area of sparse matrices in a while. But if someone is > already working in this area I would definitely like to collaborate.
I am not aware of any work on this. Your proposal to contribute is very appealing to me, so if you want to start something, please go ahead. I will be glad to review and commit the patches. The best way to do this is to open a JIRA issue at http://issues.apache.org/jira/browse/MATH and to attach patches to this issue. > > As for ideas for what needs to be in the linear package, I think that some > of the most important applications of matrix computation revolve around fast > solvers that exploit the matrix structure effectively (i.e., solvers based > on Cholesky, Triangular, LU, etc. decompositions). Do you think our current architecture is OK for your work ? Can you fit sparse matrices in it easily ? Luc > > Regards, > John > > On Thu, May 22, 2008 at 5:30 AM, Edward J. Yoon <[EMAIL PROTECTED]> wrote: > >> Just FYI, If you interested in the matrix, you can also take a look at >> http://wiki.apache.org/incubator/HamaProposal (Hams has accepted by >> IPMC) >> >> Edward >> >> On Thu, May 22, 2008 at 4:02 PM, <[EMAIL PROTECTED]> wrote: >>> Improving the linear algebra package is a clear need and one of the >> objectives for version 2.0. >>> The link to the (short) thread may be this one: >>> http://markmail.org/message/esfeuzzazz6yeqyk >>> >>> I have also have another discussion ongoing outside from potential users >> of [math] who would like to have fast linear algebra with support for >> vectors. >>> I tried to get involved with Mahout developers who where seeking for some >> math skills at some point, but the thread ended after my last message >> without answers (http://markmail.org/message/a743yyz5ycqr2kdv). The team >> seems to want to develop its own implementation and not being really >> interested in our work. I gave up, but I really think this is a pity, we >> should be able to work together. >>> Perhaps is it time to put some ideas on what we need in our linear >> algebra package to meet most users needs (SVD, sparse matrices, vectors, >> efficiency, perhaps complex numbers ...) >>> Luc >>> > --------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
