[
http://issues.apache.org/jira/browse/MATH-157?page=comments#action_12446796 ]
Tyler Ward commented on MATH-157:
-
Not bad. Looks like everything is in the right place, modulo a transpose or two
perhaps, but looks good. The eigenvector
[
http://issues.apache.org/jira/browse/MATH-157?page=comments#action_12446426 ]
Tyler Ward commented on MATH-157:
-
One more thing. I don't think you can just use the Q from the QR reduction
(slipped my mind earlier). Try it, you'll s
[
http://issues.apache.org/jira/browse/MATH-157?page=comments#action_12446423 ]
Tyler Ward commented on MATH-157:
-
Also, about testing. I've found that the best way to test this is just with
random matrices. If you can use, say, 10,000 r
[
http://issues.apache.org/jira/browse/MATH-157?page=comments#action_12446421 ]
Tyler Ward commented on MATH-157:
-
I think I see the confusion. Take a look at my first comment here. You don't
want to invert (or find eigenvectors of
[
http://issues.apache.org/jira/browse/MATH-157?page=comments#action_12446155 ]
Tyler Ward commented on MATH-157:
-
A few notes.
You don't have to compute the eigenvectors of both mTm and mmT, just do which
ever one is smaller. You kno
I think a very significant part of this problem is just the profusion
of jars anyway. The real incompatibility you're worried about is when
two projects are using different versions of subcomponents, or at
least the multitude of subcomponents makes it that much more likely.
Commons isn't
[
http://issues.apache.org/jira/browse/MATH-157?page=comments#action_12431127 ]
Tyler Ward commented on MATH-157:
-
Forgive me if this seems like condescension, you can find this in any decent
math book, and it is about one of the most well
Add support for SVD.
Key: MATH-157
URL: http://issues.apache.org/jira/browse/MATH-157
Project: Commons Math
Issue Type: New Feature
Reporter: Tyler Ward
SVD is probably the most important feature in any
Issue Type: Improvement
Reporter: Tyler Ward
Hi guys, I noticed that your brent solver isn't using the initial guess given
by the user. This can often radically improve the performance of the solver. In
my tests, it improved it by roughly 30% or more, with a decent guess.
Basic
I'm going to send in basically the textbook description of SVD, in case
someone wants to implement it. If you guys get your householder reductions,
then that's the most difficult part. I have code for all of this, but it is
held up by legal, so I'll just give you guys the basic ideas instead.
For
Hi guys, I noticed that your brent solver isn't using the initial guess
given by the user. This can often radically improve the performance of the
solver. In my tests, it improved it by roughly 30% or more, with a decent
guess. Basically, we should try the guess first. If that's close enough,
rret
fiability (availability of test data or ability to validate
against other packages).
See http://jakarta.apache.org/commons/math/developers.html for more
general information about contributing to [math].
On 7/17/06, Tyler Ward <[EMAIL PROTECTED]> wrote:
> This is why the householder transfor
here.
How should I send this code in, should I just attach the files to this email
thread?
-Tyler
On 7/17/06, Simon Kitching <[EMAIL PROTECTED]> wrote:
On Mon, 2006-07-17 at 06:50 +0200, Luc Maisonobe wrote:
> Tyler Ward wrote :
> > Hi guys, I haven't sent mail to this
Hi guys, I haven't sent mail to this list before. I've noticed that apache
doesn't have an SVD algorithm, so I put one together. Are you guys
interested?
This is entirely of my creation, I didn't copy or examine code from
anywhere. I wrote all the transformations by hand just browsing through
lin
14 matches
Mail list logo
