Thanks to everyone for this interesting discussion.
I've attached a new file to the MATH-581 JIRA ticket, including your comments
(I hope).
Hope this will get integrated into Commons-Math !
Sebastien
-
To unsubscribe, e-mail: dev-
As your measurements indicate, the allocation of large objects is rarely a
speed issue. It can be a memory fragmentation issue, but a full GC will fix
that as well.
People who are stressed about allocation of result vectors such as what you
are doing are mostly worried about the wrong thing.
201
Le 08/07/2011 12:28, Sébastien Brisard a écrit :
I would go to the natural type we get when building the exception. If we may
build the exception from both types, then I would select RealVector. We do
not intend to perform large processing on elements, but we want to display
them and I think we
On Fri, Jul 08, 2011 at 12:28:23PM +0200, Sébastien Brisard wrote:
> >
> > I would go to the natural type we get when building the exception. If we may
> > build the exception from both types, then I would select RealVector. We do
> > not intend to perform large processing on elements, but we want
>
> I would go to the natural type we get when building the exception. If we
may
> build the exception from both types, then I would select RealVector. We do
> not intend to perform large processing on elements, but we want to display
> them and I think we do have formatting for vectors.
>
> best r
>
> I would go to the natural type we get when building the exception. If we may
> build the exception from both types, then I would select RealVector. We do
> not intend to perform large processing on elements, but we want to display
> them and I think we do have formatting for vectors.
>
> best r
Le 08/07/2011 08:44, Sebastien Brisard a écrit :
Hi everyone,
well the message below did not raise much of an interest... My apology, I
forgot to quote [math] in the title.
Sorry, I forgot to reply in a timely fashion.
Anyway, I have a new proposal regarding this issue. I do not know what you
On Fri, Jul 08, 2011 at 08:44:15AM +0200, Sebastien Brisard wrote:
> Hi everyone,
> well the message below did not raise much of an interest... My apology, I
> forgot to quote [math] in the title.
> Anyway, I have a new proposal regarding this issue. I do not know what your
> view on the subject is
Hi everyone,
well the message below did not raise much of an interest... My apology, I
forgot to quote [math] in the title.
Anyway, I have a new proposal regarding this issue. I do not know what your
view on the subject is, but I tend to think that exceptions should return
references and not deep c
That's great.
There's not much yet, really. It's more of a framework based on the
discussions we've already had. Once this framework has been agreed upon, I'll
commit some more serious stuff.
Sébastien aka "Stéphane" ;)
-
To unsub
Le 02/06/2011 14:32, Sébastien Brisard a écrit :
Dear all,
I've submitted recently some pieces of code relating to iterative linear
solvers (see JIRA MATH-581).
I vwould be grateful for any feedback you could provide. Should I go on
submitting new pieces of code, or do you think it will not fit i
Dear all,
I've submitted recently some pieces of code relating to iterative linear
solvers (see JIRA MATH-581).
I vwould be grateful for any feedback you could provide. Should I go on
submitting new pieces of code, or do you think it will not fit in
commons-math?
Best,
Sébastien
It isn't unusual for Commons Math and Mahout to have similar code. They
have different goals and different compatibility constraints.
Where it works, sharing between the projects is great, though.
On Wed, Apr 13, 2011 at 1:10 AM, Luc Maisonobe wrote:
> I think both Mahout and Apache Commons Mat
Le 13/04/2011 08:22, Sébastien Brisard a écrit :
> Dear All,
> first of all, I should say that I'm new to the Commons-Math community, and I
> therefore apologize for any mistake/inconsistency.
> I'm currently developing a numerical method for the simulation of highly
> heterogeneous materials. This
Le 13/04/2011 09:58, Sebastien Brisard a écrit :
> Do you think MAHOUT would be more appropriate to host such solvers?
Hi Sébastien,
I think both Mahout and Apache Commons Math could benefit from such code.
Luc
> SB
>
> -
> To
Do you think MAHOUT would be more appropriate to host such solvers?
SB
-
To unsubscribe, e-mail: dev-unsubscr...@commons.apache.org
For additional commands, e-mail: dev-h...@commons.apache.org
I have an initial implementation of LSMR that I will be checking into the
Mahout project shortly. If you want to have such an algorithm in commons
math, you would be entirely free to grab that and adapt it for possible
inclusion in commons math (or your own use).
2011/4/12 Sébastien Brisard
> D
Dear All,
first of all, I should say that I'm new to the Commons-Math community, and I
therefore apologize for any mistake/inconsistency.
I'm currently developing a numerical method for the simulation of highly
heterogeneous materials. This requires the inversion of a (big) linear system,
whose mat
18 matches
Mail list logo
