Re: [boost] different matrix library?

Subject: Re: [boost] different matrix library?
From: joel (joel.falcou_at_[hidden])
Date: 2009-08-17 03:11:06

> Would it be possible to drop the user-unfriendly stuff? :-) Positive 
> definite alone already implies a vector.
>
> x = solve( positive_definite( A ), b )
>
> should do it :-)
>   
We can provide such short-cut of course. as_matrix is needed to apss the 
complete settings in case like

// acces a vector as a diagonal FORTRAN one by lazyly reindex it & 
reshape it
as_matrix<settings(diagonal,FORTRAN_indexing)(v);
> Of course, pick that, then :-) I guess CUBLAS will be faster than SIMD :-)
I'll tell you this ina bout 8 month when the research project we're 
running on exactly that is complete.

> I see NT2 is under LGPL. Are you considering submitting it for review / 
> changing the license. 
FYI, v3.x o NT2 (the one on the cooking plant) is using Boost License 
and not LGPL anymore.

I started by proposing slicing NT2 components into boost library 
(Boost.SIMD is one).
We *will* end up  boostifying NT2 and submit it for review but that's a 
long term goal (read 3-6 months).
Currently, our deadline is refurbishing actual code so our own user get 
an update NT2.
Then ONLY after that , we'll do NT2->Boost.

Note that I'm really shy on this considering the early response I got 
from just proposing Boost.SIMD.

That's why I also proposed DE to team up so we can brainstorm and don't 
dilute effort as we already have a huge code base that was working even 
if it was ugly.

> Suppose in that case I would like to contribute ~60k 
> lines of generated code to glue against (the whole of) BLAS and LAPACK. 
>   
We have such glue already but sharing/cheking/getting better is of 
course good.

> But, I'm interested in the syntax, first. I think it is probably one of the 
> most important pieces for a user to accept/reject a library. Do you have 
> some documentation on how to write more complex expressions?
There is an old documentation (NT2 is beign rewritten form ground up 
with proto) that show most of what was able at the point
on the sourceforge site.

-- 
___________________________________________
Joel Falcou - Assistant Professor
PARALL Team - LRI - Universite Paris Sud XI
Tel : [phone]