[SciPy-dev] Why does orth use svd instead of QR ?

David Cournapeau david@silveregg.co...
Fri Feb 5 02:33:49 CST 2010


Gael Varoquaux wrote:
> On Fri, Feb 05, 2010 at 04:18:02PM +0900, David Cournapeau wrote:
>> So would be it ok to use this column-rotated QR in place of svd for 
>> every case in orth ? I would have to check that QR with column rotation 
>> is still significantly faster than svd, but I would surprised if if were 
>> not the case. QR has also the advantage of being implemented in PLASMA 
>> already contrary to eigen/svd solvers,
> 
> Out of curiosity, what's PLASMA, in this context?

http://icl.cs.utk.edu/projectsfiles/plasma/html/README.html

"""
The main purpose of PLASMA is to address the performance shortcomings of 
the LAPACK and ScaLAPACK libraries on multicore processors and 
multi-socket systems of multicore processors. PLASMA provides routines 
to solve dense general systems of linear equations, symmetric positive 
definite systems of linear equations and linear least squares problems, 
using LU, Cholesky, QR and LQ factorizations. Real arithmetic and 
complex arithmetic are supported in both single precision and double 
precision.
"""

It is under BSD license, and as a bonus point, may be compiled easily on 
windows (MS is one of the sponsor of the project). The main drawback I 
can see is that it requires a serial BLAS, and ATLAS cannot be switched 
dynamically between serial and parallel (you have to relink).

I am hoping to provide a basic set of wrappers for scipy,

cheers,

David


More information about the SciPy-Dev mailing list