[Top][All Lists]
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [Help-gsl] Matrix / Vector multiplication
From: |
Jordi Gutierrez Hermoso |
Subject: |
Re: [Help-gsl] Matrix / Vector multiplication |
Date: |
Sun, 11 Feb 2007 11:58:55 -0600 |
On 11/02/07, Patrick Alken <address@hidden> wrote:
The "d" means double precision and I don't off hand know what the ge
means, but you can read the blas docs for more info.
It means general.
Multiplication of a vector by a matrix, and basic linear algebra in
general, is a more delicate task that may be apparent at first sight.
If you're familiar with vectorisation from GNU Octave and related
languages, you may appreciate the need for the BLAs routines,
especially on specialised hardware when it's available.
Golub and Van Loan have a nice discussion of this subject in great
depth, why we need saxpy and gaxpy operations, and why a function like
"gsl_matrix_multiply()" wouldn't be as good as what we already have in
BLAS.
Cheers,
- Jordi G. H.