This is the mail archive of the gsl-discuss@sources.redhat.com mailing list for the GSL project.

Index Nav:	[Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav:	[Date Prev] [Date Next]	[Thread Prev] [Thread Next]

Re: GSL matrix transpose view

To: gsl-discuss at sources dot redhat dot com
Subject: Re: GSL matrix transpose view
From: "E. Robert Tisdale" <edwin at netwood dot net>
Date: Wed, 19 Jul 2000 21:22:19 +0000
References: <39751741.267E10B8@netwood.net> <14710.5590.895075.272072@localhost>

Brian Gough wrote:

> The GSL vector matrix model is designed to be compatible with BLAS,
> which only supports a stride for the trailing dimension.
> That's a design decision based on the fact that
> we're working to the existing standards used by NAG, IMSL etc.

A transpose view is NOT incompatible with BLAS.
For example, the gsl_blas_dgemv function
would simply drop the CBLAS_TRANSPOSE_t argument
and examine the strides to determine whether to process
the matrix or it's transpose.

References:
- GSL matrix transpose view
  - From: E. Robert Tisdale
- Re: GSL matrix transpose view
  - From: Brian Gough

Index Nav:	[Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav:	[Date Prev] [Date Next]	[Thread Prev] [Thread Next]