This is the mail archive of the
mailing list for the GSL project.
Re: Sparse matrix extension
- From: Patrick Alken <alken at colorado dot edu>
- To: <gsl-discuss at sourceware dot org>
- Date: Mon, 15 Feb 2016 10:16:58 -0700
- Subject: Re: Sparse matrix extension
- Authentication-results: sourceware.org; auth=none
- Authentication-results: sourceware.org; dkim=none (message not signed) header.d=none;sourceware.org; dmarc=none action=none header.from=colorado.edu;
- References: <CAMWWPT3uJj4Vrn7ut6+F18gY===zd6+1r1UJhz0hcCj--zwtdg at mail dot gmail dot com> <CAMWWPT1mog0HrviL1tMo=f-rrSc2PhamWDGg7ZYLrVnnqkY3ng at mail dot gmail dot com> <56B77E13 dot 1000306 at colorado dot edu> <CAMWWPT2KTLjufNdwW=y-xaJ1dKkxDNQwgat4DB4ZwKDrK_fZSA at mail dot gmail dot com> <56B7A59D dot 5040707 at colorado dot edu> <CAMWWPT0CX+ti2j6QM8YmHqv+n8yg7sK69dtSdzKOQMFx3pXeBA at mail dot gmail dot com> <56B7B85C dot 10508 at colorado dot edu> <CAMWWPT1ELN9_BFLG=KUQ=Z6rC6ytYGTm0_K-=4-urJ6HGxfFJQ at mail dot gmail dot com> <CAMWWPT0qUccM6cNkX6fWezmj09SdcgYqYzGE3eoqetpuyZW6Vg at mail dot gmail dot com> <56BB5D84 dot 20804 at colorado dot edu> <CAMWWPT0a77RXNkg=a=Y8fJoxbYKpz-7tgvsXGPQNd4X_e=pNag at mail dot gmail dot com> <56BF86FC dot 8010509 at colorado dot edu> <CAMWWPT10XABEnoovh7GgoumYU-mjpGzSzjHOZbDvV=eHgTUSfw at mail dot gmail dot com> <56C0C326 dot 8090701 at colorado dot edu> <56C0C69D dot 1070700 at gladman dot plus dot com> <56C15DC9 dot 8060708 at colorado dot edu> <56C1B1D0 dot 9000303 at gladman dot plus dot com> <CAMWWPT0J9ENRZjJHLO=cxot4DGdSLer+n2HkBVnhFnO0oiVV8g at mail dot gmail dot com> <CAMWWPT08uPGzje-NU6Mt9gkEFbtaK=xxRA2r0Ee+_EvhuwfH_g at mail dot gmail dot com>
- Spamdiagnosticmetadata: NSPM
- Spamdiagnosticoutput: 1:23
On 02/15/2016 06:56 AM, Alexis Tantet wrote:
I've had a quick check at your modifications and all the tests pass.
Some very minor details:
- I find _transpose2 an inexpressive name. What about something like
_transpose_shallow or _transpose_swap.
Well, many routines the linear algebra section append a "2" when the
routine does the same thing but using a different method/algorithm, so I
was trying to be consistent with the naming, and I tried to document
this routine thoroughly in the manual. I will think some more about
whether it would make sense to make a longer descriptive name.
- Regarding _set, wouldn't it be more secure to raise an error when
the matrix dimensions have to be increased? Do you know of
applications in which the matrix dimensions are not known beforehand?
This was originally done since the code was based on the CSparse
library, which did the same thing. You're probably right it makes sense
to do more strict checking...although this code has been through a few
releases so people's programs might break if we change the behavior now.
All the best,
On Mon, Feb 15, 2016 at 12:09 PM, Brian Gladman <firstname.lastname@example.org> wrote:
On 15/02/2016 05:10, Patrick Alken wrote:
I think the problem should be fixed now, can you test the latest git and
let me know?
Yes, all tests now pass.