[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: find first and last
From: |
David Bateman |
Subject: |
Re: find first and last |
Date: |
Wed, 27 Sep 2006 09:42:48 +0200 |
User-agent: |
Thunderbird 1.5.0.7 (X11/20060921) |
John W. Eaton wrote:
> On 27-Sep-2006, David Bateman wrote:
>
> | John W. Eaton wrote:
> | > Oh, it would also be nice if spfind also handled the N and
> | > "last"/"first" arguments. So let me register that as a feature
> | > request.
> | >
> | > jwe
> | >
> | >
> |
> | What about something like the attached.
>
> If this provides compatibility with Matlab's find function for sparse
> matrices, then I'd say check it in as a replacement for the current
> spfind. I see that the current version may also return NR and NC. Is
> there any need to preserve that behavior?
>
> jwe
>
Yes it seems to be compatible with matlab.
In fact yes there is a need to keep that since some of the scripts that
depend on spfind use that feature, Otherwise I'd need to replace the
spfind by a size and an spfind. Easier just to add the additional arguments.
I'd also like to change the direction of the search for "last" to avoid
cache thrashing, and I see a trivial change to allow it. I'll submit a
patch later, but I think it might be better to split spfind out of sparse.cc
D.
- find first and last, (continued)
- find first and last, John W. Eaton, 2006/09/26
- Re: find first and last, Bill Denney, 2006/09/26
- Re: find first and last, David Bateman, 2006/09/26
- Re: find first and last, Bill Denney, 2006/09/26
- Re: find first and last, John W. Eaton, 2006/09/26
- Re: find first and last, Bill Denney, 2006/09/26
- Re: find first and last, John W. Eaton, 2006/09/26
- Re: find first and last, John W. Eaton, 2006/09/26
- Re: find first and last, David Bateman, 2006/09/26
- Re: find first and last, John W. Eaton, 2006/09/27
- Re: find first and last,
David Bateman <=
- Re: find first and last, David Bateman, 2006/09/27
- Re: find first and last, John W. Eaton, 2006/09/27