Date: Tue, 04 Aug 2009 08:46:14 +0200
From: Mogens Kjaer <mk@xxxxxx>
Subject: Re: OFF-TOPIC: Algol 60 guru required
To: "Community assistance, encouragement, and advice for using
Fedora." <fedora-list@xxxxxxxxxx>
Message-ID: <4A77D936.10404@xxxxxx>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
On 08/04/2009 08:28 AM, Hiisi wrote:
....
> D(N+1):=N(N+2):=0.0;
...
> It's accessible here:
http://portal.acm.org/citation.cfm?id=355694.355701
Maybe a mistyping in the .gz file on that page?
Is the original code published in CACM available as a PDF
file somewhere?
Mogens
It isn't. Or at least I couldn't find it.
Date: Tue, 04 Aug 2009 01:20:09 -0700
From: Jonathan Ryshpan <jonrysh@xxxxxxxxxxx>
The same function is available in FORTRAN at:
http://portal.acm.org/citation.cfm?doid=357456.357465 -> 600.gz
These pages refer to the actual papers as .pdf files and to the
algorithms as .gz files, both downloadable.
The FORTRAN implementation doesn't seem to have the perplexing line in
it at all (if present it would be around line 296 of the FORTRAN). I
suspect that the line is a typo, and should read:
D(N+1):=D(N+2):=0.0;
or some such thing. It appears to be initialization for the following
lines, which are:
FOR I:=N STEP -1 UNTIL N1 DO
D(I):=(D(I)-C(I)*D(I+1)-D(I+2))*B(I);
The FORTRAN takes advantage of the fact that (if my rusty FORTRAN is
correct) in FORTRAN arrays are initialized to 0.0 by the run time
system, so the initialization is not necessary. To verify all this
someone will have to take a careful look at the original math in
the .pdf files.
(Fun, but *really* OT.)
You're right. And the person, who needs to verify all this is me. I just
was trying to do it easy. Now I understand - it's impossible. Need to
write the whole program from scratch. By the way, I also have to use
another algorithm. The initial procedure builds quintic natural spline.
Semi-local smoothing spline would suit better to my task. But that's too
more as for this list, as for this off-topic thread...
--
Hiisi.
Registered Linux User #487982. Be counted at: http://counter.li.org/
--
fedora-list mailing list
fedora-list@xxxxxxxxxx
To unsubscribe: https://www.redhat.com/mailman/listinfo/fedora-list
Guidelines: http://fedoraproject.org/wiki/Communicate/MailingListGuidelines
