From: eachus@spectre.mitre.org (Robert I. Eachus)
Subject: Re: Help with Translation, please!
Date: 1996/03/15
Date: 1996-03-15T00:00:00+00:00 [thread overview]
Message-ID: <EACHUS.96Mar15160821@spectre.mitre.org> (raw)
In-Reply-To: 4h2gu9$hp6@watnews1.watson.ibm.com
In article <4hvvm2$ird@ds8.scri.fsu.edu> jac@ds8.scri.fsu.edu (Jim Carr) writes:
eachus@spectre.mitre.org (Robert I. Eachus) writes:
>
> .... The same holds for factorials. They can be generated
>exactly in (small) constant time and space. (Actually, if you are not
>using floating point, the time is proportional to the square of the
>length of the result...)
> Make up your mind. ;-)
> Constant or square? A loop is better than linear in the length.
It depends on whether you use a computation model where arithmetic
operations require constant time or require time proportional to the
length of the operands. In the case of factorial it matters.
> And your comment about a sqrt based method, if one exists, seems
> to ignore the extra time required -vs- multiplication. Anyway,
> my main application looks up log(n!) so its a moot point.
Well since no one took the challenge, how about an extra hint.
The ONE square root you need is the square root of five..
--
Robert I. Eachus
with Standard_Disclaimer;
use Standard_Disclaimer;
function Message (Text: in Clever_Ideas) return Better_Ideas is...
next parent reply other threads:[~1996-03-15 0:00 UTC|newest]
Thread overview: 3+ messages / expand[flat|nested] mbox.gz Atom feed top
[not found] <4gvta2$531@ds8.scri.fsu.edu>
[not found] ` <4h2gu9$hp6@watnews1.watson.ibm.com>
1996-03-15 0:00 ` Robert I. Eachus [this message]
1996-02-26 0:00 Help with Translation, please! Robert Gelb
1996-02-27 0:00 ` Pascal OBRY
replies disabled
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox