This is the mail archive of the
gsl-discuss@sources.redhat.com
mailing list for the GSL project.
Re: Root finding page manual suggestion
- To: Francesco Potorti` <pot at gnu dot org>
- Subject: Re: Root finding page manual suggestion
- From: Brian Gough <bjg at network-theory dot co dot uk>
- Date: Fri, 19 Oct 2001 11:08:45 +0100 (BST)
- Cc: gsl-discuss at sources dot redhat dot com
- References: <E15uDcB-00021q-00@pot.cnuce.cnr.it>
Francesco Potorti` writes:
> I've used this criterion in some small numerical program I've written,
> and it works ok. The above means that Newton's method wins only if
> computing f'(x) is more than twice faster than computing f(x), a quite
> rare occurrence in practice. I suggest this fact to be mentioned in the
> manual.
Thanks for the information, I wasn't aware of that. I'll add a note
in the manual.
One factor in favor of Newton's method might be better numerical
stability on roots of higher multiplicity -- the subtraction of
f(a)-f(b) in the secant method could introduce a cancellation error
there.
regards
Brian Gough