[Date Index] [Thread Index] [Author Index]

Re: Spline

• To: mathgroup at smc.vnet.net
• Subject: [mg58307] Re: Spline
• From: Bill Rowe <readnewsciv at earthlink.net>
• Date: Sun, 26 Jun 2005 23:13:09 -0400 (EDT)
• Sender: owner-wri-mathgroup at wolfram.com

On 6/26/05 at 1:33 AM, davidannetts at aapt.net.au (David Annetts)
wrote:

>If you know the values at the endpoints, then tell the spline
>routine. There's no way of extending the spline, since it is only
>defined (in the original) on [.1, .9].

Not true. For a given set of points a natural cubic spline is created by setting the second derivative to 0 at the end points and requiring continuity of both the first and second derivative at all points. So, past the end points, the curve will be a line (continuous first derivative, second and higher derivatives = 0)

However, it is true the package SplineFit will not extend the spline past the end points in the original data set.
--
To reply via email subtract one hundred and four