MathGroup Archive 2005

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

Search the Archive

Re: Why won't Hornerwork?

  • To: mathgroup at smc.vnet.net
  • Subject: [mg59097] Re: Why won't Hornerwork?
  • From: Bill Rowe <readnewsciv at earthlink.net>
  • Date: Thu, 28 Jul 2005 02:28:16 -0400 (EDT)
  • Sender: owner-wri-mathgroup at wolfram.com

On 7/27/05 at 1:24 AM, carlw at u.washington.edu (Carl K. Woll) wrote:

>> In:= Fit[ {{1,200},{200,0},{400,200}}, {1,x,x^2}, x]


>Note that if you are interested in doing polynomial fits, then
>another option is to use the package NumericalMath`PolynomialFit.

In fact, it would be better to use the package NumericalMath`PolynomialFit` than Fit. Powers of x do not form a good set of basis functions for regression. They lead to ill conditioned matrices and numerical instability. The problem becomes more severe as the degree of the polynomial fitted increases. The package NumericalMath`PolynomialFit` is designed to avoid this issue.
--
To reply via email subtract one hundred and four


  • Prev by Date: Re: Simplification question
  • Next by Date: Re: LegendreP error (bug?) in Mathematica
  • Previous by thread: Re: Why won't Hornerwork?
  • Next by thread: "Substract one and add one" algorithm