MathGroup Archive 2003

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

Search the Archive

Re: 3D Data approximation

  • To: mathgroup at smc.vnet.net
  • Subject: [mg41501] Re: 3D Data approximation
  • From: "Kevin J. McCann" <kmccann at umbc.edu>
  • Date: Fri, 23 May 2003 03:25:11 -0400 (EDT)
  • Organization: University of Maryland, Baltimore County
  • References: <baiabe$dhn$1@smc.vnet.net>
  • Sender: owner-wri-mathgroup at wolfram.com

Can you not just check the residuals for the FIT result and the book result
to see which gives a better (smaller) residual? I am making the assumption
that you are interested in the minimization of the least-squares error.

Kevin

--
Kevin J. McCann
Joint Center for Earth Systems Technology (JCET)
Department of Physics
UMBC
Baltimore MD 21250



"Martin Wieloch" <marwiel at lycos.com> wrote in message
news:baiabe$dhn$1 at smc.vnet.net...
> Dear All
>
> Please help me if you can in solving my problem with data
> approximation.
> It is probably simple but I have had Mathematica just for one week.
> I have got a set of experimental results {x,y,z} and I am trying
> to approximate (the least square method) them with a function of
> the form: a +a1x+a2x2+a3x3+a4x4+b1y+b2y2+b3y3+b4y4+cxy+c1x2y2
>
> Is there is any better function than FIT.  I used the fit function
> to solve an example, but my results are slightly different from
> those I found in a book-I have obtained higher coefficients a, b, c.
>
> I will be very much grateful if someone can help me.
>
> Regards,
> Martin
>
>




  • Prev by Date: Re: Packed array and ReadList
  • Next by Date: RE: spanning braces in graphics?
  • Previous by thread: 3D Data approximation
  • Next by thread: Re: 3D Data approximation