• To: mathgroup at smc.vnet.net
• Subject: [mg5084] Re: Fit in Excel Mathlink
• From: rubin at msu.edu (Paul A. Rubin)
• Date: Fri, 25 Oct 1996 22:48:45 -0400
• Organization: Michigan State University
• Sender: owner-wri-mathgroup at wolfram.com

```In article <53gpva\$5am at dragonfly.wolfram.com>,
Mark Dowell <mark.dowell at jrc.it> wrote:
->Dear All,
->
->Can anyone hellp I'm trying to fit some data under Excel with the
->
->y=a(0)*Exp(-a(1)*(x-400))+a(2)
->
->I just need value for best fit in a(0),a(1),a(2). for a given range in
Excel
->cells can anyone help with the syntax
->
->Mark
->Mark Dowell					Ispra, I-21020, (VA)
->Marine Environment Unit TP 272			Italy.
->Space Applications Institute			E-mail: mark.dowell at jrc.it
->Joint Research Centre				Talk: dowell at biscay.jrc.it
->Ispra Site					Phone: +39-332-789873
->Commission of the European Communities		Fax: +39-332-789034

If you're already in Excel, and you just want the parameter estimates,
and assuming that you want least squares estimates (not maximum
likelihood), why not set up a cell containing the sum of squared errors and
use the Excel Solver to minimize it?

Since nonlinear fits are sensitive to starting estimates for the
parameters, it would help to have good initial guesses.  You might want to
try a lit search on "exponential peeling" and see if you can come up with a
reference.  As I recall (and this dates back a couple of decades), it's a
method for fitting sums of decaying exponentials to data.  Your model is a
special case in that the second term is a constant (or a decaying
exponential that doesn't decay much :-).

-- Paul

**************************************************************************
* Paul A. Rubin                                  Phone: (517) 432-3509   *
* Department of Management                       Fax:   (517) 432-1111   *
* Eli Broad Graduate School of Management        Net:   RUBIN at MSU.EDU    *
* Michigan State University                                              *
* East Lansing, MI  48824-1122  (USA)                                    *
**************************************************************************
Mathematicians are like Frenchmen:  whenever you say something to them,
they translate it into their own language, and at once it is something
entirely different.                                    J. W. v. GOETHE

```

• Prev by Date: Q: About PAUSING in V2.2
• Next by Date: Re: Nonlinear fit to diff eqs.