Re: nonlinear regression

Rusty wrote:
<orangeKDS@xxxxxxx> wrote in message news:1114433365.826801.278000@xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx

I need to find formula for this graph:

http://qrange.150m.com/misc/plot.htm


What is good program for nonlinear regression that could help me?



As the curves resemble exponentials a linear combination of exp(-a(i).x) terms with a(i)'s to be determined might fit well. Prony's method solves for everything. Either Google for it, or its in Hildebrand's rather geriatric numerical analysis book.

rusty


Prony's method is so ill-conditioned that it might also solve the earlier question about random number generators!

Cheers,

Phil Hobbs
.

Relevant Pages

  • More on serendipity
    ... I agree that if one fits the linear ... in the presence of noise if one fit both models would b1 ... equations and nonlinear regression equations is not empty, ...
    (sci.stat.math)
  • Re: nlinfit help
    ... I originally did it as a linear fit using the ... regress function. ... significant errors from the measured data at the same predictors. ...
    (comp.soft-sys.matlab)
  • Re: Deseasonalization and detrending of Keeling curve
    ... fit with a linear gain factor. ... The "fit" is based on a stiff spline ... MODEL STATISTICS AND EQUATION FOR THE CURRENT EQUATION (DETAILS ... Standard Error of the Mean =Standard Dev/ .148530E-01 ...
    (sci.stat.math)
  • Re: forecasting?
    ... increase as you add more polynomial terms, but you will quickly begin to ... A simple straight line clearly does not fit the entire data set very well, ... week 7, or linear in SQRT, leading to 46.9 at week 7. ... regression equation and the RSQ value. ...
    (microsoft.public.excel.worksheet.functions)
  • Re: Piecewise Linear Function of 3d Points
    ... Now I want to fit a piecewise ... at the grid determine now a piecewise linear and continuous interpolant uniquely. ...
    (sci.math.num-analysis)