| T O P I C R E V I E W |
| srira |
Posted - 10/25/2003 : 08:15:31 AM
I have fitted Lorentzian function to few curves.I am not sure how to decide whether the fitting is good.I get the chiQ^2 value around 2 or so.Generally I guess chiq^2 around 1 is the best fit.What are the allowed limits of chiq^2 values to have a better Lorentzian fit?
If any lit. is available on this,kindly let me know.Thanks! |
| 1 L A T E S T R E P L I E S (Newest First) |
| easwar |
Posted - 10/26/2003 : 1:14:53 PM
Hi,
Origin reports reduced chi-sq, which is chi-sq divided by degrees of freedom. The nonlinear fitter minimizes this quantity to obtain best fit to the data with the given model and parameter values. When you continue to perform iterations in the fitter (NLSF or Wizard), once the fit has converged, you will get the message "Chi-sq is not reduced". At this point, you can generate a fitting parameter sheet using the Action->Results menu of the NLSF tool. This parameter sheet will give further information such as R^2, SSR etc.
You will need to make your own interpretation on whether the fit is good or not. Quantities such as chi-sq and R^2 can be used to make this determination. In general, a large chi-sq value is indicative that the dataset and model/parameters is not a good match.
Also, take a look at this article:
http://www.tipmagazine.com/tip/INPHFA/vol-9/iss-2/p24.html
Easwar
OriginLab.
|
|
|
