| T O P I C R E V I E W |
| Dominik Paulkowski |
Posted - 04/16/2007 : 10:37:51 AM
Origin Version (Select Help-->About Origin): Origin7.5G SR6 (v7.5886)
Operating System: Windows XP
Hi there!
My question aims at the non-linear fit modul of Origin. When I use the same user-difened fit function with the same initial condition in different programms, I got different coefficients/results. Both programms use the Levenberg-Marquardt algorithm to fit the data.
The curve fit with Origin looks very well. But I cannot explain the reference material. So I could not use this values for the right sample. The other programm have different coefficients and can all this. But I want to use Origin to fit my data.
Why can there be so different results? Any ideas?
A tolerence of 1E-190 in Origin should be right enough. Chi^2 is in both programs a very big value. R^2 is near 1 in both programms.
Ciao,
Dominik
-------------------
:-Dipl.-Phys. Dominik Paulkowski
Fraunhofer Institute for Surface Engineering and Thin Films (IST)
Braunschweig
Germany |
| 1 L A T E S T R E P L I E S (Newest First) |
| larry_lan |
Posted - 04/16/2007 : 11:23:50 AM
Hi Dominik:
First of all, it's normal that results from different software are slightly different.
If you think the results differ too much, please send your testing data, the fitting function, the name of the other software, and describe your question in detail to tech@originlab.com, we will try to figure out what's the problem.
Thanks!
Larry
OriginLab Technical Services |
|
|
