The Origin Forum
File Exchange
Try Origin for Free
The Origin Forum
Home | Profile | Register | Active Topics | Members | Search | FAQ | Send File to Tech support
Username:
 Password:
Save Password
Forgot your Password? | Admin Options

 All Forums
 Origin Forum
 Origin Forum
 Non-linear fit with low error in fitting paramete		  New Topic  Reply to Topic
 Printer Friendly
Author Previous Topic Topic Next Topic Lock Topic Edit Topic Delete Topic New Topic Reply to Topic

udhaya

India
2 Posts
Posted - 07/22/2023 :  09:08:59 AM  Show Profile  Edit Topic  Reply with Quote  View user's IP address  Delete Topic
I am trying to fit a dataset to an exponential decay function (nonlinear fit). As I start the fitting with the rough estimates of parameters, for the first few cycles of iteration I get a fitting curve which looks visually acceptable. The error values are low (< 5% of the three fitting parameters) and COD (R2) > 0.985. As soon as I increase the iterations by hitting "fit until converge", the resulting final parameters have error values > 100% of the fitting parameters with COD (R2) being > 0.99. I understand, the default criteria of fitting is to achieve COD(R2) as close to one as possible.

Is there a way to achieve low error in fitting parameters at the cost of COD(R2)? It is possible to tell origin to iterate to maintain the error within a limit, while achieving the COD(R2) closest to unity! Will appreciate any help.

Thanks

YimingChen

1592 Posts
Posted - 07/24/2023 :  08:41:12 AM  Show Profile  Edit Reply  Reply with Quote  View user's IP address  Delete Reply
If the error of the fitted parameter is large, it means the fitting function is over-parameterized. You may consider reducing the number of parameters in the model. Please refer to this page.
https://www.originlab.com/doc/Origin-Help/The_Reason_Why_Fail_to_Converge#Over-parameterized_functions

James
Go to Top of Page
  Previous Topic Topic Next Topic Lock Topic Edit Topic Delete Topic New Topic Reply to Topic
 New Topic  Reply to Topic
 Printer Friendly
Jump To:
The Origin Forum © 2020 Originlab Corporation Go To Top Of Page
Snitz Forums 2000