| T O P I C R E V I E W |
| taniap |
Posted - 12/05/2023 : 09:07:47 AM
Hello!
I am trying to fit some arrhenius plots with the non linear fitting of the origin but it keeps giving me a very high value of error. The equation that I'm using is the double exponential:
(I= IO/(1+Aexp(-E1/kbT) + Bexp(-E2/kbT)).
After trying to define the single exponential with good values of error, when I am trying to use this defined value in the double exponential the error keeps gaining high values. The tolerance factor that is used is 1E-9 and the max number of iterations is 400. Should I change something in these values? Can I reduce the error within 10%? |
| 3 L A T E S T R E P L I E S (Newest First) |
| YimingChen |
Posted - 12/06/2023 : 11:18:26 AM
Changing the tolerance factor won't help if the fitting is already converged. Could you share your data here? Thanks.
James |
| taniap |
Posted - 12/06/2023 : 06:17:58 AM
quote:
Originally posted by YimingChen
Have you manually set the initial values of the fitting parameters? Please refer to the page:
https://www.originlab.com/doc/Origin-Help/The_Reason_Why_Fail_to_Converge#Poor_initial_parameter_values
James
Hello,
yes i have tried fixing some of the initial parameters. I wonder if by changing the tolerance factor it will still be the same. In addition, do you know if there is any option to limit the error while fitting by defining the value in the settings of the non-linear fit? |
| YimingChen |
Posted - 12/05/2023 : 09:37:09 AM
Have you manually set the initial values of the fitting parameters? Please refer to the page:
https://www.originlab.com/doc/Origin-Help/The_Reason_Why_Fail_to_Converge#Poor_initial_parameter_values
James |
