| T O P I C R E V I E W |
| Alisa_LGA |
Posted - 10/21/2018 : 05:04:32 AM
Hi guys!
Could you please help me to find an approximation of experimental data with the non-linear equation:
k1*y+k2*x+ln((k3*x)/(x-y))-k4=0
I need to find constants k1,k2,k3 and k4. I know a set of values (around 100 values) for x andy.
Which function should I use in Origin to find the constants?
Any of your replies are very appreciated!
Origin Ver.2018 and Service Release (Select Help-->About Origin):
Operating System:Win 10, 64
|
| 4 L A T E S T R E P L I E S (Newest First) |
| yuki_wu |
Posted - 10/25/2018 : 02:23:56 AM
Hi,
1# Please take a look at this page to find the reason why fail to converge:
https://www.originlab.com/doc/Origin-Help/The_Reason_Why_Fail_to_Converge
2# I am sorry I don’t understand your question, but you could definitely define a fitting function with multiple exponential functions in Origin, for example, the built-in functions ExpGro2, ExpGro3 and etc. You could press F9 to open Fitting Function Organizer and expand Exponential tree node to look up the functions I mentioned.
Regards,
Yuki
OriginLab
|
| Alisa_LGA |
Posted - 10/24/2018 : 11:26:34 AM
quote:
Originally posted by yuki_wu
Hi,
Would you like to perform fitting on your dataset? If so, you could defiend a user-defined fitting function first and then perform fitting using that function. Please take a look at this page, it should help:
https://www.originlab.com/doc/Tutorials/UserDef-FitFunc
Regards,
Yuki
OriginLab
Moreover, I found that Origin cannot describe the combination of exponential functions at all.
When I built a combination of 2 exponential functions (with known parameters), I cannot describe it with the same formula unless I am putting exactly those parameters I inserted.
P.S. Sometimes, it cannot be described even with the same parameters initially inserted.
:( |
| Alisa_LGA |
Posted - 10/24/2018 : 09:57:55 AM
Hi,
thank you for the reply!
But unfortunately, it just does not want to fit the data and optimize the constants.
I checked everything manually (put the guess values into the column) and it calculates fine. But the non-linear curve fit gives me:
"(4) Fit did not converge - reason unknown."
Please see the attached image.
 |
| yuki_wu |
Posted - 10/22/2018 : 02:22:17 AM
Hi,
Would you like to perform fitting on your dataset? If so, you could defiend a user-defined fitting function first and then perform fitting using that function. Please take a look at this page, it should help:
https://www.originlab.com/doc/Tutorials/UserDef-FitFunc
Regards,
Yuki
OriginLab
|
|
|
