| Author |
 Topic  |
|
|
dkimia1001
S. Korea
Posts |
 Posted - 01/16/2018 : 1:08:44 PM
|
Origin Ver. and Service Release (Select Help-->About Origin): OriginPro 2018 (64 bit) b.9.5.193
Operating System:Windows 7
Hello,
I have two data sets that need to be fitted simultaneously with two different fitting functions, which share common fitting parameters. What would be the best way of doing so? It seems I can only choose one fitting equation even for global fitting.
For example, Y1=a^b*sin(pi*c*x), Y2=d^b*cos(pi*c*x). Y1 and Y2 are different and the corresponding fitting functions are different as well, but the fitting parameters b and c are shared between the two. So the purpose is to simultaneously fit both data sets and find a, b, c, d that lead to the least error globally.
|
|
|
AmandaLu
439 Posts |
Posted - 01/16/2018 : 9:58:29 PM
|
Hi,
You can combine 2 functions and use a parameter to switch between them. Take functions
Y1=a^b*sin(pi*c*x),
Y2=d^b*cos(pi*c*x)
for example. You can combine them like the following
Y=s*(a^b*sin(pi*c*x)) + (s-1)(d^b*cos(pi*c*x))
when fit with Y1 data, fix s = 1, when fit with Y2 data, fix s = 0. And you can share b and c as needed.
Please refer to the following tutorial for the details:
https://www.originlab.com/doc/Tutorials/Multi-Functions-Global-Fitting
Thanks,
Amanda
OriginLab Technical Service
|
 |
|
| |
Topic |
|
|
|
global fit with shared fit parameters