| Author |
 Topic  |
|
|
zaneslovakia
Slovakia
7 Posts |
 Posted - 10/14/2014 : 11:02:46 AM
|
Origin Ver. and Service Release (Select Help-->About Origin):OriginPro 8
Operating System:Windows 7
Hello everyone!
I need to fit my spectra that has 5 peaks: four of them have Lorentzian line shape, and one has Gaussian line shape. I thought to make an user-defined function (Analysis/Non linear curve fit) and came to the point where I need help. My question is: what should I write in the Fitting function organizer window under the field ''Function''?
Any help will be strongly appreciated! |
|
|
greg
USA
1378 Posts |
Posted - 10/14/2014 : 12:34:13 PM
|
No need for some custom function.
First, make sure you are running the latest Service Release of version 8 : OriginPro 8.0.63.988 SR6
Run Help : Check for Updates if need be.
Select your data and choose Analysis : Peaks and Baseline : Peak Analyzer.
Use this as a guide:
http://www.originlab.com/Index.aspx?go=Support/VideoTutorials&pid=1883
When you get to the "Fit Peaks" page, click "Fit Control" where you can select individual functions for individual peaks. |
 |
|
|
zaneslovakia
Slovakia
7 Posts |
Posted - 10/23/2014 : 05:02:25 AM
|
Unfortunately I couldn't follow the steps suggested by Greg. However, I made a custom fitting function in Origin and in Fitting function organizer i put function:
y=y0 + (A1/(w1*sqrt(PI/2)))*exp(-2*((x-xc1)/w1)^2) +2*(A2/PI)*(w2/(4*(x-xc2)^2 + w2^2)) +2*(A3/PI)*(w3/(4*(x-xc3)^2 + w3^2)) +2*(A4/PI)*(w4/(4*(x-xc4)^2 + w4^2)) +2*(A5/PI)*(w5/(4*(x-xc5)^2 + w5^2))
and defined parameters:
y0,A1,xc1,w1,A2,w2,xc2,A3,w3,xc3,A4,w4,xc4,A5,w5,xc5. I left the ''x'' as a independent variable and ''y'' as a dependent. When I compiled the function everything went well, but I was unable to fit my curve with this custom function. Any suggestion about my further steps? |
|
|
|
zaneslovakia
Slovakia
7 Posts |
Posted - 10/23/2014 : 10:38:49 AM
|
To explain my problem a bit in detail: When I enter parameters for position of maximums and width of peaks, after first iteration they change to unreasonable values (6,365E39 etc.) and the final fit result is usually one straight line. The message that appears is: Fit did not converge - mutual dependency exists between parameters.
You may have overparameterized the fitting function. Fixing one of them may eliminate this problem.
|
|
|
| |
Topic |
|
|
|
user defined fitting function (Lorentz + Gauss)