| Author |
 Topic  |
|
|
vmukund
USA
4 Posts |
 Posted - 07/28/2011 : 09:20:13 AM
|
Origin Ver. 8 and Service Release6 (Select Help-->About Origin):
Operating System:Windows 7
Hi,
Have been using Origin8 for a year now for non-linear fitting. Presently i have some data which does not have a constant baseline(the peaks are on a incline) and Origin doesnt fits the plots well.
1. I tried defining a user-defined fitting function using the fitting function organizer as y=y0+b*x+ (2*A/PI)*(w/(4*(x-xc)^2 + w^2)).This is to get the slope of the background(baseline). But this doesnt seem to do the job. Also i am able to fit the data by going to Analysis->Fitting->Nonlinear fit->Open Dialog box.
Am not able to use a program(a very simple code to call the user-defined function & use the nlsf.fit and related parameter initialization) to do so.When i run the program, it does not take the replicas into consideration but fits only one peak with lorentzian and not the line.I tried changing the number of replica in the controls-section of the user-defined-function's FDF file but its not fitting.
2.Another idea i have not tried so far is to do piecewise fitting with linear-fit to get the correct slope & intercept.But here again my data is such that the peaks position are not same and a code will have to figure out to cut the data to do the linear fit. Is there a built-in function to do so? or how to go about writing a simple code for piecewise linear fit?
Thanks,
VM |
|
|
easwar
USA
1964 Posts |
Posted - 07/28/2011 : 09:57:25 AM
|
Hi VM,
Looks like you have Standard Origin and not Pro. In the Professional version, the Peak Analyzer tool has capability to fit a baseline along with the peaks, allowing you to vary the baseline AND peak parameters during the final fitting process.
The other option is to use the Baseline goal in your Standard-version Peak Analyzer, and then subtract your baseline, and then fit the baseline-corrected data with NLFit and replicas.
You can download the demo for the latest version 8.5.1, where there is also a Quick Fit gadget which makes fitting with baseline and multiple peaks much easier than using Peak Analyzer, so that is worth trying.
Easwar
OriginLab |
 |
|
|
vmukund
USA
4 Posts |
Posted - 07/28/2011 : 1:59:30 PM
|
Hi Easwar,
Thanks for the reply.But i dont see any difference in the fitting(between Origin8 and Origin8.5) for my data.
Please let me know as to how i can fit my data(2 peaks on a incline) to a Lorentzian whose background is a straight line(with slope and intercept)in Origin 8.
Like i said before,i have a user-defined fitting function as y=y0+b*x+ (2*A/PI)*(w/(4*(x-xc)^2 + w^2)).It does not find slope and it does not fit more than one peak.
Thanks,
VM. |
|
|
|
easwar
USA
1964 Posts |
Posted - 07/28/2011 : 2:16:37 PM
|
Hi VM,
Please send your OPJ with the data to tech support and refer to this post. We can then send you the OPJ back with fit results. You can click on the "Send File to Tech Support" link on the top right of this page, to send the file.
Easwar
OriginLab |
|
|
| |
Topic |
|
|
|
Fitting using Lorentzian and non-constant linear b