| Author |
 Topic  |
|
|
ararix
S. Korea
Posts |
 Posted - 02/01/2016 : 3:25:11 PM
|
Origin Ver. and Service Release (Select Help-->About Origin):
Operating SystemOrigin Ver. and Service Release (Select Help-->About Origin):
Operating System:Window 7
Hi, I am trying to fit my peak functions double sided exponentials.
y(x) = A*Exp(-|x-x0|/xc)
To consider Gaussian broadening of the peak induced by the experimental system.
I used Integral functions
Integration variable:t
Argument : x, IRF, x0 , xc
Return exp(-(x-t)^2/(2*IRF^2))*exp(-abs(t-x0)/xc);
Also defined the main functions and parameters as below.
Parameters : A=1, IRF=0.1, x0=0, xc=3,
Integral elements : Lower lmt = -inf, upper lmt : inf
x=x, IRF=IRF, x0=x0, xc=xc
Main function
y= A*integral(Integral, -inf ,inf ,x ,IRF ,x0, xc)
With this fitting, the function fit the data, but the fitted curve has several regions with a null value (0). Therefore, the fitted curve looks discrete. As shown below (the actual data and fitting functions are more complicate than I describe above, but basically the same.)
These null points disappear if I increase the IRF value from 0.1 to 0.5. So I guess this phenomena are affected by some integration time steps.
Could you let me know what causes this and how can I solve this unexpecting discontinuity fitted curve with integral functions?
 : |
|
|
jasonzhao
China
262 Posts |
Posted - 02/02/2016 : 04:55:04 AM
|
Hello,
Would you please send us the OPJ and FDF file via tech@originlab.com and quote this post.
Best regards!
Jason
OriginLab Technical Service |
 |
|
| |
Topic |
|
|
|
Fitting curve including integrals