Hi @ll,
i´m stuck with origin c and hopefully somebody can help. Im currently writing my phd thesis and encountered a problem with data evaluation that has to be solved. Unfortunatly I can´t do it and within my workgroup nobody can help me. Anyways it has to be done..
---
Currently I´m fitting pre-fourier transformed (with another program) autocorrelation data with a spectral function S(q,T,omega). (see code snippet 1 if you are interested).
For a more reliable evaluation I should do the following:
1.) calculate the numeric fourier transformed S(q,T,omega) (which is the autocorrelation function c(q,T,t) following the wiener-kientchine theorem) using literature values for all variables used within S(q,T,omega) but with q,T as input and omega as independent variable. (0 < omega < 100000).
2.) fit the experimental data with the result of 1.):
y = A*B*c(q,T,t)+A*c(q,T,t)^2
(this is related to homodyne and heterodyne contribution to the autocorrelation function)
Summary: I need to know how to let Origin first calculate S(q,T,omega) than apply inverse fourier transform to it. Afterwards fit my data with the modified result c^2 and c for hetero/homodyne contribution.
If anybody could help me with maybe support me with the code needed - you will be mentioned in caps in my thesis ackknowledgement.
code snippet 1
#include <origin.h>
double f(double x, double k, double eta, double T, double sigma)
{
double rho=(1.219-7.17E-4*(T-273.16))*1000;
double nu = eta/rho;
double epsilon = nu*k^2;
double omegas = sqrt(9.81*k+sigma*k^3/rho);
complex D1 = omegas^2-(x+2*1i*epsilon)^2-4*epsilon^2*(1-1i*x/epsilon)^(1/2);
return -2/x*k/rho*Im(D1)/cabs(D1)^2;
}
//---Here´s the non editable part...
{
// Beginning of editable part
y = A*f(x, k, eta, T, sigma);
}
numeric FFT for data evaluation - Need help
Posted - 03/07/2010 : 10:17:24 AM
