I usually do this in Excel or Matlab, but with OriginPro you can also do it by following this video, the idea is the same just define your formula:
https://www.youtube.com/watch?v=lw5oJwv9JME
BR
Saeed
Dear Sir,
thank you for your reply.....
In given tutorial guass model is applied but i want to use redlich-kister polynomial equation. i have also feed it's equation in origin but it shows link failed in code builder. do you have any idea for this error ?
one thing is that guass model is perfect for spectra type graph. but generally excess property give sigmoid or parabola or polynomial type behavior.
change the fitting coefficients of redlich-kister polynomial equation.
Your function is linear in the coefficient. Now the solution of the problem depend on the number of data (x1,y) you have. If you have 5 values, you can use these points to build a system of 5 linear equations. Otherwise if your data are >5, you should use the least square method http://en.wikipedia.org/wiki/Least_squares which gives you a linear system of 5 equations. Many available software can do the job, like origin, GNU plot, matlab, mathematica, etc. However, it is not difficult to implement it by yourself, you need only a solver of linear equations. Gaussian elimination is OK for this system.
I forgot to say that for linear fitting, the initial guess is not necessary.
Dear patel
Have you got the idea of getting the coefficients RK equation. If so Please do share with us
Thanks
We must first write this equation in standard form in terms of "X" coefficients (it could be possible with "collect" command in Matlab). Then, easily you can fit any array of data to this equation by "van Der Mond" matrix by means of "polyfit" in Matlab or by hand.
Personally, I am using "gnu-plot" (a free, powerful, python base and light weight application) to plot and fit my data.
- Badges
- Science topic
- Similar topics
- Mathematics
- Statistics
- Regression