Fit Experimental Data with MATLAB: Best Practices

[sounouhid]

Hello everyone
help please
I need to fit experimental data with a theoretical model using the MATLAB software what is the best way to do it?
thx for advance

 

[jedishrfu]

Why not try plotting the experimental data vs the theoretical data to see where they differ?

 

[sounouhid]

thank you
but i don't have the the theoretical data .i just have the model ,I'm looking for the fit parameters

 

[jedishrfu]

Can't you run the model with input parameters matching the ones recorded when you did the experiment and then plot the relevant model output?

 

[sounouhid]

yes i did it using the least squares curve fit ,but the fit parameters are not constant they are function of the input.

 

[JJacquelin]

@ sounouhid : One cannot answer to your question without a clear description of your problem. What kind of "theoretical model" are you talking about ? What kind of data ? And more details on the context of the problem are essential. The fitting methods are numerous and depends on the context. A whole book should not large enough to answer if the problem is not restricted. For example, some procedures of fitting to several models are available in : https://fr.scribd.com/doc/14674814/Regressions-et-equations-integrales .(paper in French with some parts in English. All procedures are understandable in English)