Hello ! Respectful Greetings ! I’m trying to fit a data set, plotted through TGraph, using the Fit Panel and with a “user” defined function. The same is a Lorentzian and is being expressed as ([0]*[1]**2)/([1]**2+(x-[2])**2). There are 512 data points with several absorption “peaks” and, to begin…

There is an undeclared identifier in your code. Can you post a macro example reproducing the problem ?

[fitpanel] Thank you very much for your response. I have tried capturing the “look” of the issue and the same is attached herewith. Best regards.

Thanks for the screen dump, I let @moneta give you some hints. I see that selecting the "user’ choice in the Type menu produces the errors you mentioned earlier.

Gratitude ! I’ll eagerly look forward to it. However, as I mentioned earlier, those errors that appear in the terminal following the choice of the “user” option is probably not afflicting the fit procedure. I have exercised the same option for simpler fits.

Hi, In the Fit panel, if you want to type your formula, you need to fill it in the texts field above selected, as I see in your screen shot. But afterwards you need to type enter, to pass the formula to the fit panel. I see in the screen shot that you have “user” as selected. This is not correct. …

For this kind of data, you might also like this old thread: [image] Fit unable to minimize ROOT Hello, my fitting routine seems to be unable to produce any proper results. I am not that experienced yet, so it might be a rather simple mistake, but please take a look none…

Thank you very much, for this insight. I’ve implemented the same. However, I don’t think this is the issue. Since, I have been able to fit other simpler (linear) data through the “user” defined function and even without passing the formula. Nevertheless, many thanks for the education and I’ll do i…

Gratitude for sharing the thread. I’ll try implementing the function therein. Best regards.

Problem with Fitting Data using Fit Panel

Newbie

Wile_E_Coyote March 17, 2021, 11:52am 7

For this kind of data, you might also like this old thread: