Best software to fit molecular spectroscopy data

[kelly0303]

Hello! I have some data from a molecular spectroscopy experiment, containing vibrational and rotational spectra, and I want to fit the peaks with Voigt profiles (one for each peak) in order to obtain the centers of the peaks. Do you know any software suitable for this kind of fit? I usually use Python (scipy, lmfit), but so far it didn't really work, as there are many peaks in one spectrum, so many parameters to fit for and I think that the fit gets stuck in a local minima, before reaching the actual parameters (but if you know a python package good enough for what I need, I am willing to try). Also, should I fit all the spectrum at once, or one peak at a time. If I do one peak at a time I get rid of the dimensionality problem, but the fit might not be as accurate as when I fit everything at once. Thank you!

 

[NateD]

The best software package for this kind of fitting problem is probably MATLAB, which has a couple of functions specifically designed for fitting spectra with Voigt profiles. You can also use the curve fitting toolbox in the software to get more control over the fit and adjust the parameters more precisely. It is probably a better idea to fit one peak at a time as it simplifies the problem and usually gives better results. However, if you are looking for a more general solution that can fit multiple peaks simultaneously, then Python's scipy package does have a function called curve_fit() that can be used to fit data to arbitrary functions with multiple parameters.