Fitting routine: how can increase order of Gaussian until certain convergence criterion?

Dear Users!

I am currently analyzing a big amount of data. The data has (ideally) the shape of Gaussians (1,2 or 3). Now I want to make a simple routine that tries to fit first 1 Gaussian, then a combination or 2 Gaussians and then 3, until a certain GOF is reached or the last one is accepted.

I am not sure how to do this most efficiently as there are so many ways to fit. I usually use something like this:

[f1 gof1]=fit(xData,yData,'gauss1');

but interestingly the cftool gives me much better (say with the naked eye in better agreement), is however a little bit more cumbersome to call (or maybe I am not doing it correctly) since I have to pass a function to it.

What is your recommended procedure for this kind of routine work?

Yours, Chris