LabVIEW
lev-mar multivariate data
You need to rewrite the nonlinear Lev-Mar fit so the fit funtion operates on the entire array, instead of one point at the time and everything will be much easier.
Then simply reshape your 2D gaussian data and model function to a 1D array of size (x*y) and fit as usual.
To get you started we've done some of the work for you: 🙂
Quite a while ago, David Thomson, Don Wagner and me had some discussions about improving Lev-Mar.
David has distilled a version and posted it on his site at the following link:
http://www.originalcode.com/programs_scientific.html
Look for the section "Nonlinear Fit", Version 2 (version 1 would not help much).
I have worked out a much more elaborate versions but the above should get you started.
Then simply reshape your 2D gaussian data and model function to a 1D array of size (x*y) and fit as usual.
To get you started we've done some of the work for you: 🙂
Quite a while ago, David Thomson, Don Wagner and me had some discussions about improving Lev-Mar.
David has distilled a version and posted it on his site at the following link:
http://www.originalcode.com/programs_scientific.html
Look for the section "Nonlinear Fit", Version 2 (version 1 would not help much).
I have worked out a much more elaborate versions but the above should get you started.
Eugene,
Remember that the x and y values are not really needed, because they are given by the array itself.
The fit function must just calculate the 2D gaussian as another such 2D array which is then compared to the experimental data.
In your particular case, the 5 fittable variables would be e.g. [amplitude, x-center, x-width, y-center, y-width]. Your function does not need an x input at all, it is sufficient if it produces a 2D array from these 5 parameters, to be compared to the experimental array. The axes are implicit.
I am very busy today, but please contact me directly at caltenba(at)ucla.edu. Maybe send me your data and function and I'll have a look at it over the weekend. This seems not too complicated.
Remember that the x and y values are not really needed, because they are given by the array itself.
The fit function must just calculate the 2D gaussian as another such 2D array which is then compared to the experimental data.
In your particular case, the 5 fittable variables would be e.g. [amplitude, x-center, x-width, y-center, y-width]. Your function does not need an x input at all, it is sufficient if it produces a 2D array from these 5 parameters, to be compared to the experimental array. The axes are implicit.
I am very busy today, but please contact me directly at caltenba(at)ucla.edu. Maybe send me your data and function and I'll have a look at it over the weekend. This seems not too complicated.
