LabVIEW

cancel
Showing results for 
Search instead for 
Did you mean: 

multiple peak fitting

Solved!
Go to solution

Dear all, i'm working on a small project for the development of an time domain interferometer and in particular on the data analysis.

I' m quite new on labview programming and i have a problem on the fit of multiple peaks.

i used the peak detector but it is not the right way to implement the analysis.

I would like to fit every peak presents on the data and found the position. Every evident peak is an ovelap of multiple peak. my aim is to find the mamimum of this group of peak. Some suggestions?

Attached an example of the data obtained.

 

Regards and thank's in advance

 

 

 

 

0 Kudos
Message 1 of 3
(3,423 Views)
Solution
Accepted by vittoc13

Do you know the mathematical shape of the peaks (Gaussian, lorentzian, etc.). In this case you can uses nonlinear fitting.

You could use the output of the peak detector to generate the initial estimate.

Can you make any other simplifying assumptions, e.g. do all peaks have the same width (just differening in position and amplitude).

 

 

0 Kudos
Message 2 of 3
(3,392 Views)
Solution
Accepted by vittoc13

Think very carefully about what criteria you will use to determine when to stop adding peaks.  In particular, if you have N points, you can get a perfect fit by using very narrow fitting curves (smaller than the spacing between points) and fitting each point to a peak at that height.

 

Fitting peaks is not a trivial topic.  Not only do you need to determine what you mean by a "peak", but what criteria you will use for "goodness of fit", (as Altenbach noted) whether the peaks have a common attribute (shape, size, width, etc.), and (as I just noted) when to stop fitting.

 

Bob Schor

0 Kudos
Message 3 of 3
(3,376 Views)