How can I fit this data by using fit function?
1 Ansicht (letzte 30 Tage)
Ältere Kommentare anzeigen
I need to fit the data, I tried to fit it by using gauss fit functions. Can anybody help?
2 Kommentare
dpb
am 25 Feb. 2020
Not w/o data, no...just looking at a picture doesn't let anybody do much of anything.
Need to subtract a baseline and then mayhaps a muti-peak model might have some chance. Looks like interfering peaks may be present, however, which makes things much tougher...
Am familiar with the experiment but never actually did a lab with it to have collected actual data to try to fit...
Walter Roberson
am 2 Mär. 2020
My experience suggests that fitting that data well would require 16 to 24 guassians.
Fitting guassians well is difficult to do.Very small changes in proposed position can make a big difference in goodness of fit.
I have long forgotten the details, but when you fit a guassian you need four parameters (ummm, position, half-width, phase, and amplitude it might have been). It turns out that given the right two of those, the other two can be calculated linearly, but I have long forgotten how. The result is that you have to do a nonlinear search for two parameters for each guassian involved, and do a linear calculation at each step, and then calculate the goodness of fit, and then move slightly and try again. Gradient minimization does not work well for the nonlinear search, as it has many many small local minima. A modified simplex search worked better at the time.
Antworten (1)
PRABHANSHU PURWAR
am 2 Mär. 2020
Try these:
https://in.mathworks.com/help/curvefit/curve-fitting.html (Curve Fitting)
And few demo files: http://www.mathworks.com/matlabcentral/fileexchange/38006-demo-functions-for-peak-detection-and-fitting (Demo functions for peak detection and fitting)
Hope it helps!!
0 Kommentare
Siehe auch
Kategorien
Mehr zu Linear and Nonlinear Regression finden Sie in Help Center und File Exchange
Community Treasure Hunt
Find the treasures in MATLAB Central and discover how the community can help you!
Start Hunting!