Fitting the curve problem

Yumeng Yin

2022 3 10

0 个回答

4 次查看（30 天）

0 个投票

Ch4.mat

I try to fit an s-shape curve with the function that I define. Actually I get a not-so-bad fitting by setting a more appropriate boundary and starting point. One remaining question is that although the fit is better, the parameters' range is still big. Anyone has any advice for this? Attached file is the data file for Ch4.

Final_time=10;
recordLength=size(Ch4);
recordLength=recordLength(1);
Time=[linspace(0,Final_time*1000,recordLength)]';
Ch4_mV=Ch4.*1000;
[pks,locs]=findpeaks(Ch4_mV,'MinPeakProminence',30);
for j=3
   duration_fit=Time(locs(j)-6:locs(j)+10);
   y_duration_fit=Ch4_mV(locs(j)-6:locs(j)+10);
   FitFunction=@(a,b,A,B,F,e,x)F.*(b.*A-(x-a).*B)./((x-a).^2+b^2)+e;
   options=fitoptions('Method','NonlinearLeastSquares','Upper',[Time(locs(j)+50) Time(8)-Time(1) 1 1 pks(j)+10 0],'Lower',[Time(locs(j)-50) 0 -1 -1 pks(j)-10 -15],'StartPoint',[Time(locs(j)),1,1,Time(locs(j)+5)-Time(locs(j)-2),pks(j),1]); 
   [fitcurve,gof]=fit(duration_fit,y_duration_fit,FitFunction,options)
   plot(fitcurve,duration_fit,y_duration_fit);
end