Data fitting with correlation coefficient

4 次查看(过去 30 天)
I have 1 dimensional array of data (nonlinear regression) that I need to figure out the best curve fitting (normal, weibull, Exponential, ....) for, according to the higher correlation coefficient (R). what is the Matlab function to use "without third party code, like mike's allfitdist or fitmethis codes"? thanks in advance

回答(1 个)

Amr Hashem
Amr Hashem 2015-10-29
If you want to find " The best curve fitting", you can use
cftool
which is a graphical user interface.
or you can use the functions of each, for example:
expfcn = @(b,x) b(1).*exp(b(2).*x); %exponential with 2 parameters
B0=[50;0] % intial parameters
B=nlinfit(values,ydata,expfcn,B0);
Exp=expfcn(B0,xvalues);
plot(xvalues,Exp)
but if you want to find " The best distribution fitting", you can use
exppdf % exponential
fitdist(data,'normal'); % normal
wblpdf % weibull

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!

Translated by