how to solve the error in Lsqcurvefit

hi guys

i need help for the error i got it , with the function lsqcurvefit, i don't know exactely how to solve this probleme.

Thank you,

this is my script:

clear all
clc
load('RCS1190_460A.mat')

xdata=Times; 

i=460
%%    RCS1190_460A             %%%%%%%%%%%%%%%%%ù
ydata= QC_plusC-45;
fun=@(x,xdata)((x(1)./x(2)).*(  1-   (exp(-(x(2)).*xdata))))
x0=[2,1]
x=lsqcurvefit(fun,x0,xdata,ydata)

[Torsten]

Execute

load('RCS1190_460A.mat')
xdata=Times; 
i=460;
%%    RCS1190_460A             %%%%%%%%%%%%%%%%%ù
ydata= QC_plusC-45;
fun=@(x,xdata)((x(1)./x(2)).*(  1-   (exp(-(x(2)).*xdata))));
x0=[2,1];
test_y = fun(x0,xdata)

and check whether "test_y" has Inf or NaN elements.

[Sou Fat]

I ran your code and found that test_y has a value,

[Sou Fat]

I have the same error you saw it in the picture when i run the code with this "x=lsqcurvefit(fun,x0,xdata,ydata)"

[Torsten]

test_y must return a list of values of the same size as your "ydata" and there mustn't be any NaN or Inf values in it.

Any of these conditions cannot be true in your case because lsqcurvefit gets garbage back when it calls "fun" with your xdata and your initial parameter vector x0.

[Sou Fat]

exactly what I found a list of value of the same size as my ydata but when i tried to run the function lsqcurvefit, so i don't how to solve this probleme . thank you for your answer

[Torsten]

But it says your objective function returns undefined values at the initial point.

So "test_y" must contain Inf or NaN values.

[Sou Fat]

i don't know, look at these picture that when i run the code without lesqcurvefit, it's work

[Sou Fat]

you think i have problem in the ydata ?

[Torsten]

Try what you get from

any(isnan(fun(x0,xdata)))
any(isinf(fun(x0,xdata)))

If one of them returns 1, you have a problem with the xdata and/or the evaluation of your function.

Do the same for your ydata.

[Sou Fat]

i found any(isnan(fun(x0,xdata))) return 1.

and second any(isinf(fun(x0,xdata))) return 0

[Sou Fat]

i did the same for ydata and i found the both return 0

[Sou Fat]

I rectified in the tebalu that it misses a value I rectified it then I verified the two any(isnan(fun(x0,xdata)))

any(isinf(fun(x0,xdata))) and fo ydata which return 0

[Torsten]

So no more errors ?

Answer 1

[Sou Fat]

Yes I didn't find the error thank you for your help