minimize the error between calculated values and experimental results

Hello, I am writing the following code to minimize the error between calculated values and experimental results:

K=[1 2 2.8 3.3 3.5 3.6]; %some data to test

v = sym('v',[1 2]);
a = v(1); b = v(2);

e=[1 2 3 4 5 6]
eo=2
for i=1:1:6
    D(i)=@(v)a*(e(i)-eo)^b
    E(i)=e(i)*(1-D(i));
    R(i)=(K(i)-E(i))^2;
end

RM=0
for i=1:1:6
    RM=RM+R(i)
end

RMS=(RM)^(1/2)
v = [0.12 2];
c=fminsearch(RMS,v)

Its showing error. Can you please help to solve the issue.

Answer 1

[Guru Mohanty]

Hi, I understand you are trying to minimize error for given set of data. The error in your code is due the fact that the function fminsearch takes function handle as input, instead of sym. First you need to convert the ‘RMS’ to function handle using matlabFunction and then compute the minimum. Here is the modified code for it.

clear all;
clc;
K=[1 2 2.8 3.3 3.5 3.6]; %some data to test
v = sym('v',[1 2]);
a = v(1); b = v(2);
e=[1 2 3 4 5 6];
eo=2;
RM=0;
for i=1:1:6
    D(i)=a*(e(i)-eo)^b;
    E(i)=e(i)*(1-D(i));
    R(i)=(K(i)-E(i))^2;
    RM=RM+R(i);
end
RMS=(RM)^(1/2);
RMS=matlabFunction(RMS,'Vars',{[v(1),v(2)]});
v = [0.12 2];
c = fminsearch(RMS, v);
disp(c);

[MEENAKSHI SHARMA]

@Guru Mohanty - Thanks a lot for your answer.