Inconsistency in the fitted curve and the plot of obtained fitting parameters

Question

Shaily_T 2022-5-30

0
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/1729655-inconsistency-in-the-fitted-curve-and-the-plot-of-obtained-fitting-parameters

评论： Mathieu NOE 2022-5-31

kkrebook2.m

I am trying to fit an equation to a model and I need to call a function "kkrebook2" in my fit type funtion (For "kkrebook2", the inputs are two vectors and its output is a vector). I have attached the way I call "kkrebook2" in my fit type function. When I run my fit code I see the obtained fitted curve (red figure) is different from the plot of the fit type function by using the obtained fitting parameters (yellow figure) which doesn't make sense. This is what I see:

This is the snippet of my fit type function code where I call "kkrebook2" function.

function p = SCR(nu,numGaussians,a,center,sigma)

for i = 1:length(nu)

for k = 1 : numGaussians

if k==1

b = Start;

else

b = Start + (center * (k-1));

end

thisGaussian(i) = a.*exp(-((nu(i)-b).^2)/(2.*(sigma.^2)));

% Add into accumulator array:

gaussEqn1(i) = gaussEqn1(i) + thisGaussian(i);

z(i)= gaussEqn1(i);

end

Rn = kkrebook2(nu,z,0);

for i = 1:length(nu)

hi2(i) = (4*pi*n.*nu(i)*L)/c;

hi1(i)=(Rn(i));

f(i)=(-r1+r2*exp(-(z(i))).*exp(-1i*hi1(i)).*exp(-1i*hi2(i)));

g(i)=(1-r1*r2*exp(-(z(i))).*exp(-1i*hi1(i)).*exp(-1i*hi2(i)));

h(i)= f(i)./g(i);

m(i)= abs(h(i));

p(i)= (m(i).^2);

end

I have also attached "kkrebook2" function that I call in my code.

Could you please let me know where is the problem? I think it should be in the way I am calling "kkrebook2" but I cannot capture it.

Thanks in advance!

4 个评论
显示 2更早的评论隐藏 2更早的评论

Mathieu NOE 2022-5-30

在 MATLAB Online 中打开

hello

i am not really answering the question, but the poor's man solution (without the curve fiiting toolbox) based on fminsearch does not show any "bizarre" output

clear
load('42nsC.mat')
nu = freq;
p = REF;
Startpoint = 3.778651249339824e+14;
Endpoint = 3.778653249339824e+14;
% Initial conditions
% numGaussians = 9;
% center = (Endpoint - Startpoint)/numGaussians;
% sigma = center/10;
% d0 = 0.05;
% a = 0.23;
numGaussians = 9;
center = 24.1E6;
sigma = 3.4E6;
d0 = 0.04;
a = 0.8;
% curve fit using fminsearch
f = @(a,center,sigma,d0,nu) SCR(nu,numGaussians,a,center,sigma,d0);
obj_fun = @(params) norm(f(params(1), params(2), params(3), params(4),nu)-p);
options = optimset('MaxIter',100);
sol = fminsearch(obj_fun, [a,center,sigma,d0],options);
a_sol = sol(1);
center_sol = sol(2);
sigma_sol = sol(3);
d0_sol = sol(4);
y_fit = f(a_sol, center_sol,sigma_sol, d0_sol, nu); % same output as line below
Sol1= SCR(nu,numGaussians,a_sol,center_sol,sigma_sol,d0_sol); % p = SCR(nu,numGaussians,a,center,sigma,d0)
figure(1)
clf
hold on
plot(nu, p)
plot(nu,Sol1) % This plots the solution above. 
plot(nu,y_fit,'k.') % This plots the solution as well (fit). 
hold off
set(gca,'XTick',Startpoint:1E8:Endpoint)
set(gca,'XTickLabel',-2.87:0.1:-2.67)
xlim([Startpoint Endpoint+15E6])
xlabel('Frequency Offset (GHz)','FontSize', 20)

Shaily_T 2022-5-30

@Mathieu NOE Thank you so much for your response! I appreciate that.

Mathieu NOE 2022-5-31

My pleasure !

请先登录，再进行评论。

请先登录，再回答此问题。

Inconsistency in the fitted curve and the plot of obtained fitting parameters

4 个评论
显示 2更早的评论隐藏 2更早的评论

回答（0 个）

另请参阅

类别

标签

Community Treasure Hunt

Inconsistency in the fitted curve and the plot of obtained fitting parameters

4 个评论 显示 2更早的评论隐藏 2更早的评论

回答（0 个）

另请参阅

类别

标签

Community Treasure Hunt

4 个评论
显示 2更早的评论隐藏 2更早的评论