Legend Color doesn't Match the Curves Color

Question

A.J.M 2020-9-15

0
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/593911-legend-color-doesn-t-match-the-curves-color

评论： A.J.M 2020-9-15

PDF vs Distance.jpg

Good Day;

Here my question is happened often to occur as I see when I searched to solve my problem,but each case it depend on their code. I faced this problem as shown in the attached figure, and I hope to find the solution from you. I used 'hold on'but nothing changed, the code as shown below:

Z=100:250:5000; %Distance Between The Transmitter and Reciever in m

pointing_Err_angle=1*10^-5; %Pointing Error Angle in mrad

lambda=1550e-9; %Wavelength in nm

wo=5*10^-2; %Beam Waist at Z=0 in m

Cn=sqrt(5e-15); %Refractive Index Structure Parameter

% --------------------Calculation Section----------------------------------

k=2*pi/lambda; %The number of optical wave

po=(0.55*(Cn)^2*k^2*Z).^(-3/5); %The Coherence Length

E=(1+2*wo^2./po.^2);

wz=wo*(1+E.*((lambda.*Z)./(pi*wo^2)).^2).^(1/2); %The Beam Waist

r=Z*(pointing_Err_angle); %The Radial Displacement B

%----------------------------------------Calculate Hp-----------------------------

v=(sqrt(pi)*a)./(sqrt(2)*wz);

a0=(erf(v)).^2; %Maximal Fraction of Collected Power at r=0

Wzeq=sqrt(wz.^2*((sqrt(pi).*erf(v))/(2*v.*exp(-v.^2))));

Hp=a0.*exp((-2*r.^2)./((Wzeq).^2));

%----------------------------------Intensity Distribution-------------------------

Rytov_var=[0.1,0.2,0.5,0.8]; %Log irradiance variance (Roytov variance)

for i=1:length(Rytov_var)

for j=1:5:length(Hp)

A1(i,j)=1./(Hp(j).*sqrt(2*pi*Rytov_var(i)));

A2(i,j)=((log(Hp(j))+0.5.*Rytov_var(i)).^2)./2.*Rytov_var(i).^2;

PDF_Hp(i,j)=A1(i,j).*exp(-(A2(i,j)));

end

%--------------------------Plot Section------------------------------------

z=1250:1250:5000; %Distance for Plot Purpose

figure(1)

semilogy(z,PDF_Hp,'LineWidth',4);

xlabel('The Distance Z')

ylabel('The Probability Density Function ')

title('The PDF of Hp Vs The Distance')

legend('σ^2_I=0.1','σ^2_I=0.2','σ^2_I=0.5','σ^2_I=0.8')

grid on

figure(2)

semilogy(Rytov_var,PDF_Hp,'LineWidth',3);

xlabel('Roytov variance')

ylabel('The Probability Density Function ')

title('The PDF of Hp Vs Rytove Variance')

legend('σ^2_I=0.1','σ^2_I=0.2','σ^2_I=0.5','σ^2_I=0.8')

grid on

0 个评论
显示 -2更早的评论隐藏 -2更早的评论

请先登录，再进行评论。

请先登录，再回答此问题。

Answer 1

Adam Danz 2020-9-15

0
链接

此回答的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/593911-legend-color-doesn-t-match-the-curves-color#answer_494794

编辑：Adam Danz 2020-9-15

在 MATLAB Online 中打开

I see the problem. Look at the values you're plotting. Assuming a=0.5e-2;

semilogy(z,PDF_Hp,'LineWidth',4);

PDF_Hp =
  Columns 1 through 12
       59.204            0            0            0            0       74.572            0            0            0            0       130.59            0
       33.684            0            0            0            0       41.243            0            0            0            0        66.92            0
       5.2821            0            0            0            0       5.3479            0            0            0            0       5.1881            0
      0.42448            0            0            0            0      0.30818            0            0            0            0      0.12088            0
  Columns 13 through 16
            0            0            0       242.89
            0            0            0       112.93
            0            0            0       4.5326
            0            0            0     0.032869

A line is created for each column so your plot actually has 16 lines, not 4.

The reason they don't appear is because in a log plot, 0 is not defined. However, those line objects are still produced - they are just not shown. As evidence of that, check out the 16 object handles.

h = semilogy(z,PDF_Hp,'LineWidth',4)
%     h = 
%       16×1 Line array:
%     
%       Line
%       Line
%       Line
%       Line
%       Line
%       Line
%       Line
%       Line
%       Line
%       Line
%       Line
%       Line
%       Line
%       Line
%       Line
%       Line

The reason the "wrong" colors are defined is because you're only asking for the first 4 objects to appear in the legend. But objects 2-3-4 are all non-visible lines.

legend('σ^2_I=0.1','σ^2_I=0.2','σ^2_I=0.5','σ^2_I=0.8')

Solution

Get rid of the columns with 0s in PDF_Hp

1 个评论
显示 -1更早的评论隐藏 -1更早的评论

A.J.M 2020-9-15

Thanks a lot, Mr. Adam Danz, the problem was solved. I appreciate your hard efforts

Accept my respect.

请先登录，再进行评论。

Answer 2

Bjorn Gustavsson 2020-9-15

0
链接

此回答的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/593911-legend-color-doesn-t-match-the-curves-color#answer_494719

在 MATLAB Online 中打开

That's peculiar. What happens if you keep the plot-handle from your calls to semilogy and send that to legend?

ph = semilogy(z,PDF_Hp,'LineWidth',4);
xlabel('The Distance Z')
ylabel('The Probability Density Function ')
title('The PDF of Hp Vs The Distance')
legend(ph,'σ^2_I=0.1','σ^2_I=0.2','σ^2_I=0.5','σ^2_I=0.8')