Legend for multiple plots

I want to generate a legend for multiple plots on the same figure using 'hold on;'. I am using a certain method I found through googling but I am not getting expected results. Here is my code.

 sf = 1;         % scaling factor
lambda_b = 10;  % BS process density
lambda_c = 10;  % cluster process density c <= b
lambda_u = 50;  % user process density; large enough to make each BS active

 % Base station process
npointsb = poissrnd(lambda_b*sf);    % poisson variable; no. of points
pprocb = rand(npointsb,2);         % uniform dist of points
xb = pprocb(:,1);
yb = pprocb(:,2);

 % Cluster process
npointsc = poissrnd(lambda_c*sf);
pprocc = rand(npointsc,2);
xc = pprocc(:,1);
yc = pprocc(:,2);

 % User process
npointsu = poissrnd(lambda_u*sf);
pprocu = rand(npointsu,2);
xu = pprocu(:,1);
yu = pprocu(:,2);

 figure();
[vxb,vyb] = voronoi(xb,yb);
[vxc,vyc] = voronoi(xc,yc);
p1 = plot(vxb,vyb,'b','userdata','BS tessellation'); hold on;
p2 = plot(xb,yb,'rs','userdata','BS');
p3 = plot(vxc,vyc,'r','userdata','Cluster tessellation'); 
p4 = plot(xu,yu,'go','userdata','User'); hold off;
legend(get(gca, 'children'), get(get(gca, 'children'), 'userdata'));
axis([0 1 0 1]);

The result of this code is illustrated in the image.

You can see there are multiple entries and I suspect that is because the variables returned by the voronoi function are matrices and not vectors like the other variables. It seems this is making the plot function return a column vector of Line objects and this is why I am getting the multiple legend entries.

How do I get around this and plot the correct legend ignoring the multiple entries?

[Vijayanand Sivakumar]

Thank you very much for the solution.

Answer 1

[Jonathan Mayers]

Hi all,

I have obtained the desired legend output using handles. See the code below. The code generating the points did not change.

 figure();
[vxb,vyb] = voronoi(xb,yb);
[vxc,vyc] = voronoi(xc,yc);
p1 = plot(vxb,vyb,'b'); hold on;
p2 = plot(xb,yb,'rs');
p3 = plot(vxc,vyc,'r'); 
p4 = plot(xu,yu,'go'); hold off;
axis([0 1 0 1]);

 % Create column vector of plot handles
% Variables returned from voronoi are matrices and plotting them
%   returns a column vector of handles so extract the first one
h = [p1(1);p2;p3(1);p4];

 % Now call the legend function passing the handle h and specify the text
legend(h,'BS tessellation','BS','Cluster tessellation','User');

The below image verifies correct output of the code.

[Héctor Carrasco López]

You are my hero.

[Shivam Chopra]

Hi, what should I do if I need to insert legends for these plots plotted in loop??

a = 1;
p = [1];
figure();

for B = [1E-9 1E-8 1E-7 1E-6 1E-5 1E-4 1E-3]
abs_z = diffusion(B);
f = (0:10:10^5);
p(a) = plot(log10(f),log10(abs_z));
a = a+1;
hold on;
end
hold off;

function abs_z = diffusion(B)
t_n = 1E-6;
d_n = 1E+6;
q = 1.6E-19;
m_n = 9.11E-31;
f = (0:10:10^5);
w_c = (q * B)/(m_n);
w = f.*(2*pi);

d_real = 1 + ((t_n^2) * ((w.^2) + (w_c^2))) ; 
d_img = w.*((t_n)*((t_n.^2) * (((w_c.^2) - w.^2)-1))) ; 
d_divider = (1 + ((t_n.^2) * (((w_c.^2) - w.^2).^2)) + ((w.*2*(t_n)).^2)) ;

z = complex(d_real,d_img);
d_m = d_n./d_divider;
abs_z = abs(z).*d_m;
end

Thanks in advance

[Walter Roberson]

p(a) = plot(log10(f), log10(abs_z), 'DisplayName', sprintf('B = %.0e', B));

and at the end of the loop.

legend show

[Tingmingke Lu]

This comes in handy! Thanks!

[anup mallick]

Thanks...After a long search here and there, I got your answer, it worked for me..

Answer 2

[Taniel Winner]

I have found this code INCREDIBLY HELPFUL!!! It works!! Download it!

https://www.mathworks.com/matlabcentral/fileexchange/67646-legendunq

Answer 3

[Miguel Marques]

You are my hero too.