Multiple lines in one graph

Dear All,

I am drawing several lines on one graph and I would like to make them thicker where they coincide. All advice gratefully accepted.

John Hey

[Rik]

Could you give an example of realistic data? And what did you try so far?

[John Hey]

Dear Rik, Thanks for replying. Here is a graph I have produced. At the end all the lines coincide. I would like to make the thickness of the lines indicate how many lines coincide. Ant help would be greatly appreciated. John

[Stephen23]

@John Hey: do all of the plotted lines use exactly the same X values?

[John Hey]

Stephen, Yes - they do. Does that help? John

Answer 1

[Stephen23]

Here is one solution using NaN's to mask out parts of the lines, and plotting them for each thickness. For this to work it is also required to interpolate the data (try it without to see the difference). I don't claim that this is very efficient, but for a small number of lines and nodes it will work.

Xi = [1;2;3;4;5;6];
Yi = [1,2,2,2,1,1;2,2,2,2,3,4;3,3,3,3,3,4;4,3,3,3,4,5;5,5,2,2,4,4].';
plot(Xi,Yi,'LineWidth',3)
hold on
ylim([0,6])
N = numel(Xi);
Xo = interp1(1:N,Xi,1:0.5:N);
Yo = interp1(1:N,Yi,1:0.5:N);
D = bsxfun(@minus,Yo,permute(Yo,[1,3,2]))==0;
D = D(1:end-1,:,:) & D(2:end,:,:);
D = convn(D,[1;1])>0;
S = sum(D,3);
for k = 2:max(S(:))
    Yo(S<k) = NaN;
    plot(Xo,Yo,'k','LineWidth',k*3)
end

Giving:

You can adjust the colors and line thickness as you see fit. With some effort you could simplify and make it more efficient (e.g. remove columns with all NaN, or use indexing to extract only the non-NaN portions of M and plot only them).

Answer 2

[elham kreem]

let

y = 1:2:40

x1=4:2:43

x2 = 8:3:66

x3 = 1:4:80

figure ;

plot( x1 , y , ' r ' , x2 , y , ' b--o' , x3,y , 'g-*')