Legend problem in a multiple axes graph

I want to avoid the tick misalignment of plotyy (and the ticks on the top x axis), so I create a multiple axes plot manually...

Ex :

x = [1 2 3 4 5 6 7 ];
y1 = x;
y2 = [10 20 30 70 80 90 100]

line(x, y1)

ax1 = gca;

ax2 = axes('Position',get(ax1,'Position'),...
           'XAxisLocation','bottom',...
           'YAxisLocation','right',...
           'Color','none',...
           'XTick',[],...
           'XColor','k','YColor','k');

line(x,y2,'color','black','linestyle','--')

legend('y1','y2')

But then I can't display both lines in my legend box and get the following message: "Warning: Ignoring extra legend entries."

Any suggestion on how to solve this problem? Thanks!

Answer 1

[Walter Roberson]

x = [1 2 3 4 5 6 7 ];
y1 = x;
y2 = [10 20 30 70 80 90 100]

L1 = line(x, y1);

ax1 = gca;

ax2 = axes('Position',get(ax1,'Position'),...
           'XAxisLocation','bottom',...
           'YAxisLocation','right',...
           'Color','none',...
           'XTick',[],...
           'XColor','k','YColor','k');

L2 = line(x,y2,'color','black','linestyle','--');

legend([L1, L2],'y1','y2')

[Étienne Gaudette]

Thanks!

[Alborz Sakhaei]

I have a figure with two axes (left and right). 2 lines on left axis and one line on right axis. I can not get all legends in one box using following code.

PS. I have read previous discussions (links below) on this topic but does not seem to work for me!

Single legend for plot with two axes

Legend problem in a multiple axes graph

function
x = 0:0.01:10;
y11 = sin(x);
y12 = cos(x);
y2  = ones(1001,1);
figure;
ax(1) = axes('position',[0.1 0.1 0.8 0.8]);
ax(2) = axes('position',[0.1 0.1 0.8 0.8], 'yaxislocation','right','color','none');
line('parent',ax(1),'xdata',x,'ydata',y11,'color','b');
line('parent',ax(1),'xdata',x,'ydata',y12,'color','g');
line('parent',ax(2),'xdata',x,'ydata',y2,'color','r');
legend( [ax(1) ; ax(2)] , {'sin','cos','one'} )

[Walter Roberson]

Do not legend the axes handles, legend the line() handles.

[Alborz Sakhaei]

Thanks Walter! Works perfect.