Shade between time series lines (log-y scale)

15 views (last 30 days)
Hello,
I'm attempting to plot between two lines on a graph that's set to a log-y scale. I've tried almost every iteration of the "fill" command that seems to appear from various sources online, which usually gives me somewhat psychedelic (but useless) results.
Suppose that:
x = 0:100
y1 = rand(1,100)+1.5
y2 = rand(1,100)+0.5
How can I fill just between the y1 and y2 lines??? I want nothing but empty space below the y2 line. Also, this is on a log-y scale, which is why I can't just tweak the "area" commands (like I did for a similar project that was not on a log-y scale.
Please help!!!
Thanks, Ryan

Accepted Answer

Walter Roberson
Walter Roberson on 19 Jan 2012
Your x starts from 0 (and is one point longer than y1 or y2.) If you set the axis XScale to log, then you are attempting to take log(0), which leads to problems in drawing the patch.
Without the 0 I had no problem.
x = 1:100;
y1 = rand(1,100)+1.5;
y2 = rand(1,100)+0.5;
fill([x fliplr(x)],[y1 fliplr(y2)],'r')
pause(3)
set(gca,'XScale','log','YScale','log')
drawnow()
  4 Comments
Walter Roberson
Walter Roberson on 4 Nov 2018
You did not use my revised code that has the x.' and so on.
To give it again in full:
A = xlsread('TAW_NZ');
x = A(1:100,1);
min1 = A(1:100,6);
max1 = A(1:100,7);
figure
plot(x, min1, x, max1);
fill([x.',fliplr(x.')], [min1.',fliplr(max1.')], 'r')

Sign in to comment.

More Answers (3)

Venn Ravichandran
Venn Ravichandran on 19 Jan 2012
Using stem plot might be a quick work around. I don't know why the area plots don't work.

the cyclist
the cyclist on 19 Jan 2012
Will this suffice?
x = 1:100;
y1 = rand(1,100)+1.5;
y2 = rand(1,100)+0.5;
figure
plot(x,y1,x,y2)
fill([x,fliplr(x)],[y1,fliplr(y2)],'r')
set(gca,'YScale','log')
[Edited to include the additional fliplr, in response to Walter's comment.]
  1 Comment
Walter Roberson
Walter Roberson on 19 Jan 2012
I think you have to flip y2 as well.
Amusing that we choose the same color and same basic technique :)

Sign in to comment.


Ryan Utz
Ryan Utz on 19 Jan 2012
Thanks Walter and the cyclist! This does work on my more complex data. The command "fliplr" is new to me, and useful...

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!