This routine will shade the area of a 2-D plot between two user defined vectors. Simply pass the two vectors, and their corresponding horizontal coordinates, select your fill color and transparency level and viola! You have a figure that is colored in between the two input vectors.
Thanks for the code! I adapted it to add regions of interest so you can shade just particular parts of the two curves (such as statistically different sections), using a logical vector. I also fixed the orientation thing.
%% ADAPTED BY K. OLFERS 28-08-2016
%This function will fill a region with a color between the two vectors provided
%using the Matlab fill command.
%fillhandle is the returned handle to the filled region in the plot.
%xpoints= The horizontal data points (ie frequencies). Note length(Upper)
% must equal Length(lower)and must equal length(xpoints)!
%upper = the upper curve values (data can be less than lower)
%lower = the lower curve values (data can be more than upper)
%roi = a *(logical) vector specifying the areas to be shaded
%color = the color of the filled area
%edge = the color around the edge of the filled area
%add = a flag to add to the current plot or make a new one.
%transparency is a value ranging from 1 for opaque to 0 for invisible for
%the filled color only.
%John A. Bockstege November 2006;
% a=rand(1,20);%Vector of random data
% b=a+2*rand(1,20);%2nd vector of data points;
% x=1:20;%horizontal vector
% grid on
if nargin<8;transparency=.5;end %default is to have a transparency of .5
if nargin<7;add=0;end %default is to add to current plot
if nargin<6;edge='k';end %dfault edge color is black
if nargin<5;color='b';end %default color is blue
if nargin<4; roi = ones(length(xpoints),1); end
if ~isrow(lower); lower = lower'; end
if ~isrow(upper); upper = upper'; end
if ~isrow(xpoints); xpoints = xpoints'; end
[clusters cnum] = bwlabeln(roi);
add = 1;
for iClust = 1:cnum
iStart = find(clusters==iClust,1);
clength = length(find(clusters == iClust))-1;
tmpUpper = upper(iStart:iStart+clength);
tmpLower = lower(iStart:iStart+clength);
tmpXpoints = xpoints (iStart:iStart+clength);
if length(tmpUpper)==length(tmpLower) && length(tmpLower)==length(tmpXpoints)
fillhandle=fill(tmpXpoints,filled,color);%plot the data
set(fillhandle,'EdgeColor',edge,'FaceAlpha',transparency,'EdgeAlpha',transparency);%set edge color
msg='Error: Must use the same number of points in each vector';
Excellent tool! Thank you for sharing it.
That's correct. As Paul pointed out, the input vectors need to be rows. Thanks for the function and thanks Paul for the comment.
would be nice if it didn't freak out with column vectors. works well if i transpose my data vectors, but it took me a bit to figure that out. thanks!
I applied spline interpolation to fit the curve. In this case, how can I use 'jbfill' function to shade area between two curves.
x=[0 100 200 300 400 500 600 700 800 900 1000 1100 1200 1300];
y=[0 28.23 27.3 20.94 13.89 6.268 2.468 0.1701 0.3704 0.1852 0.1852 0 0 0];
errBar=[0 3.366 4.936 4.969 2.294 2.353 1.568 0.1701 0.3704 0.1852 0.1852 0 0 0];
topcurve = y+errBar;
bottomcurve = y-errBar;
Do you have a way to figure this out?
This only works with data input as rows. To allow for column or row vector input, add the following:
xpoints = xpoints';
upper = upper';
lower = lower';
looks excellent, but somehow it doesnt save as pdf properly...
I was using this program but unable to save my figure as jpeg/tif/bmp or copy this figure. This just give a black picture. I do not understand this have anyone came across this before?
thanks, very very useful
Very useful function. That really helped!
Thanks a lot!
I use this function for plotting confidence intervals... it's very useful, thanks!
Works great, thanks!
Easy to use, and very useful... at least for those of us working in geophysics. Thanks
nice function, nice help, straightforward use
Simple and straight forward! Just what I needed.
Very nice. But because of the transparancy, Matlab switches to OpenGL, which leads to vanishing axis. To avoid that I just make the axis lines thicker set(gca,'LineWidth',1.0).
Because logarithmic plots do not support OpenGL, you can not draw transparant areas in logarithmic plots.
Very smart solution.
This function is great, thanks very much. I was hoping to use it to highlight the time-periods for which there was a statistical difference between the two lines. After creating the basic plot, I created a new matrix which specifies the time-points of interest (statistcially significant) in the first column, and the relevant data points for the two lines in the 2nd and 3rd column. I then ran jbfill on this. Unfortunately because the periods of interest are non-contiguous the function fills in the area in between. Is there a way to make it so that it wil only shade in areas at the specific time-points I have specified?
very good, and commented well (does what it says on the tin)
good but row vectors VS column vectors as input does matter.
the "transparancy" option comes with a caveat: it will switch the renderer to opengl. After that, a vector export is not possible anymore (matlab limitation). set to 1.0 to avoid this problem.
thank you!!!! this came in REALLY handy! easy to use, well documented! :)
Very, very helpful and extremely easy to use. Great for confidence intervals!
Great. This was very helpful and saved me lots of time to figure out the exact work that you've done. Thanks!
On another note, I need a few vector-couples the area between which need to be shadded, on the same plot. How can I incorporate that?
Good file, but the colour disappears when I switch to a vertical log scale. How can I fix this?
Thanks, exactly what I needed!
Also thanks to Chris: put data in row vectors!
if you make I run the following example, matlab logs me off my user account on SuSe 10.2
% copy the jbfill.m example from the comments here.
Very good file, thank you.
lovely function but it is removing box around figure as consequence of transparency.
Great job. Thanks for all.
Just what I needed. Thanks!
Exactly what I needed. Thanks for posting!
Very nice. Thanks!
Works great, but apparently only when datapoint are horizontally organized (one row). If it's a column it doesn't work. You could mention this in the file.
Great! Solves a common problem.
Added a retroactive BSD license
Download apps, toolboxes, and other File Exchange content using Add-On Explorer in MATLAB.