how to stack hist?

Does anyone knows how to stack 2 histogram. I have certain area data for 20 years. I am calculating yearly number of events by hist command. And also I have another area data for 20 years. Doing the same, calculating yearly number of events by hist command for the area also. Note for 2 data time the same, 1980-2000 in years.

Now I wanted 2 hist data stack. Is there any way to do it?

Thanks!

[José-Luis]

What do you mean by stack?

[davit petraasya]

adding y values, y1 and y2 values. Notice they are different size of vectors.

[José-Luis]

I still don't get it. Could you show a figure as an example? And some data?

[Stephen Licata]

That is a very good and fun example - many thanks!

BTW, if you want a special color, like 'gray' (which does not have a Matlab shortcut symbol code), do this bar command

bar(binrng,counts2,'FaceColor', [0.75 0.75 0.75])

which is equivalent to this version using the RGB integer values:

bar(binrng,counts2,'FaceColor', [192 192 192]/255)

Answer 1

[Star Strider]

The hist function does not offer a 'stacked' option, but you can create the effect easily enough with the histc function and a bar plot.

Experiment with this to get the result you want with your data:

d1 = randi(9, 50, 1);                                           % Create Data
d2 = randi(9, 50, 1);                                           % Create Data
binrng = 1:9;                                                   % Create Bin Ranges

counts1 = histc(d1, binrng);                                    % Histogram For ‘d1’
counts2 = histc(d2, binrng);                                    % Histogram For ‘d2’
counts3 = counts1 + counts2;                                    % Histogram Sum ‘d1’+‘d2’

figure(1)
bar(binrng, counts3, 'b')
hold on
bar(binrng, counts1, 'y')
hold off
legend('Data 1', 'Data 2')

The idea is straightforward: create histogram counts for both sets of data, add them, then use the bar plot to first plot the sum, then overplot with one of the others. That will create your stacked histogram plot.

[davit petraasya]

Thanks Strider, this is exactly what I needed. One thing more to ask when you impose binrng are you giving command the number of bins? If I want 20 bins it will be binrng=1:20 ?

[Star Strider]

My pleasure.

‘If I want 20 bins it will be binrng=1:20 ?’

Yes. Just change the ‘binrng’ assignment to that, assign your data as ‘d1’ and ‘d2’ (duplicating the assignment for the purposes of plotting it), and you can run my code without any further changes, unless you also want to change the colours of the bars.

[davit petraasya]

Okay, Thanks Strider, It worked perfectly!

[Star Strider]

My pleasure.

[Wesser]

I know this is an old question, but can you do something similar if d1 and d2 are different length vectors? I also want to stack histograms, but each histogram does not have the same number of data points.

Thanks!

[Star Strider]

@Jessica Lynne Meeks —

The lengths of the vectors should not matter, so long as they are both column vectors.

Example —

d1 = randi(9, 50, 1);                                           % Create Data
d2 = randi(9, 90, 1);                                           % Create Data
binrng = 1:9;                                                   % Create Bin Ranges
counts1 = histc(d1, binrng);                                    % Histogram For ‘d1’
counts2 = histc(d2, binrng);                                    % Histogram For ‘d2’
counts3 = counts1 + counts2;                                    % Histogram Sum ‘d1’+‘d2’

figure(1)
bar(binrng, counts3, 'b')
hold on
bar(binrng, counts1, 'y')
hold off
legend('Data 1', 'Data 2')

.

[Wesser]

Thanks!

[Star Strider]

My pleasure!

[Jordan]

Hi, I know this is an old question too. But what if there are 6 separate vectors I want stacked? I can ask it as a new question if you would prefer that. Thanks!

[Star Strider]

That’s a bit ambiguous.

Perhaps:

v1 = randi(9,1,50);
v2 = randi(9,1,50);
v3 = randi(9,1,50);
v4 = randi(9,1,50);
v5 = randi(9,1,50);
v6 = randi(9,1,50);
vm = [v1; v2; v3; v4; v5; v6];
binrng = 1:9;

for k = 1:size(vm,1)
    counts(k,:) = histc(vm(k,:), binrng);
end

figure
bar(binrng,counts,'stacked')
grid

Experiment, if necessary, to get different results.

.

[Joris Bockhofer]

Make sure to use histc!

MatLab recommended using histcounts instead but this will not work because histcounts yields one less value than #bins for whatever reason.