How to visualize two histograms for comparative analysis?

Basically I have two datasets and I want to perform a comparative analysis by showing how many data fall under a specific range. Any idea as how to do it?

Thanks Alex

[Pieter VL]

You can change the transparancy of the bars of the second histogram using the 'FaceAlpha' parameter to create a visible overlay as follows:

hist1 = histogram( ..., 'EdgeColor', 'blue', 'FaceColor', 'blue' );

hold on;

hist2 = histogram( ..., 'EdgeColor', 'green', 'FaceColor', 'green', 'FaceAlpha', 0.2 );

[Anupam Sharma]

This is not working in following code. I am using it exactly as mentioned above but it shows the second histogram only. Can you help ?

% MATLAB R2019a
% Setup
N = [1:5 10 20 40];
LB = 0;
UB = 3;
n = 10000;

% Generate random variates
X = LB + (UB - LB)*rand(max(N),n);
Sn = cumsum(X); 

mu = 1.5;
sigma = .75;
S_1 = mu + sigma.*randn(n, 1)

hist1= histogram(Sn(1,:),'Normalization','pdf','EdgeColor', 'blue', 'FaceColor',  'blue')
hold on
hist2 = histogram(S_1(:), 'EdgeColor', 'green', 'FaceColor',  'green', 'FaceAlpha', 0.2);

[Image Analyst]

Use subplot to put them into different axes.

Answer 1

[Image Analyst]

The histograms will tell you that - how many counts in each range. Use histogram() for visualization, or histcounts() and bar().

[Alex]

Thank you Image Analyst. Indeed, I have two datasets (Nx1). For both the datasets, I want to see the number of counts in a specific range. If I use hold on and plot the histograms, it gets overlapped and it is not easy to visualize the difference.

So, I was wondering if I can introduce some sort of transparency or sort of 3D plot.

[the cyclist]

Here is an example of using Image Analyst's idea of using both histcounts and bar:

x = randn(2000,1);
y = 0.1 + randn(2000,1);

binRange = -3:0.5:3;

hcx = histcounts(x,[binRange Inf]);
hcy = histcounts(y,[binRange Inf]);

figure
bar(binRange,[hcx;hcy]')

[Alex]

Thanks a lot "the cyclist".

Answer 2

[the cyclist]

As shown in the documentation for the histogram histogram, you can plot two overlapping histograms on one figure like this

x = randn(2000,1);
y = 1 + randn(5000,1);
h1 = histogram(x);
hold on
h2 = histogram(y);

If you have the Statistics and Machine Learning Toolbox, you can also use the Kolmogorov-Smirnov test to determine whether the CDFs of the two distributions are statistically different.

[Alex]

Thank you cyclist for the answer. Yes, I was using the hold on command. Unfortunately, it is difficult to visualize when I want to check number of counts in a range. So, I wondering if I could introduce any transparency or maybe plot 3D?

[Milagre MANHIQUE]

Hallo Matlab community

I would like to plot two histograms from two different datasets on the same graph (overlapping and transparent), but each of the histograms containing the respective fit curve.

I tried to use the histfit command on each dataset but was unsuccessful.

I ask for help please

[Image Analyst]

Try histcounts() and bar()

[Milagre MANHIQUE]

Thank you very much, I will proceed as you advise.

I thank you deeply.

Milagre MANHIQUE