How to show bin counts' percentage values in Histogram2 tile plot?

Asked by Luna on 4 Nov 2019

Latest activity Commented on by Luna on 27 Nov 2019 at 8:19

35 views (last 30 days)

Hi Dear Community

I have x and y values*, I want to plot it like a 3D Histogram but on the top view. (It should look like a heatmap with a colorbar).

I made the plot, but I want to show the values percentages or counts on the each rectangle. Here is the code below:

hFig = figure;
hAxes = axes(hFig);
xVals = randn(1,10000);
yVals = randn(1,10000)*5;
xEdges = min(xVals):1:max(xVals);
yEdges = min(yVals):5:max(yVals);
hHist = histogram2(hAxes,xVals,yVals,xEdges,yEdges,'DisplayStyle','tile');
colorbar;

This code only plots the tiled histogram. I looked the hHist object and saw that it has the BinCounts property which actually gives the number of points in each rectangle.

So I calculated percentages as follows:

percentagesMatrix = (hHist.BinCounts / sum(hHist.BinCounts,'all'))*100;

I don't know how to update the rectangles with this percentage values. I want all that matrix is written inside the boxes. And update the colorbar accordingly.

Any help will be appreciated.

Thank you so much! :)

*Note: x and y values normally comes from a tall table as a tall array.

Here is the sample plot below. I want to see numbers inside these squares:

0 Comments

Products

MATLAB

Release

R2018b

2 Answers

Link

Answer by Luna on 21 Nov 2019 at 16:03

Edited by Luna on 21 Nov 2019 at 16:11

Accepted Answer

I have solved it writing a function manually add text to each rectangle.

Here is the function:

function hHist = normalized_histogram2(hAxes,xVals,yVals,xEdges,yEdges)
% Creates a 2D Histogram plot of data counts percentage in each bins.
% Add percentage values as text inside each bin. 
% Font size and font weight can be changed manually inside the code or
% added to function inputs. 
% NOTE: If text does not fit the rectangles, change font size or maximize
% figure plot. 
%% inputs: 
% hAxes - axis handle object for the plot
% xVals - a double vector of x values
% yVals - a double vector of y values
% NOTE: xVals and yVals must be the same length
% xEdges - a double vector of x bins
% yEdges - a double vector of y bins
%% outputs:
% hHist - histogram object handle
hist_Counts = histcounts2(xVals,yVals,xEdges,yEdges);
hist_Counts_normalized = (hist_Counts / sum(hist_Counts,'all'))*100;
hHist = histogram2(hAxes,'XBinEdges',xEdges,'YBinEdges',yEdges,'DisplayStyle','tile','BinCounts',hist_Counts_normalized);
colorbar;
xBinCenters = hHist.XBinEdges - [0,diff(hHist.XBinEdges)/2];
yBinCenters = hHist.YBinEdges - [0,diff(hHist.YBinEdges)/2];
for i = 2:numel(xBinCenters)
    
    for j = 2:numel(yBinCenters)
        x_loc = xBinCenters(i);
        y_loc = yBinCenters(j);
        if ~(round(hHist.Values(i-1,j-1),2) == 0)
            text(hAxes,x_loc,y_loc,[num2str(round(hHist.Values(i-1,j-1),2)),'%'],'Color',[0 0 0],'FontSize',10,'FontWeight','bold');
        else
            continue
        end
        
    end
end

Here is the demo:

%% Data Creation:
xVals = randn(1,10000);
yVals = randn(1,1000)*5;
xEdges = min(xVals):1:max(xVals);
yEdges = min(yVals):5:max(yVals);
%% Demo:
hFig = figure;
hAxes = axes(hFig);
normalized_histogram2(hAxes,xVals,yVals,xEdges,yEdges);
% make plot look beautiful:
xlabel('X Data');
ylabel('Y Data');
title('Normalized Histogram 2');
hFig.WindowState = 'maximized';

It will look like below:

0 Comments

Link

Answer by Guillaume on 4 Nov 2019

It's all explained in the doc of histogram2. You don't even have to make the percentage calculation yourself, histogram2 will do it for you if you specify 'Normalization', 'probability':

hHist = histogram2(hAxes,xVals,yVals,xEdges,yEdges,'DisplayStyle','tile', 'Normalization', 'probability');

5 Comments

Show 2 older comments

Luna on 21 Nov 2019 at 16:04

Yes, I wrote an answer below. You can check it.

Shashank Bhatia on 22 Nov 2019 at 5:10

Thanks it works for me!!

Best Regards,

SHashank

Luna on 27 Nov 2019 at 8:19

Your welcome ;)

Products

MATLAB

Release

R2018b

You are now following this question

Luna (view profile)